[Federal Register Volume 85, Number 84 (Thursday, April 30, 2020)]
[Rules and Regulations]
[Pages 24174-25278]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2020-06967]



[[Page 24173]]

Vol. 85

Thursday,

No. 84

April 30, 2020

Part IV

Book 2 of 3 Books

Pages 24173-25278





Environmental Protection Agency





-----------------------------------------------------------------------



40 CFR Parts 86 and 600





Department of Transportation





-----------------------------------------------------------------------



National Highway Traffic Safety Administration

49 CFR Parts 523, 531, 533, et al.



-----------------------------------------------------------------------



The Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model 
Years 2021-2026 Passenger Cars and Light Trucks; Final Rule

  Federal Register / Vol. 85 , No. 84 / Thursday, April 30, 2020 / 
Rules and Regulations  

[[Page 24174]]


-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 86 and 600

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration

49 CFR Parts 523, 531, 533, 536, and 537

[NHTSA-2018-0067; EPA-HQ-OAR-2018-0283; FRL 10000-45-OAR]
RIN 2127-AL76; RIN 2060-AU09


The Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for 
Model Years 2021-2026 Passenger Cars and Light Trucks

AGENCY: Environmental Protection Agency and National Highway Traffic 
Safety Administration.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: EPA and NHTSA, on behalf of the Department of Transportation, 
are issuing final rules to amend and establish carbon dioxide and fuel 
economy standards. Specifically, EPA is amending carbon dioxide 
standards for model years 2021 and later, and NHTSA is amending fuel 
economy standards for model year 2021 and setting new fuel economy 
standards for model years 2022-2026. The standards set by this action 
apply to passenger cars and light trucks, and will continue our 
nation's progress toward energy independence and carbon dioxide 
reduction, while recognizing the realities of the marketplace and 
consumers' interest in purchasing vehicles that meet all of their 
diverse needs. These final rules represent the second part of the 
Administration's action related to the August 24, 2018 proposed Safer 
Affordable Fuel-Efficient (SAFE) Vehicles Rule. These final rules 
follow the agencies' actions, taken September 19, 2019, to ensure One 
National Program for automobile fuel economy and carbon dioxide 
emissions standards, by finalizing regulatory text related to 
preemption under the Energy Policy and Conservation Act and withdrawing 
a waiver previously provided to California under the Clean Air Act.

DATES: This final rule is effective on June 29, 2020.
    Judicial Review: NHTSA and EPA undertake this joint action under 
their respective authorities pursuant to the Energy Policy and 
Conservation Act and the Clean Air Act. Pursuant to CAA section 307(b), 
42 U.S.C. 7607(b), any petitions for judicial review of this action 
must be filed in the United States Court of Appeals for the D.C. 
Circuit. Given the inherent relationship between the agencies' action, 
any challenges to NHTSA's regulation under 49 U.S.C. 32909 should also 
be filed in the United States Court of Appeals for the D.C. Circuit.

ADDRESSES: EPA and NHTSA have established dockets for this action under 
Docket ID Nos. EPA-HQ-OAR-2018-0283 and NHTSA-2018-0067, respectively. 
All documents in the docket are listed in the http://www.regulations.gov index. Although listed in the index, some 
information is not publicly available, e.g., confidential business 
information (CBI) or other information whose disclosure is restricted 
by statute. Certain other material, such as copyrighted material, will 
be publicly available in hard copy in EPA's docket, and electronically 
in NHTSA's online docket. Publicly available docket materials can be 
found either electronically in www.regulations.gov by searching for the 
dockets using the Docket ID numbers above, or in hard copy at the 
following locations:
    EPA: EPA Docket Center, EPA/DC, EPA West, Room 3334, 1301 
Constitution Ave. NW, Washington, DC. The Public Reading Room is open 
from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal 
holidays. The telephone number for the Public Reading Room is (202) 
566-1744.
    NHTSA: Docket Management Facility, M-30, U.S. Department of 
Transportation (DOT), West Building, Ground Floor, Rm. W12-140, 1200 
New Jersey Ave. SE, Washington, DC 20590. The DOT Docket Management 
Facility is open between 9 a.m. and 5 p.m. Eastern Time, Monday through 
Friday, except Federal holidays.

FOR FURTHER INFORMATION CONTACT: EPA: Christopher Lieske, Office of 
Transportation and Air Quality, Assessment and Standards Division, 
Environmental Protection Agency, 2000 Traverwood Drive, Ann Arbor, MI 
48105; telephone number: (734) 214-4584; fax number: (734) 214-4816; 
email address: [email protected], or contact the Assessment 
and Standards Division, email address: [email protected]. NHTSA: James Tamm, 
Office of Rulemaking, Fuel Economy Division, National Highway Traffic 
Safety Administration, 1200 New Jersey Avenue SE, Washington, DC 20590; 
telephone number: (202) 493-0515.

SUPPLEMENTARY INFORMATION: 

Does this action apply to me?

    This action affects companies that manufacture or sell new light-
duty vehicles, light-duty trucks, and medium-duty passenger vehicles, 
as defined under EPA's CAA regulations,\1\ and passenger automobiles 
(passenger cars) and non-passenger automobiles (light trucks) as 
defined under NHTSA's CAFE regulations.\2\ Regulated categories and 
entities include:
---------------------------------------------------------------------------

    \1\ ``Light-duty vehicle,'' ``light-duty truck,'' and ``medium-
duty passenger vehicle'' are defined in 40 CFR 86.1803-01. Generally 
speaking, a ``light-duty vehicle'' is a passenger car, a ``light-
duty truck'' is a pick-up truck, sport-utility vehicle, or minivan 
up to 8,500 lbs. gross vehicle weight rating, and a ``medium-duty 
passenger vehicle'' is a sport-utility vehicle or passenger van from 
8,500 to 10,000 lbs. gross vehicle weight rating.
    \2\ ``Passenger car'' and ``light truck'' are defined in 49 CFR 
part 523.

---------------------------------------------------------------------------

[[Page 24175]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.000

    This list is not intended to be exhaustive, but rather provides a 
guide regarding entities likely to be regulated by this action. To 
determine whether particular activities may be regulated by this 
action, you should carefully examine the regulations. You may direct 
questions regarding the applicability of this action to the person 
listed in FOR FURTHER INFORMATION CONTACT.

I. Executive Summary
II. Overview of Final Rule
III. Purpose of the Rule
IV. Purpose of Analytical Approach Considered as Part of Decision-
Making
V. Regulatory Alternatives Considered
VI. Analytical Approach as Applied to Regulatory Alternatives
VII. What does the analysis show, and what does it mean?
VIII. How do the final standards fulfill the agencies' statutory 
obligations?
IX. Compliance and Enforcement
X. Regulatory Notices and Analyses

I. Executive Summary

    NHTSA (on behalf of the Department of Transportation) and EPA are 
issuing final rules to adopt and modify standards regulating corporate 
average fuel economy and tailpipe carbon dioxide (CO2) 
emissions and use/leakage of other air conditioning refrigerants for 
passenger cars and light trucks for MYs 2021-2026.\3\ These final rules 
follow the proposal issued in August 2018 and respond to each agency's 
legal obligation to set standards based on the factors Congress 
directed them to consider, as well as the direction of the United 
States Supreme Court in Massachusetts v. EPA, which stated that ``there 
is no reason to think the two agencies cannot both administer their 
obligations and yet avoid inconsistency.'' \4\ These standards are the 
product of significant and ongoing work by both agencies to craft 
regulatory requirements for the same group of vehicles and vehicle 
manufacturers. This work aims to facilitate, to the extent possible 
within the statutory directives issued to each agency, the ability of 
automobile manufacturers to meet all requirements under both programs 
with a single national fleet under one national program of fuel economy 
and tailpipe CO2 emission regulation.
---------------------------------------------------------------------------

    \3\ Throughout this document and the accompanying FRIA, the 
agencies will often use the term ``CO2'' or ``tailpipe 
CO2'' to refer broadly to EPA's suite of light duty 
vehicle GHG standards.
    \4\ 549 U.S. 497, 532 (2007).
---------------------------------------------------------------------------

    The CAFE and CO2 emissions standards established by 
these final rules will increase in stringency at 1.5 percent per year 
from MY 2020 levels over MYs 2021-2026. The ``1.5 percent'' regulatory 
alternative is new for the final rule and was not expressly analyzed in 
the NPRM, but it is a logical outgrowth of the NPRM analysis, being 
well within the range of alternatives then considered and consistent 
with discussions by both the agencies and commenters that there are 
benefits to having standards that increase at the same rate for all 
fleets. These standards apply to light-duty vehicles, which NHTSA 
divides for purposes of regulation into passenger cars and light 
trucks, and EPA divides into passenger cars, light-duty trucks, and 
medium-duty passenger vehicles (i.e., sport utility vehicles, cross-
over utility vehicles, and light trucks). Both the CAFE and 
CO2 standards are vehicle-footprint-based, as are the 
standards currently in effect. These standards will become more 
stringent for each model year from 2021 to 2026, relative to the MY 
2020 standards. Generally, the larger the vehicle footprint, the less 
numerically stringent the corresponding vehicle CO2 and 
miles-per-gallon (mpg) targets. As a result of the footprint-based 
standards, the burden of compliance is distributed across all vehicle 
footprints and across all manufacturers. Each manufacturer is subject 
to individualized standards for passenger cars and light trucks, in 
each model year, based on the vehicles it produces. When standards are 
carefully crafted, both in terms of the footprint curves and the rate 
of increase in stringency of those curves, manufacturers are not

[[Page 24176]]

compelled to build vehicles of any particular size or type.
    Knowing that many readers are accustomed to considering CAFE and 
CO2 emissions standards in terms of the mpg and grams-per-
mile (g/mi) values that the standards are projected to eventually 
require, the agencies include those projections here. EPA's standards 
are projected to require, on an average industry fleet-wide basis, 201 
grams per mile (g/mi) of CO2 in model year 2030, while 
NHTSA's standards are projected to require, on an average industry 
fleet-wide basis, 40.5 miles per gallon (mpg) in model year 2030. The 
agencies note that real-world CO2 is typically 25 percent 
higher and real-world fuel economy is typically 20 percent lower than 
the CO2 and CAFE compliance values discussed here, and also 
note that a portion of EPA's expected ``CO2'' improvements 
will in fact be made through improvements in minimizing air 
conditioning leakage and through use of alternative refrigerants, which 
will not contribute to fuel economy but will contribute toward 
reductions of climate-related emissions.
    In these final rules, NHTSA and EPA are reaching similar 
conclusions on similar grounds: even though each agency has its own 
distinct statutory authority and factors, the relevant considerations 
overlap in many ways. Both agencies recognize that they are balancing 
the relevant considerations in somewhat different ways from how they 
may have been balanced previously, as in the 2012 final rule and in 
EPA's Initial Determination, but the current balancing is called for in 
light of the facts before the agencies. The balancing in these final 
rules is also somewhat different from how the agencies balanced their 
respective considerations in the proposal, in part because of updates 
to analytical inputs and methodologies, previewed in the NPRM and made 
in response to public comments, that collectively resulted in changes 
to the analytical outputs. For example, between the notice and final 
rule, the agencies updated fuel price projections to somewhat greater 
values, updated the analysis fleet to MY 2017, updated estimates of the 
efficacy and cost of fuel-saving technologies, revised procedures for 
calculating impacts on vehicle sales and scrappage, updated models for 
estimating highway safety impacts, updated estimates of highway 
congestion costs, and updated estimates of annual mileage accumulation, 
holding VMT (before applying the rebound effect) constant between 
regulatory alternative. Moreover, the cost-benefit analysis conducted 
for these final rules has even been overtaken by events in many ways 
over recent weeks. Based upon current events, and for additional 
reasons discussed in Section VI.D.1 the benefits of saving additional 
fuel through more stringent standards are potentially even smaller than 
estimated in this rulemaking analysis.
    The standards finalized today fit the pattern of gradual, tough, 
but feasible stringency increases that take into account real world 
performance, shifts in fuel prices, and changes in consumer behavior 
toward crossovers and SUVs and away from more efficient sedans. This 
approach ensures that manufacturers are provided with sufficient lead 
time to achieve standards, considering the cost of compliance. The 
costs to both industry and automotive consumers would have been too 
high under the standards set forth in 2012, and by lowering the auto 
industry's costs to comply with the program, with a commensurate 
reduction in per-vehicle costs to consumers, the standards enhance the 
ability of the fleet to turn over to newer, cleaner and safer vehicles.
    More stringent standards also have the potential for overly 
aggressive penetration rates for advanced technologies relative to the 
penetration rates seen in the final standards, especially in the face 
of an unknown degree of consumer acceptance of both the increased costs 
and of the technologies themselves--particularly given current 
projections of relatively low fuel prices during that timeframe. As a 
kind of insurance policy against future fuel price volatility, 
standards that increase at 1.5 percent per year for cars and trucks 
will help to keep fleet fuel economy higher than they would be 
otherwise when fuel prices are low, which is not improbable over the 
next several years.\5\ At the same time, the standards help to address 
these issues by maintaining incentives to promote broader deployment of 
advanced technologies, and so provides a means of encouraging their 
further penetration while leaving manufacturers alternative technology 
choices. Steady, gradual increases in stringency ensure that the 
benefits of reduced GHG emissions and fuel consumption are achieved 
without the potential for disruption to automakers or consumers.
---------------------------------------------------------------------------

    \5\ For example, EIA currently expects U.S. retail gasoline 
prices to average $2.14/gallon in 2020, compared to $2.69/gallon in 
2019 (see https://www.eia.gov/outlooks/steo/archives/mar20.pdf), and 
$3.68/gallon in 2012 (see https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=EMM_EPM0_PTE_NUS_DPG&f=A). While gasoline 
prices may foreseeably rise over the rulemaking time frame, it is 
also very foreseeable that they will not rise to the $4-5/gallon 
that many Americans saw over the 2008-2009 time frame, that caused 
the largest shift seen toward smaller and higher-fuel-economy 
vehicles. See, e.g., Figure VIII-2 below.
---------------------------------------------------------------------------

    Standards that increase at 1.5 percent per year represent a 
reasonable balance of additional technology and required per-vehicle 
costs, consumer demand for fuel economy, fuel savings and emissions 
avoided given the foreseeable state of the global oil market and the 
minimal effect on climate between finalizing 1.5 percent standards 
versus more stringent standards. The final standards will also result 
in year-over-year improvements in fleetwide fuel economy, resulting in 
energy conservation that helps address environmental concerns, 
including criteria pollutant, air toxic pollutant, and carbon 
emissions.
    The agencies project that under these final standards, required 
technology costs would be reduced by $86 to $126 billion over the 
lifetimes of vehicles through MY 2029. Equally important, purchase 
prices costs to U.S. consumers for new vehicles would be $977 to $1,083 
lower, on average, than they would have been if the agencies had 
retained the standards set forth in the 2012 final rule and originally 
upheld by EPA in January 2017. While these final standards are 
estimated to result in 1.9 to 2.0 additional billion barrels of fuel 
consumed and from 867 to 923 additional million metric tons of 
CO2 as compared to current estimates of what the standards 
set forth in 2012 would require, the agencies explain at length below 
why the overall benefits of the final standards outweigh these 
additional costs.\6\
---------------------------------------------------------------------------

    \6\ 1.9 to 2.0 barrels of fuel is approximately 78 to 84 gallons 
of fuel.
---------------------------------------------------------------------------

    For the CAFE program, overall (fleetwide) net benefits vary from 
$16.1 billion at a 7 percent discount rate to -$13.1 billion at a 3 
percent discount rate. For the CO2 program, overall 
(fleetwide) societal net benefits vary from $6.4 billion at a 7 percent 
discount rate to -$22.0 billion at a 3 percent discount rate. The net 
benefits straddle zero, and are very small relative to the scale of 
reduced required technology costs, which range from $86.3 billion to 
$126.0 billion for the CAFE and CO2 programs across 7 
percent and 3 percent discount rates. Likewise, net benefits are very 
small relative to the scale of reduced retail fuel savings over the 
full life of all vehicles manufactured during the 2021 through 2029 
model years, which range from $108.6 billion to $185.1 billion for the 
CAFE and CO2 programs across 7 percent and 3 percent 
discount rates. Similarly, all of the alternatives have small net 
benefits, ranging from $18.4 billion to -$31.1

[[Page 24177]]

billion for the CAFE and CO2 programs across 7 percent and 3 
percent discount rates.\7\
---------------------------------------------------------------------------

    \7\ See Table II-12 to Table II-15 for costs, benefits and net 
benefits.
---------------------------------------------------------------------------

    NHTSA and EPA believe their analysis of the final rule represents 
the best available science, evidence, and methodologies for assessing 
the impacts of changes in CAFE and CO2 emission standards. 
In fact, the agencies note that today's analysis represents a marked 
improvement over prior rulemakings. Previously, the agencies were 
unable to model the impact of the standards on new vehicle sales or the 
retirement of older vehicles in the fleet, and, instead, were forced to 
assume, contrary to economic theory and empirical evidence, that the 
number of new vehicles sold and older vehicles scrapped remained static 
across regulatory alternatives. Today's analysis--as commenters to 
previous rulemakings and EPA's Science Advisory Board have argued is 
necessary \8\--quantifies the sales and scrappage impacts of the 
standards, including the associated safety benefits, and represents a 
significant step forward in agencies' ability to comprehensively 
analyze the impacts of CAFE and CO2 emission standards.
---------------------------------------------------------------------------

    \8\ Science Advisory Board, U.S. EPA. Review of EPA's Proposed 
SAFE rule at 4 (Feb. 27, 2020), available at https://
yosemite.epa.gov/sab/sabproduct.nsf/LookupWebProjectsCurrentBOARD/
1FACEE5C03725F268525851F006319BB/$File/EPA-SAB-20-003+.pdf 
[hereinafter ``SAB Report''].
---------------------------------------------------------------------------

    However, the agencies also believe it is important to be 
transparent about analytical limitations. For example, EPA's Science 
Advisory Board stressed that the agencies account for ``evolving 
consumer preferences for performance and other vehicle attributes,'' 
\9\ yet due to limitations on the agencies' current ability to model 
buyers' choices among combinations of various attributes and their 
costs, the primary analysis does not account for the consumer benefits 
of other vehicle features that may be sacrificed for costly 
technologies that improve fuel economy. The agencies' analysis assumes 
that under these final standards, attributes of new cars and light 
trucks other than fuel economy would remain identical to those under 
the baseline standards, so that changes in sales prices and fuel 
economy would be the only sources of benefits or costs to new car and 
light truck buyers. In other words, the agencies' primary analysis does 
not consider that producers will likely respond to buyers' demands by 
reallocating some their savings in production costs due to lower 
technology costs to add or improve other attributes that consumers 
value more highly than the increases in fuel economy the augural 
standards would have required. The agencies have long debated whether 
and how best to model the consumer benefits of other vehicle 
attributes, and note that they have made considerable progress.\10\ 
However, despite these potential analytical shortcomings, the agencies 
reaffirm that today's analysis represents the most complete and 
rigorous examination of CAFE and CO2 emission standards to 
date, and provide decision-makers a powerful analytical tool--
especially since the limitations are known, do not bias the central 
analysis' results, and are afforded due consideration.
---------------------------------------------------------------------------

    \9\ SAB at 10.
    \10\ In their evaluations of previous CAFE and CO2 
rules, the agencies attempted to account for this possibility by 
conducting sensitivity analyses that reduced the fuel savings and 
other benefits to vehicle buyers by a significant fraction. For 
example, NHTSA's analysis supporting the Final Rule establishing 
CAFE standards for model year 2012-16 cars and light trucks tested 
the sensitivity of their central estimates of social costs and 
benefits to the assumptions that 25 percent and 50 percent of 
benefits to buyers were offset by opportunity costs of foregone 
improvements in attributes other than fuel economy; see NHTSA, Final 
Regulatory Impact Analysis: Corporate Average Fuel Economy for Model 
year 2012-16 Passenger Cars and Light Trucks, March 2010, at 563-565 
and Table X-9, at 566-56; see also, NHTSA, Final Regulatory Impact 
Analysis: Corporate Average Fuel Economy for Model year 2017-25 
Passenger Cars and Light Trucks, August 2012, at 1087 and Tables X-
18a, X-18b, and X-18c, at 1099-1104. The agencies acknowledged that 
this was not a completely satisfactory way to represent the 
sacrifices in vehicles' other attributes that car and light truck 
manufacturers might find it necessary to make in order to comply 
with the increasingly stringent standards those previous rules 
established. At the time, however, the agencies were unable to 
identify specific attributes that manufacturers were most likely to 
sacrifice, measure the tradeoffs between increased fuel economy and 
improvements in those attributes, or assess the potential losses in 
utility to car and light truck buyers. In an effort to improve on 
their previous treatment of this issue, the agencies' evaluation of 
this final rule includes a sensitivity case that assumes 
manufacturers redirect their technology cost savings from complying 
with less stringent standards to instead improve a combination of 
cars' and light trucks' other attributes that offers benefits to 
their buyers significantly exceeding those costs. The magnitude of 
these (net) benefits is interpreted as the opportunity cost of the 
improvements in vehicles' other attributes that would have been 
sacrificed if the augural standards had been enacted. The method the 
agencies use to approximate these benefits, together with its effect 
on the rule's overall benefits and costs, is discussed in detail in 
Section VI.D.1.b)(8). Briefly, the results of this sensitivity 
analysis suggest the Final Rule would generate net benefits for the 
CAFE and CO2 programs ranging from $34.9 to $55.4 billion 
at 3% and 7% discount rates.
---------------------------------------------------------------------------

    In terms of the agencies' respective statutory authorities, EPA is 
setting national tailpipe CO2 emissions standards for 
passenger cars and light trucks under section 202(a) of the Clean Air 
Act (CAA),\11\ and taking other actions under its authority to 
establish metrics and measure passenger car and light truck fleet fuel 
economy pursuant to the Energy Policy and Conservation Act (EPCA),\12\ 
while NHTSA is setting national corporate average fuel economy (CAFE) 
standards under EPCA, as amended by the Energy Independence and 
Security Act (EISA) of 2007.\13\ As summarized above and as discussed 
in much greater detail below, the agencies believe that these represent 
appropriate levels of CO2 emissions standards and maximum 
feasible CAFE standards for MYs 2021-2026, pursuant to their respective 
statutory authorities. Sections III and VIII below contain detailed 
discussions of both agencies' statutory obligations and authorities.
---------------------------------------------------------------------------

    \11\ 42 U.S.C. 7521(a).
    \12\ 49 U.S.C. 32904(c).
    \13\ 49 U.S.C. 32902.
---------------------------------------------------------------------------

    Section 202(a) of the CAA requires EPA to establish standards for 
emissions of pollutants from new motor vehicles that cause or 
contribute to air pollution that may reasonably be anticipated to 
endanger public health or welfare. Standards under section 202(a) thus 
take effect only ``after providing such period as the Administrator 
finds necessary to permit the development and application of the 
requisite technology, giving appropriate consideration to the cost of 
compliance within such period.'' \14\ In establishing such standards, 
EPA must consider issues of technical feasibility, cost, and available 
lead time, among other things.
---------------------------------------------------------------------------

    \14\ CAA Sec. 202(a); 42 U.S.C. 7512(a)(2).
---------------------------------------------------------------------------

    EPCA, as amended by EISA, contains a number of provisions governing 
how NHTSA must set CAFE standards. EPCA requires that the Department of 
Transportation establish separate passenger car and light truck 
standards \15\ at ``the maximum feasible average fuel economy level 
that the Secretary decides the manufacturers can achieve in that model 
year,'' \16\ based on the agency's consideration of four statutory 
factors: technological feasibility, economic practicability, the effect 
of other standards of the Government on fuel economy, and the need of 
the United States to conserve energy.\17\ EPCA does not define these 
terms or specify what weight to give each concern in balancing them--
such considerations are left within the discretion of the Secretary of 
Transportation (delegated to NHTSA) based upon current information. 
Accordingly, NHTSA interprets these factors and determines the 
appropriate weighting that leads to the maximum

[[Page 24178]]

feasible standards given the circumstances present at the time of 
promulgating each CAFE standard rulemaking. While EISA, for MYs 2011-
2020, additionally required that standards increase ``ratably'' and be 
set at levels to ensure that the CAFE of the industry-wide combined 
fleet of new passenger cars and light trucks reach at least 35 mpg by 
MY 2020,\18\ EISA requires that standards for MYs 2021-2030 simply be 
set at the maximum feasible level as determined by the Secretary (and 
by delegation, NHTSA).\19\
---------------------------------------------------------------------------

    \15\ 49 U.S.C. 32902(b)(1).
    \16\ 49 U.S.C. 32902(a).
    \17\ 49 U.S.C. 32902(f).
    \18\ 49 U.S.C. 32902(b)(2)(A) and (C).
    \19\ 49 U.S.C. 32902(b)(2)(B).
---------------------------------------------------------------------------

    In the NPRM, the agencies sought comment on a variety of possible 
changes to existing compliance flexibilities that have been created 
over the past several years. The vast majority of the existing 
compliance flexibilities are not being changed, but a small number of 
flexibilities related to real-world fuel efficiency improvements are 
being finalized. In addition, EPA will continue to allow manufacturers 
to make improvements relating to air conditioning refrigerants and 
leakage and will credit those improvements toward CO2 
compliance, and EPA is making no changes in the amounts of credits 
available. EPA is also not making any changes to the existing 
CH4 and N2O standards. EPA is also extending the 
``0 g/mi upstream'' incentive for electric vehicles beyond its current 
sunset of MY 2021, through MY 2026. EPA is also establishing a credit 
multiplier for natural gas vehicles through the 2026 model year. 
Otherwise, compliance flexibilities in the two programs do not change 
significantly for the final rule. These changes should help to 
streamline manufacturer use of those flexibilities in certain respects. 
While manufacturers and suppliers sought a number of other additional 
compliance flexibilities, the agencies have concluded that the 
aforementioned existing flexibilities are reasonable and appropriate, 
and that additional flexibilities are not justified.
    Table I-1 and Table I-2 present the total costs, benefits, and net 
benefits for the 2021-2026 preferred alternative CAFE and 
CO2 levels, relative to the MY 2022-2025 existing/augural 
standards (with the MY 2025 standards repeated for MY 2026) and current 
MY 2021 standard. The preferred alternative exhibits a stringency rate 
increase of 1.5 percent per year for both passenger cars and light 
trucks. The values in Table I-1 and Table I-2 display (in total and 
annualized forms) costs for all MYs 1978-2029 vehicles, and the 
benefits and net benefits represent the impacts of the standards over 
the full lifetimes of the vehicles sold or projected to be sold during 
model years 1978-2029.
    For this analysis, negative signs are used for changes in costs or 
benefits that decrease from those that would have resulted from the 
existing/augural standards. Any changes that would increase either 
costs or benefits are shown as positive changes. Thus, an alternative 
that decreases both costs and benefits, will show declines (i.e., a 
negative sign) in both categories. From Table I-1 and Table I-2, the 
preferred alternative (Alternative 3) is estimated to decrease costs 
relative to the baseline by $182 to $280 billion over the lifetime of 
MYs 1978-2029 passenger vehicles (range determined by discount rate 
across both CAFE and CO2 programs). It will also decrease 
benefits from $175 to $294 billion over the life of these MY fleets. 
The net impact will be a decrease from $22 billion to an increase of 
$16 billion in total net benefits to society over this roughly 52-year 
timeframe. Annualized, this amounts to roughly -$0.8 to 1.2 billion in 
net benefits per year.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.001

[GRAPHIC] [TIFF OMITTED] TR30AP20.002


[[Page 24179]]


    Table I-3 and Table I-4 lists costs, benefits, and net benefits for 
all seven alternatives that were examined.
[GRAPHIC] [TIFF OMITTED] TR30AP20.003

[GRAPHIC] [TIFF OMITTED] TR30AP20.004

    Table I-5 and Table I-6 show a summary of various impacts of the 
preferred alternative for CAFE and CO2 standards. Impacts 
are presented in monetized and non-monetized values, as well as from 
the perspective of society and the consumer.

[[Page 24180]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.005


[[Page 24181]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.006

BILLING CODE 4910-59-C
    The agencies note that the NPRM drew more public comments (and, 
particularly, more pages of substantive comments) than any rulemaking 
in the history of the CAFE or CO2 tailpipe emissions 
programs--exceeding 750,000 comments. The agencies recognized in the 
NPRM that the proposal was significantly different from the final rules 
set forth in 2012, and explained at length the reasons for those 
differences--namely, that new information and considerations, along 
with an expanded and updated analysis, had led to different tentative 
conclusions. Today's final rules represent a further evolution of the 
work that supported the proposal, based on improved quantitative 
methodology and in careful consideration of the hundreds of thousands 
of public comments and deep reflection on the serious issues before the 
agencies. Simply put, the agencies have heard the comments, and today's 
analysis and decision reflect the agencies' grappling with the issues 
commenters raised, as well as all of the other information before the 
agencies. These programs and issues are weighty, and the agencies 
believe that a reasonable balance has been struck in these final rules 
between the many competing national needs that these regulatory 
programs collectively address.

II. Overview of Final Rule

A. Summary of Proposal

    In the NPRM, the National Highway Traffic Safety Administration 
(NHTSA) and the Environmental Protection Agency (EPA) (collectively, 
``the

[[Page 24182]]

agencies'') proposed the ``Safer Affordable Fuel-Efficient (SAFE) 
Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light 
Trucks'' (SAFE Vehicles Rule). The proposed SAFE Vehicles Rule would 
set Corporate Average Fuel Economy (CAFE) and carbon dioxide 
(CO2) emissions standards, respectively, for passenger cars 
and light trucks manufactured for sale in the United States in model 
years (MYs) 2021 through 2026.\20\
---------------------------------------------------------------------------

    \20\ NHTSA sets CAFE standards under the Energy Policy and 
Conservation Act of 1975 (EPCA), as amended by the Energy 
Independence and Security Act of 2007 (EISA). EPA sets 
CO2 standards under the Clean Air Act (CAA).
---------------------------------------------------------------------------

    The agencies explained that they must act to propose and finalize 
these standards and do not have discretion to decline to regulate. 
Congress requires NHTSA to set CAFE standards for each model year.\21\ 
Congress also requires EPA to set emissions standards for light-duty 
vehicles if EPA has made an ``endangerment finding'' that the pollutant 
in question--in this case, CO2--``cause[s] or contribute[s] 
to air pollution which may reasonably be anticipated to endanger public 
health or welfare.'' \22\ NHTSA and EPA proposed the standards 
concurrently because tailpipe CO2 emissions standards are 
directly and inherently related to fuel economy standards,\23\ and, if 
finalized, the rules would apply concurrently to the same fleet of 
vehicles. By working together to develop the proposals, the agencies 
aimed to reduce regulatory burden on industry and improve 
administrative efficiency.
---------------------------------------------------------------------------

    \21\ 49 U.S.C. 32902.
    \22\ 42 U.S.C. 7521; see also 74 FR 66495 (Dec. 15, 2009) 
(``Endangerment and Cause or Contribute Findings for Greenhouse 
Gases under Section 202(a) of the Clean Air Act'').
    \23\ See, e.g., 75 FR 25324, at 25327 (May 7, 2010) (``The 
National Program is both needed and possible because the 
relationship between improving fuel economy and reducing tailpipe 
CO2 emissions is a very direct and close one. The amount 
of those CO2 emissions is essentially constant per gallon 
combusted of a given type of fuel. Thus, the more fuel efficient a 
vehicle is, the less fuel it burns to travel a given distance. The 
less fuel it burns, the less CO2 it emits in traveling 
that distance. [citation omitted] While there are emission control 
technologies that reduce the pollutants (e.g., carbon monoxide) 
produced by imperfect combustion of fuel by capturing or converting 
them to other compounds, there is no such technology for 
CO2. Further, while some of those pollutants can also be 
reduced by achieving a more complete combustion of fuel, doing so 
only increases the tailpipe emissions of CO2. Thus, there 
is a single pool of technologies for addressing these twin problems, 
i.e., those that reduce fuel consumption and thereby reduce 
CO2 emissions as well.'').
---------------------------------------------------------------------------

    The agencies discussed some of the history leading to the proposal, 
including the 2012 final rule, the expectations regarding a mid-term 
evaluation as required by EPA regulation, and the rapid process over 
2016 and early 2017 by which EPA issued its first Final Determination 
that the CO2 standards set in 2012 for MYs 2022-2025 
remained appropriate based on the information then before the EPA 
Administrator.\24\ The agencies also discussed President Trump's 
direction in March 2017 to restore the original mid-term evaluation 
timeline, and EPA's subsequent information-gathering process and 
announcement that it would reconsider the January 2017 
Determination.\25\ EPA ultimately concluded that the standards set in 
2012 for MYs 2022-2025 were no longer appropriate.\26\ For NHTSA, in 
turn, the ``augural'' CAFE standards for MYs 2022-2025 were never 
final, and as explained in the 2012 final rule, NHTSA was obligated 
from the beginning to undertake a new rulemaking to set CAFE standards 
for MYs 2022-2025.
---------------------------------------------------------------------------

    \24\ See 83 FR at 42987 (Aug.24, 2018).
    \25\ Id.
    \26\ 83 FR 16077 (Apr. 2, 2018).
---------------------------------------------------------------------------

    The NPRM thus began the rulemaking process for both agencies to 
establish new standards for MYs 2022-2025 passenger cars and light 
trucks. Standards were concurrently proposed for MY 2026 in order to 
provide regulatory stability for as many years as is legally 
permissible for both agencies together. The NPRM also included revised 
standards for MY 2021 passenger cars and light trucks, because the 
agencies tentatively concluded, based on the information and analysis 
then before them, that the CAFE standards previously set for MY 2021 
were no longer maximum feasible, and the CO2 standards 
previously set for MY 2021 were no longer appropriate. Agencies always 
have authority under the Administrative Procedure Act to revisit 
previous decisions in light of new facts, as long as they provide 
notice and an opportunity for comment, and the agencies stated that it 
is plainly the best practice to do so when changed circumstances so 
warrant.\27\
---------------------------------------------------------------------------

    \27\ See FCC v. Fox Television, 556 U.S. 502 (2009).
---------------------------------------------------------------------------

    The NPRM proposed to maintain the CAFE and CO2 standards 
applicable in MY 2020 for MYs 2021-2026, and took comment on a wide 
range of alternatives, including different stringencies and retaining 
existing CO2 standards and the augural CAFE standards.\28\ 
Table II-1, Table II-2, and Table II-3 show the estimates, under the 
NPRM analysis, of what the MY 2020 CAFE and CO2 curves would 
translate to, in terms of miles per gallon (mpg) and grams per mile (g/
mi), in MYs 2021-2026, as well as the regulatory alternatives 
considered in the NPRM. In addition to retaining the MY 2020 
CO2 standards through MY 2026, EPA proposed and sought 
comment on excluding air conditioning refrigerants and leakage, and 
nitrous oxide and methane emissions for compliance with CO2 
standards after model year 2020, in order to improve harmonization with 
the CAFE program. EPA also sought comment on whether to change existing 
methane and nitrous oxide standards that were finalized in the 2012 
rule. The proposal was accompanied by a 1,600 page Preliminary 
Regulatory Impact Analysis (PRIA) and, for NHTSA, a 500 page Draft 
Environmental Impact Statement (DEIS), with more than 800 pages of 
appendices and the entire CAFE model, including the software source 
code and documentation, all of which were also subject to comment in 
their entirety and all of which received significant comments.
---------------------------------------------------------------------------

    \28\ The agencies noted that this did not mean that the miles 
per gallon and grams per mile levels that were estimated for the MY 
2020 fleet in 2012 would be the ``standards'' going forward into MYs 
2021-2026. Both NHTSA and EPA set CAFE and CO2 standards, 
respectively, as mathematical functions based on vehicle footprint. 
These mathematical functions that are the actual standards are 
defined as ``curves'' that are separate for passenger cars and light 
trucks, under which each vehicle manufacturer's compliance 
obligation varies depending on the footprints of the cars and trucks 
that it ultimately produces for sale in a given model year. It was 
the MY 2020 CAFE and CO2 curves that the agencies 
proposed would continue to apply to the passenger car and light 
truck fleets for MYs 2021-2026. The mpg and g/mi values which those 
curves would eventually require of the fleets in those model years 
would be known for certain only at the ends of each of those model 
years. While it is convenient to discuss CAFE and CO2 
standards as a set ``mpg,'' ``g/mi,'' or ``mpg-e'' number, 
attempting to define those values based on the information then 
before the agency would necessarily end up being inaccurate.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24183]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.007


[[Page 24184]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.008

     
---------------------------------------------------------------------------

    \29\ The carbon dioxide equivalents of air conditioning 
refrigerant leakage, nitrous oxide emissions, and methane emissions 
were included for compliance with the EPA standards for all MYs 
under the baseline/no action alternative in the NPRM. Carbon dioxide 
equivalent is calculated using the Global Warming Potential (GWP) of 
each of the emissions.
    \30\ Beginning in MY 2021, the proposal provided that the GWP 
equivalents of air conditioning refrigerant leakage, nitrous oxide 
emissions, and methane emissions would no longer be able to be 
included with the tailpipe CO2 for compliance with 
tailpipe CO2 standards.

---------------------------------------------------------------------------

[[Page 24185]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.009

BILLING CODE 4910-59-C
    The agencies explained in the NPRM that new information had been 
gathered and new analysis performed since publication of the 2012 final 
rule establishing CAFE and CO2 standards for MYs 2017 and 
beyond and since issuance of the 2016 Draft TAR and EPA's 2016 and 
early 2017 ``mid-term evaluation'' process. This new information and 
analysis helped lead the agencies to the tentative conclusion that 
holding standards constant at MY 2020 levels through MY 2026 was 
maximum feasible, for CAFE purposes, and appropriate, for 
CO2 purposes.
    The agencies further explained that technologies had played out 
differently in the fleet from what the agencies previously assumed: 
That while there remain a wide variety of technologies available to 
improve fuel economy and reduce CO2 emissions, it had become 
clear that there were reasons to temper previous optimism about the 
costs, effectiveness, and consumer acceptance of a number of 
technologies. In addition, over the years between the previous analyses 
and the NPRM, automakers had added considerable amounts of technologies 
to their new vehicle fleets, meaning that the agencies were no longer 
free to make certain assumptions about how some of those technologies 
could be used going forward. For example, some technologies that could 
be used to improve fuel economy and reduce emissions had not been used 
entirely for that purpose, and some of the benefit of these 
technologies had gone instead toward improving other vehicle 
attributes. Other technologies had been tried, and had been met with 
significant customer acceptance issues. The agencies underscored the 
importance of reflecting the fleet as it stands today, with the 
technology it has and as that technology has been used, and considering 
what technology remains on the table at this point, whether and when it 
can realistically be available for widespread use in production, and 
how much it would cost to implement.
    The agencies also acknowledged the math of diminishing returns: As 
CAFE and CO2 emissions standards increase in stringency, the 
benefit of continuing to increase in stringency decreases. In mpg 
terms, a vehicle owner who drives a light vehicle 15,000 miles per year 
(a typical assumption for analytical purposes) \31\ and trades in a 
vehicle with fuel economy of 15 mpg for one with fuel economy of 20 
mpg, will reduce their annual fuel consumption from 1,000 gallons to 
750 gallons--saving 250 gallons annually. If, however, that owner were 
to trade in a vehicle with fuel economy of 30 mpg for one with fuel 
economy of 40 mpg, the owner's annual gasoline consumption would drop 
from 500 gallons/year to 375 gallons/year--only 125 gallons even though 
the mpg improvement is twice as large. Going from 40 to 50 mpg would 
save only 75 gallons/year. Yet each additional fuel economy improvement 
becomes much more expensive as the easiest to achieve low-cost 
technological improvement options are chosen. In CO2 terms, 
if a vehicle emits 300 g/mi CO2,

[[Page 24186]]

a 20 percent improvement is 60 g/mi, so the vehicle would emit 240 g/
mi; but if the vehicle emits 180 g/mi, a 20 percent improvement is only 
36 g/mi, so the vehicle would get 144 g/mi. In order to continue 
achieving similarly large (on an absolute basis) emissions reductions, 
the percentage reduction must also continue to increase.
---------------------------------------------------------------------------

    \31\ A different vehicle-miles-traveled (VMT) assumption would 
change the absolute numbers in the example, but would not change the 
mathematical principles.
---------------------------------------------------------------------------

    Related, average real-world fuel economy is lower than average fuel 
economy required under CAFE and CO2 standards. The 2012 
Federal Register notice announcing augural CAFE and CO2 
standards extending through MY 2025 indicated that, if met entirely 
through the application of fuel-saving technology, the MY 2025 
CO2 standards would result in an average requirement 
equivalent to 54.5 mpg. However, because the CO2 standards 
provide credit for reducing leakage of AC refrigerants and/or switching 
to lower-GWP refrigerants, and these actions do not affect fuel 
economy, the notice explained that the corresponding fuel economy 
requirement (under the CAFE program) would be 49.7 mpg. These estimates 
were based on a market forecast grounded in the MY 2008 fleet. The 
notice also presented analysis using a market forecast grounded in the 
MY 2010 fleet, showing a 48.7 mpg average CAFE requirement.
    In the real world, fuel economy is, on average, about 20% lower 
than as measured under regulatory test procedures. In the real world, 
then, these new standards were estimated to require 39.0-39.8 mpg.
    Today's analysis indicates that the requirements under the 
baseline/augural CAFE standards would average 46.6 mpg in MY 2029. The 
lower value results from changes in the fleet forecast which reflects 
consumer preference for larger vehicles than was forecast for the 2012 
rulemaking. In the real world, the requirements average about 37.1 mpg. 
Under the final standards issued today, the regulatory test procedure 
requirements average 40.5 mpg, corresponding to 33.2 mpg in the real 
world. Buyers of new vehicles experience real-world fuel economy, with 
levels varying among drivers (due to a wide range of factors). Vehicle 
fuel economy labels provide average real-world fuel economy information 
to buyers.
[GRAPHIC] [TIFF OMITTED] TR30AP20.010

    Vehicle owners also face fuel prices at the pump. The agencies 
noted in the NPRM that when fuel prices are high, the value of fuel 
saved may be enough to offset the cost of further fuel economy/
emissions reduction improvements, but the agencies recognized that 
then-current projections of fuel prices by the Energy Information 
Administration did not indicate particularly high fuel prices in the 
foreseeable future. The agencies explained that fundamental structural 
shifts had occurred in global oil markets since the 2012 final rule, 
largely due to the rise of U.S. production and export of shale oil. The 
consequence over time of diminishing returns from more stringent fuel 
economy/emissions reduction standards, especially when combined with 
relatively low fuel prices, is greater difficulty for automakers to 
find a market of consumers willing to buy vehicles that meet the 
increasingly stringent standards. American consumers have long 
demonstrated that in times of relatively low fuel prices, fuel economy 
is not a top priority for the majority of them, even when highly fuel 
efficient vehicle models are available.
    The NPRM analysis sought to improve how the agencies captured the 
effects of higher new vehicle prices on fleet composition as a whole by 
including an improved model for vehicle scrappage rates. As new vehicle 
prices increase, consumers tend to continue using older vehicles for 
longer, slowing fleet turnover and thus slowing improvements in fleet-
wide fuel economy, reductions in CO2 emissions, reductions 
in criteria pollutant emissions, and advances in safety. That aspect of 
the analysis was also driven by the agencies' updated estimates of 
average per-vehicle cost increases due to

[[Page 24187]]

higher standards, which were several hundred dollars higher than 
previously estimated. The agencies cited growing concerns about 
affordability and negative equity for many consumers under these 
circumstances, as loan amounts grow and loan terms extend.
    For all of the above reasons, the agencies proposed to maintain the 
MY 2020 fuel economy and CO2 emissions standards for MYs 
2021-2026. The agencies explained that they estimated, relative to the 
standards for MYs 2021-2026 put forth in 2012, that an additional 0.5 
million barrels of oil would be consumed per day (about 2 to 3 percent 
of projected U.S. consumption) if that proposal were finalized, but 
that they also expected the additional fuel costs to be outweighed by 
the cost savings from new vehicle purchases; that more than 12,700 on-
road fatalities and significantly more injuries would be prevented over 
the lifetimes of vehicles through MY 2029 as compared to the standards 
set forth in the 2012 final rule over the lifetimes of vehicles as more 
new and safer vehicles are purchased than the current (and augural) 
standards; and that environmental impacts, on net, would be relatively 
minor, with criteria and toxic air pollutants not changing noticeably, 
and with estimated atmospheric CO2 concentrations increasing 
by 0.65 ppm (a 0.08 percent increase), which the agencies estimated 
would translate to 0.003 degrees Celsius of additional temperature 
increase relative to the standards finalized in 2012.
    Under the NPRM analysis, the agencies tentatively concluded that 
maintaining the MY 2020 curves for MYs 2021-2026 would save American 
auto consumers, the auto industry, and the public a considerable amount 
of money as compared to EPA retaining the previously-set CO2 
standards and NHTSA finalizing the augural standards. The agencies 
explained that this had been identified as the preferred alternative, 
in part, because it appeared to maximize net benefits compared to the 
other alternatives analyzed, and recognizing the statutory 
considerations for both agencies. Relative to the standards issued in 
2012, under CAFE standards, the NPRM analysis estimated that costs 
would decrease by $502 billion overall at a three-percent discount rate 
($335 billion at a seven-percent discount rate) and benefits were 
estimated to decrease by $326 billion at a three-percent discount rate 
($204 billion at a seven-percent discount rate). Thus, net benefits 
were estimated to increase by $176 billion at a three-percent discount 
rate and $132 billion at a seven-percent discount rate. The estimated 
impacts under CO2 standards were estimated to be similar, 
with net benefits estimated to increase by $201 billion at a three-
percent discount rate and $141 billion at a seven-percent discount 
rate.
    The NPRM also sought comment on a variety of potential changes to 
NHTSA's and EPA's compliance programs for CAFE and CO2 as 
well as related programs, including questions about automaker requests 
for additional flexibilities and agency interest in reducing market-
distorting incentives and improving transparency; and on a proposal to 
withdraw California's CAA preemption waiver for its ``Advanced Clean 
Car'' regulations, with an accompanying discussion of preemption of 
State standards under EPCA.\32\ The agencies sought comment broadly on 
all aspects of the proposal.
---------------------------------------------------------------------------

    \32\ Agency actions relating to California's CAA waiver and EPCA 
preemption have since been finalized, see 84 FR 51310 (Sept. 27, 
2019), and will not be discussed in great detail as part of this 
final rule.
---------------------------------------------------------------------------

B. Public Participation Opportunities and Summary of Comments

    The NPRM was published on NHTSA's and EPA's websites on August 2, 
2018, and published in the Federal Register on August 24, 2018, 
beginning a 60-day comment period. The agencies subsequently extended 
the official comment period for an additional three days, and left the 
dockets open for more than a year after the start of the comment 
period, considering late comments to the extent practicable. A separate 
Federal Register notice also published on August 24, 2018, which 
announced the locations, dates, and times of three public hearings to 
be held on the proposal: One in Fresno, California, on September 24, 
2018; one in Dearborn, Michigan, on September 25, 2018; and one in 
Pittsburgh, Pennsylvania, on September 26, 2018. Each hearing started 
at 10 a.m. local time; the Fresno hearing ended at 5:10 p.m. and 
resulted in a 235 page transcript; the Dearborn hearing ran until 5:26 
p.m. and resulted in a 330 page transcript; and the Pittsburgh hearing 
ran until 5:06 p.m. and also resulted in a 330 page transcript. Each 
hearing also collected several hundred pages of comments from 
participants, in addition to the hearing transcripts.
    Besides the comments submitted as part of the public hearings, 
NHTSA's docket received a total of 173,359 public comments in response 
to the proposal as of September 18, 2019, and EPA's docket a total of 
618,647 public comments, for an overall total of 792,006. NHTSA also 
received several hundred comments on its DEIS to the separate DEIS 
docket. While the majority of individual comments were form letters, 
the agencies received over 6,000 pages of substantive comments on the 
proposal.
    Many commenters generally supported the proposal and many 
commenters opposed it. Commenters supporting the proposal tended to 
cite concerns about the cost of new vehicles, while commenters opposing 
the proposal tended to cite concerns about additional fuel expenditures 
and the impact on climate change. Many comments addressed the modeling 
used for the analysis, and specifically the inclusion, operation, and 
results of the sales and scrappage modules that were part of the NPRM's 
analysis, while many addressed the NPRM's safety findings and the role 
that those findings played in the proposal's justification. Many other 
comments addressed California's standards and role in Federal decision-
making; as discussed above, those comments are further summarized and 
responded to in the separate Federal Register notice published in 
September 2019. Nearly every aspect of the NPRM's analysis and 
discussion received some level of comment by at least one commenter. 
The comments received, as a whole, were both broad and deep, and the 
agencies appreciate the level of engagement of commenters in the public 
comment process and the information and opinions provided.

C. Changes in Light of Public Comments and New Information

    The agencies made a number of changes to the analysis between the 
NPRM and the final rule in response to public comments and new 
information that was received in those comments or otherwise became 
available to the agencies. While these changes, their rationales, and 
their effects are discussed in detail in the sections below, the 
following represents a high-level list of some of the most significant 
changes:
     Some regulatory alternatives were dropped from 
consideration, and one was added;
     updated analysis fleet, and changes to technologies on 
``baseline'' vehicles within the fleet to reflect better their current 
properties and improve modeling precision;
     no civil penalties assumed to be paid after MY 2020 under 
CAFE program;
     updates and expansions in accounting for certain over-
compliance

[[Page 24188]]

credits, including early credits earned in EPA's program;
     updates and expansions to CAFE Model's technology paths;
     updates to inputs defining the range of manufacturer-, 
technology-, and product-specific constraints;
     updates to allow the model to adopt a more advanced 
technology if it is more cost-effective than an earlier technology on 
the path;
     precision improvements to the modeling of A/C efficiency 
and off-cycle credits;
     updates to model's ``effective cost'' metric;
     extended explicit simulation of technology application 
through MY 2050;
     expanded presentation of the results to include ``calendar 
year'' analysis;
     quantifying different types of health impacts from changes 
in air pollution, rather than only accounting for such impacts in 
aggregate estimates of the social costs of air pollution;
     updated costs to 2018 dollars;
     updated fuel costs based on the AEO 2019 version of NEMS;
     a variety of technology updates in response to comments 
and new information;
     updated accounting of rebound VMT between regulatory 
alternatives;
     updated estimates of the macroeconomic cost of petroleum 
dependence;
     updated response of total new vehicle sales to increases 
in fuel efficiency and price; and
     updated response of vehicle retirement rates to changes in 
new vehicle fuel efficiency and transaction price.
    Sections IV and VI below discuss these updates in significant 
detail.

D. Final Standards--Stringency

    As explained above, the agencies have chosen to set CAFE and 
CO2 standards that increase in stringency by 1.5 percent 
year over year for MYs 2021-2026. Separately, EPA has decided to retain 
the A/C refrigerant and leakage and CH4 and N2O 
standards set forth in 2012 for MYs 2021 and beyond, and the stringency 
of the CO2 standards in this final rule reflect the 
``offset'' also established in 2012 based on assumptions made at that 
time about anticipated HFC emissions reductions.
    When the agencies state that stringency will increase at 1.5 
percent per year, that means that the footprint curves which actually 
define the standards for CAFE and CO2 emissions will become 
more stringent at 1.5 percent per year. Consistent with Congress's 
direction in EISA to set CAFE standards based on a mathematical 
formula, which EPA harmonized with for the CO2 emissions 
standards, the standard curves are equations, which are slightly 
different for CAFE and CO2, and within each program, 
slightly different for passenger cars and light trucks. Each program 
has a basic equation for a fleet standard, and then values that change 
to cause the stringency changes are the coefficients within the 
equations. For passenger cars, consistent with prior rulemakings, NHTSA 
is defining fuel economy targets as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.011

where:

TARGETFE is the fuel economy target (in mpg) applicable to a 
specific vehicle model type with a unique footprint combination,
a is a minimum fuel economy target (in mpg),
b is a maximum fuel economy target (in mpg),
c is the slope (in gallons per mile per square foot, or gpm, per 
square foot) of a line relating fuel consumption (the inverse of 
fuel economy) to footprint, and
d is an intercept (in gpm) of the same line.

    Here, MIN and MAX are functions that take the minimum and maximum 
values, respectively, of the set of included values. For example, 
MIN[40,35] = 35 and MAX(40, 25) = 40, such that MIN[MAX(40, 25), 35] = 
35.
    For light trucks, also consistent with prior rulemakings, NHTSA is 
defining fuel economy targets as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.012

where:

TARGETFE is the fuel economy target (in mpg) applicable to a 
specific vehicle model type with a unique footprint combination,
a, b, c, and d are as for passenger cars, but taking values specific 
to light trucks,
e is a second minimum fuel economy target (in mpg),
f is a second maximum fuel economy target (in mpg),
g is the slope (in gpm per square foot) of a second line relating 
fuel consumption (the inverse of fuel economy) to footprint, and
h is an intercept (in gpm) of the same second line.

    The final CAFE standards (described in terms of their footprint-
based curves) are as follows, with the values for the coefficients 
changing over time:

[[Page 24189]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.013

    These equations are presented graphically below, where the x-axis 
represents vehicle footprint and the y-axis represents fuel economy, 
showing that in the CAFE context, targets are higher (fuel economy) for 
smaller footprint vehicles and lower for larger footprint vehicles:
BILLING CODE 4910-59-C

[[Page 24190]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.014

BILLING CODE 4910-59-P

[[Page 24191]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.015

BILLING CODE 4910-59-C
    EPCA, as amended by EISA, requires that any manufacturer's 
domestically-manufactured passenger car fleet must meet the greater of 
either 27.5 mpg on average, or 92 percent of the average fuel economy 
projected by the Secretary for the combined domestic and non-domestic 
passenger automobile fleets manufactured for sale in the U.S. by all 
manufacturers in the model year, which projection shall be published in 
the Federal Register when the standard for that model year is 
promulgated in accordance with 49 U.S.C. 32902(b).\33\ Any time NHTSA 
establishes or changes a passenger car standard for a model year, the 
MDPCS for that model year must also be evaluated or re-evaluated and 
established accordingly. Thus, this final rule establishes the 
applicable MDPCS for MYs 2021-2026. Table II-8 lists the minimum 
domestic passenger car standards.
---------------------------------------------------------------------------

    \33\ 49 U.S.C. 32902(b)(4).
    [GRAPHIC] [TIFF OMITTED] TR30AP20.016
    
    EPA CO2 standards are as follows. Rather than expressing 
these standards as linear functions with accompanying minima and 
maxima, similar to the approach NHTSA has followed since 2005 in 
specifying attribute-based standards, the following tables specify flat 
standards that apply below and above specified footprints, and a linear 
function that applies between those footprints. The two approaches are 
mathematically identical. For passenger cars with a footprint of less 
than or equal to 41 square feet, the gram/mile CO2 target 
value is selected for the appropriate model year from Table II-9:

[[Page 24192]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.017

    For passenger cars with a footprint of greater than 56 square feet, 
the gram/mile CO2 target value is selected for the 
appropriate model year from Table II-10:
[GRAPHIC] [TIFF OMITTED] TR30AP20.018

    For passenger cars with a footprint that is greater than 41 square 
feet and less than or equal to 56 square feet, the gram/mile 
CO2 target value is calculated using the following equation 
and rounded to the nearest 0.1 grams/mile.


[[Page 24193]]


Target CO2 = [a x f] + b

Where f is the vehicle footprint and a and b are selected from Table 
II-11 for the appropriate model year:
[GRAPHIC] [TIFF OMITTED] TR30AP20.019

    For light trucks with a footprint of less than or equal to 41 
square feet, the gram/mile CO2 target value is selected for 
the appropriate model year from Table II-12:

[[Page 24194]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.020

    For light trucks with a footprint greater than the minimum value 
specified in the table below for each model year, the gram/mile 
CO2 target value is selected for the appropriate model year 
from Table II-13:
[GRAPHIC] [TIFF OMITTED] TR30AP20.021


[[Page 24195]]


    For light trucks with a footprint that is greater than 41 square 
feet and less than or equal to the maximum footprint value specified in 
Table II-14 below for each model year, the gram/mile CO2 
target value is calculated using the following equation and rounded to 
the nearest 0.1 grams/mile.

Target CO2 = (a x f) + b

Where f is the footprint and a and b are selected from Table II-14 
below for the appropriate model year:
[GRAPHIC] [TIFF OMITTED] TR30AP20.022

    These equations are presented graphically below, where the x-axis 
represents vehicle footprint and the y-axis represents the 
CO2 target. The targets are lower for smaller footprint 
vehicles and higher for larger footprint vehicles:
BILLING CODE 4910-59-P

[[Page 24196]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.023


[[Page 24197]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.024

BILLING CODE 4910-59-C
    Except that EPA elected to apply a slightly different slope when 
defining passenger car targets, CO2 targets may be expressed 
as direct conversion of fuel economy targets, as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.025

where 8887 g/gal relates grams of CO2 emitted to gallons 
of fuel consumed, and OFFSET reflects the fact that that HFC 
emissions from lower-GWP A/C refrigerants and less leak-prone A/C 
systems are counted toward average CO2 emissions, but 
EPCA provides no basis to count reduced HFC emissions toward CAFE 
levels.

    For the reader's benefit, Table II-15, Table II-16, and Table II-17 
show the estimates, under the final rule analysis, of what the MYs 
2021-2026 CAFE and CO2 curves would translate to, in terms 
of miles per gallon (mpg) and grams per mile (g/mi).
BILLING CODE 4910-59-P

[[Page 24198]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.026

BILLING CODE 4910-59-C

[[Page 24199]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.027

    As the following tables demonstrate, averages of manufacturers' 
estimated requirements are more stringent (i.e., for CAFE, higher, and 
for CO2, lower) under the final standards than under the 
proposed standards:
[GRAPHIC] [TIFF OMITTED] TR30AP20.028


[[Page 24200]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.029

E. Final Standards--Impacts

    This section summarizes the estimated costs and benefits of the MYs 
2021-2026 CAFE and CO2 emissions standards for passenger 
cars and light trucks, as compared to the regulatory alternatives 
considered. These estimates helped inform the agencies' choices among 
the regulatory alternatives considered and provide further confirmation 
that the final standards are maximum feasible, for NHTSA, and 
appropriate, for EPA. The costs and benefits estimated to result from 
the CAFE standards are presented first, followed by those estimated to 
result from the CO2 standards. For several reasons, the 
estimates for costs and benefits presented for the different programs, 
while consistent, are not identical. NHTSA's and EPA's standards are 
projected to result in slightly different fuel efficiency improvements. 
EPA's CO2 standard is nominally more stringent in part due 
to its assumptions about manufacturers' use of air conditioning 
leakage/refrigerant replacement credits, which are expected to result 
in reduced emissions of HFCs. NHTSA's final standards are based solely 
on assumptions about fuel economy improvements, and do not account for 
emissions reductions that do not relate to fuel economy. In addition, 
the CAFE and CO2 programs offer somewhat different program 
flexibilities and provisions, primarily because NHTSA is statutorily 
prohibited from considering some flexibilities when establishing CAFE 
standards, while EPA is not.\34\ The analysis underlying this final 
rule reflects many of those additional EPA flexibilities, which 
contributes to differences in how the agencies estimate manufacturers 
could comply with the respective sets of standards, which in turn 
contributes to differences in estimated impacts of the standards. These 
differences in compliance flexibilities are discussed in more detail in 
Section IX below.
---------------------------------------------------------------------------

    \34\ See 49 U.S.C. 32902(h); CAA Sec. 202(a).
---------------------------------------------------------------------------

    Table II-20 to Table II-23 present all subcategories of costs and 
benefits of this final rule for all seven alternatives proposed. Costs 
include application of fuel economy technology to new vehicles, 
consumer surplus, crash costs due to changes in VMT, as well as, noise 
and congestion. Benefits include fuel savings, consumer surplus, 
refueling time, and clean air.
BILLING CODE 4910-59-P

[[Page 24201]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.030


[[Page 24202]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.031


[[Page 24203]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.032


[[Page 24204]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.033


[[Page 24205]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.034


[[Page 24206]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.035


[[Page 24207]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.036


[[Page 24208]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.037


[[Page 24209]]


BILLING CODE 4910-59-C

F. Other Programmatic Elements

1. Compliance and Flexibilities
    Automakers seeking to comply with the CAFE and CO2 
standards are generally expected to add fuel economy-improving 
technologies to their new vehicles to boost their overall fleet fuel 
economy levels. Readers will remember that improving fuel economy 
directly reduces CO2 emissions, because CO2 is a 
natural and inevitable byproduct of fossil fuel combustion to power 
vehicles. The CAFE and CO2 programs contain a variety of 
compliance provisions and flexibilities to accommodate better 
automakers' production cycles, to reward real-world fuel economy 
improvements that cannot be reflected in the 1975-developed test 
procedures, and to incentivize the production of certain types of 
vehicles. While the agencies sought comment on a broad variety of 
changes and potential expansions of the programs' compliance 
flexibilities in the NPRM, the agencies determined, after considering 
the comments, to make a few changes to the flexibilities proposed in 
the NPRM in this final rule. The most noteworthy change is the 
retention, in the CO2 program, of the flexibilities that 
allow automakers to continue to use HFC reductions toward their 
CO2 compliance, and that extend the ``0 grams/mile'' 
assumption for electric vehicles through MY 2026 (i.e., recognizing 
only the tailpipe emissions of full battery-electric vehicles and not 
recognizing the upstream emissions caused by the electricity usage of 
those vehicles). In the NPRM, EPA had proposed to remove and sought 
comment on removing those flexibilities from the CO2 
program, but determined not to remove them in this final rule. EPA and 
NHTSA are also removing from the programs, starting in MY 2022, the 
credit/FCIV for full-size pickup trucks that are either hybrids or 
over-performing by a certain amount relative to their targets, and 
allowing technology suppliers to begin the petition process for off-
cycle credits/adjustments.
    Table II-24, Table II-25, Table II-26, and Table II-27 provide a 
summary of the various compliance provisions in the two programs; their 
authorities; and any changes included as part of this final rule:
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.038


[[Page 24210]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.039

BILLING CODE 4910-59-P

[[Page 24211]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.040

     
---------------------------------------------------------------------------

    \35\ The CAFE program uses an energy efficiency metric and 
standards that are expressed in miles per gallon. For PHEVs and 
BEVs, to determine gasoline the equivalent fuel economy for 
operation on electricity, a Petroleum Equivalency Factor (PEF) is 
applied to the measured electrical consumption. The PEF for 
electricity was established by the Department of Energy, as required 
by statute, and includes an accounting for upstream energy 
associated with the production and distribution for electricity 
relative to gasoline. Therefore, the CAFE program includes upstream 
accounting based on the metric that is consistent with the fuel 
economy metric. The PEF for electricity also includes an incentive 
that effectively counts only 15 percent of the electrical energy 
consumed.
---------------------------------------------------------------------------

    Providing a technology neutral basis by which manufacturers meet 
fuel economy and CO2 emissions standards encourages an 
efficient and level playing field. The agencies continue to have a 
desire to minimize incentives that disproportionately favor one 
technology over another. Some of this may involve regulations 
established by other Federal agencies. In the near future, NHTSA and 
EPA intend to work with other relevant Federal agencies to pursue 
regulatory means by which we can further ensure technology neutrality 
in this field.
2. Preemption/Waiver
    As discussed above, the issues of Clean Air Act waivers of 
preemption under Section 209 and EPCA/EISA preemption under 49 U.S.C. 
32919 are not addressed in today's final rule, as

[[Page 24212]]

they were the subject of a separate final rulemaking action by the 
agencies in September 2019. While many comments were received in 
response to the NPRM discussion of those issues, those comments have 
been addressed and responded to as part of that separate rulemaking 
action.

III. Purpose of the Rule

    The Administrative Procedure Act (APA) requires agencies to 
incorporate in their final rules a ``concise general statement of their 
basis and purpose.'' \36\ While the entire preamble document represents 
the agencies' overall explanation of the basis and purpose for this 
regulatory action, this section within the preamble is intended as a 
direct response to that APA (and related CAA) requirements. Executive 
Order 12866 further states that ``Federal agencies should promulgate 
only such regulations as are required by law, are necessary to 
interpret the law, or are made necessary by compelling public need, 
such as material failures of private markets to protect or improve the 
health and safety of the public, the environment, or the well-being of 
the American people.'' \37\ Section III.C of the FRIA accompanying this 
rulemaking discusses at greater length the question of whether a market 
failure exists that these final rules may address.
---------------------------------------------------------------------------

    \36\ 5 U.S.C. 553(c); see also Clean Air Act section 
307(d)(6)(A), 42 U.S.C. 7607(d)(6)(A).
    \37\ E.O. 12866, Section 1(a).
---------------------------------------------------------------------------

    NHTSA and EPA are legally obligated to set CAFE and GHG standards, 
respectively, and do not have the authority to decline to regulate.\38\ 
The agencies are issuing these final rules to fulfill their respective 
statutory obligations to provide maximum feasible fuel economy 
standards and limit emissions of pollutants from new motor vehicles 
which have been found to endanger public health and welfare (in this 
case, specifically carbon dioxide (CO2); EPA has already set 
standards for methane (CH4), nitrous oxide (N2O), 
and hydrofluorocarbons (HFCs) and is not revising them in this rule). 
Continued progress in meeting these statutory obligations is both 
legally necessary and good for America--greater energy security and 
reduced emissions protect the American public. The final standards 
continue that progress, albeit at a slower rate than the standards 
finalized in 2012.
---------------------------------------------------------------------------

    \38\ For CAFE, see 49 U.S.C. 32902; for CO2, see 42 
U.S.C. 7521(a).
---------------------------------------------------------------------------

    National annual gasoline consumption and CO2 emissions 
currently total about 140 billion gallons and 5,300 million metric 
tons, respectively. The majority of this gasoline (about 130 billion 
gallons) is used to fuel passenger cars and light trucks, such as will 
be covered by the CAFE and CO2 standards issued today. 
Accounting for both tailpipe emissions and emissions from ``upstream'' 
processes (e.g., domestic refining) involved in producing and 
delivering fuel, passenger cars and light trucks account for about 
1,500 million metric tons (mmt) of current annual CO2 
emissions. The agencies estimate that under the standards issued in 
2012, passenger car and light truck annual gasoline consumption would 
steadily decline, reaching about 80 billion gallons by 2050. The 
agencies further estimate that, because of this decrease in gasoline 
consumption under the standards issued in 2012, passenger car and light 
truck annual CO2 emissions would also steadily decline, 
reaching about 1,000 mmt by 2050. Under the standards issued today, the 
agencies estimate that, instead of declining from about 140 billion 
gallons annually today to about 80 billion gallons annually in 2050, 
passenger car and light truck gasoline consumption would decline to 
about 95 billion gallons. The agencies correspondingly estimate that 
instead of declining from about 1,500 mmt annually today to about 1,000 
mmt annually in 2050, passenger car and light truck CO2 
emissions would decline to about 1,100 mmt. In short, the agencies 
estimate that under the standards issued today, annual passenger car 
and light truck gasoline consumption and CO2 emissions will 
continue to steadily decline over the next three decades, even if not 
quite as rapidly as under the previously-issued standards.
    The agencies also estimate that these impacts on passenger car and 
light truck gasoline consumption and CO2 emissions will be 
accompanied by a range of other energy- and climate-related impacts, 
such as reduced electricity consumption (because today's standards 
reduce the estimated rate at which the market might shift toward 
electric vehicles) and increased CH4 and N2O 
emissions. These estimated impacts, discussed below and in the FEIS 
accompanying today's notice, are dwarfed by estimated impacts on 
gasoline consumption and CO2 emissions.
    As explained above, these final rules set or amend fuel economy and 
carbon dioxide standards for model years 2021-2026. Many commenters 
argued that it was not appropriate to amend previously-established 
CO2 and CAFE standards, generally because those commenters 
believed that the administrative record established for the 2012 final 
rule and EPA's January 2017 Final Determination was superior to the 
record that informed the NPRM, and that that prior record led 
necessarily to the policy conclusion that the previously-established 
standards should remain in place.\39\ Some commenters similarly argued 
that EPA's Revised Final Determination--which, for EPA, preceded this 
regulatory action--was invalid because, they allege, it did not follow 
the procedures established for the mid-term evaluation that EPA 
codified into regulation,\40\ and also because the Revised Final 
Determination was not based on the prior record.\41\
---------------------------------------------------------------------------

    \39\ Comments arguing that the prior record was superior to the 
current record, and thus a better basis for decision-making, will be 
addressed throughout the balance of this preamble.
    \40\ 40 CFR 86.1818-12(h).
    \41\ See, e.g., comments from the States and Cities, Attachment 
1, Docket No. NHTSA-2018-0067-11735, at 40-42; CARB, Detailed 
Comments, Docket No. NHTSA-2018-0067-11873, at 71-72; CBD et. al, 
Appendix A, Docket No. NHTSA-2018-0067-12000, at 214-228.
---------------------------------------------------------------------------

    The agencies considered a range of alternatives in the proposal, 
including the baseline/no action alternative of retaining the existing 
EPA carbon dioxide standards. As the agencies explained in the 
proposal, the proposal was entirely de novo, based on an entirely new 
analysis reflecting the best and most up-to-date information available 
to the agencies.\42\ This rulemaking action is separate and distinct 
from EPA's Revised Final Determination, which itself was neither a 
proposed nor a final decision that the standards ``must'' be revised. 
EPA retained full discretion in this rulemaking to revise the standards 
or not revise them. In any event, the case law is clear that agencies 
are free to reconsider their prior decisions.\43\ With that legal 
principle in mind, the agencies agree with commenters that the amended 
(and new) CO2 and CAFE standards must be consistent with the

[[Page 24213]]

CAA and EPCA/EISA, respectively, and this preamble and the accompanying 
FRIA explain in detail why the agencies believe they are consistent. 
The section below discusses briefly the authority given to the agencies 
by their respective governing statutes, and the factors that Congress 
directed the agencies to consider as they exercise that authority in 
pursuit of fulfilling their statutory obligations.
---------------------------------------------------------------------------

    \42\ 83 FR 42968, 42987 (Aug. 24, 2018).
    \43\ See, e.g., Encino Motorcars, LLC v. Navarro, 136 S. Ct. 
2117, 2125 (2016) (``Agencies are free to change their existing 
policies as long as they provide a reasoned explanation for the 
change.''); FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 
(2009) (When an agency changes its existing position, it ``need not 
always provide a more detailed justification than what would suffice 
for a new policy created on a blank slate. Sometimes it must--when, 
for example, its new policy rests on factual findings that 
contradict those which underlay its prior policy; or when its prior 
policy has engendered serious reliance interests that must be taken 
into account . . . . In such cases it is not that further 
justification is demanded by the mere fact of policy change, but 
that a reasoned explanation is needed for disregarding facts and 
circumstances that underlay or were engendered by the prior 
policy.'')
---------------------------------------------------------------------------

A. EPA's Statutory Requirements

    EPA is setting national CO2 standards for passenger cars 
and light trucks under Section 202(a) of the Clean Air Act (CAA).\44\ 
Section 202(a) of the CAA requires EPA to establish standards for 
emissions of pollutants from new motor vehicles which cause or 
contribute to air pollution which may reasonably be anticipated to 
endanger public health or welfare.\45\ In establishing such standards, 
EPA considers issues of technical feasibility, cost, available lead 
time, and other factors. Standards under section 202(a) thus take 
effect only ``after providing such period as the Administrator finds 
necessary to permit the development and application of the requisite 
technology, giving appropriate consideration to the cost of compliance 
within such period.'' \46\ EPA's statutory requirements are further 
discussed in Section VIII.A.
---------------------------------------------------------------------------

    \44\ 42 U.S.C. 7521(a).
    \45\ See Coalition for Responsible Regulation v. EPA, 684 F.3d 
102, 114-115 (D.C. Cir. 2012) (`` `If EPA makes a finding of 
endangerment, the Clean Air Act requires the [a]gency to regulate 
emissions of the deleterious pollutant from new motor vehicles . . . 
. Given the non-discretionary duty in Section 202(a)(1) and the 
limited flexibility available under Section 202(a)(2), which this 
court has held related only to the motor vehicle industry, . . . EPA 
had no statutory basis on which it could ground [any] reasons for 
further inaction' '') (quoting Massachusetts v. EPA, 549 U.S. 497, 
533-35 (2007).
    \46\ 42 U.S.C. 7521(a)(2).
---------------------------------------------------------------------------

B. NHTSA's Statutory Requirements

    NHTSA is setting national Corporate Average Fuel Economy (CAFE) 
standards for passenger cars and light trucks for each model year as 
required under EPCA, as amended by EISA.\47\ EPCA mandates a motor 
vehicle fuel economy regulatory program that balances statutory factors 
in setting minimum fuel economy standards to facilitate energy 
conservation. EPCA allocates the responsibility for implementing the 
program between NHTSA and EPA as follows: NHTSA sets CAFE standards for 
passenger cars and light trucks; EPA establishes the procedures for 
testing, tests vehicles, collects and analyzes manufacturers' data, and 
calculates the individual and average fuel economy of each 
manufacturer's passenger cars and light trucks; and NHTSA enforces the 
standards based on EPA's calculations.
---------------------------------------------------------------------------

    \47\ EPCA and EISA direct the Secretary of Transportation to 
develop, implement, and enforce fuel economy standards (see 49 
U.S.C. 32901 et. seq.), which authority the Secretary has delegated 
to NHTSA at 49 CFR 1.94(c).
---------------------------------------------------------------------------

    The following sections enumerate specific statutory requirements 
for NHTSA in setting CAFE standards and NHTSA's interpretations of 
them, where applicable. Many comments were received on these 
requirements and interpretations. Because this is intended as an 
overview section, those comments will be addressed below in Section 
VIII rather than here, and the agencies refer readers to that part of 
the document for more information.
    For each future model year, EPCA (as amended by EISA) requires that 
DOT (by delegation, NHTSA) establish separate passenger car and light 
truck standards at ``the maximum feasible average fuel economy level 
that the Secretary decides the manufacturers can achieve in that model 
year,'' \48\ based on the agency's consideration of four statutory 
factors: ``technological feasibility, economic practicability, the 
effect of other motor vehicle standards of the Government on fuel 
economy, and the need of the United States to conserve energy.'' \49\ 
The law also allows NHTSA to amend standards that are already in place, 
as long as doing so meets these requirements.\50\ EPCA does not define 
these terms or specify what weight to give each concern in balancing 
them; thus, NHTSA defines them and determines the appropriate weighting 
that leads to the maximum feasible standards given the circumstances in 
each CAFE standard rulemaking.\51\
---------------------------------------------------------------------------

    \48\ 49 U.S.C. 32902(a) and (b).
    \49\ 49 U.S.C. 32902(f).
    \50\ 49 U.S.C. 32902(g).
    \51\ See Center for Biological Diversity v. NHTSA, 538 F.3d 
1172, 1195 (9th Cir. 2008) (hereafter ``CBD v. NHTSA'') (``The EPCA 
clearly requires the agency to consider these four factors, but it 
gives NHTSA discretion to decide how to balance the statutory 
factors--as long as NHTSA's balancing does not undermine the 
fundamental purpose of the EPCA: Energy conservation.'')
---------------------------------------------------------------------------

    EISA added several other requirements to the setting of separate 
passenger car and light truck standards. Standards must be ``based on 1 
or more vehicle attributes related to fuel economy and express[ed] . . 
. in the form of a mathematical function.'' \52\ New standards must 
also be set at least 18 months before the model year in question, as 
would amendments to increase standards previously set.\53\ NHTSA must 
regulations prescribing average fuel economy standards for at least 1, 
but not more than 5, model years at a time.\54\ A number of comments 
addressed these requirements; for the reader's reference, those 
comments will be summarized and responded to in Section VIII. EISA also 
added the requirement that NHTSA set a minimum standard for 
domestically-manufactured passenger cars,\55\ which will also be 
discussed further in Section VIII below.
---------------------------------------------------------------------------

    \52\ 49 U.S.C. 32902(b)(3)(A).
    \53\ 49 U.S.C. 32902(a), (g)(2).
    \54\ 49 U.S.C. 39202(b)(3)(B).
    \55\ 49 U.S.C. 32902(b)(4).
---------------------------------------------------------------------------

    For MYs 2011-2020, EISA further required that the separate 
standards for passenger cars and for light trucks be set at levels high 
enough to ensure that the achieved average fuel economy for the entire 
industry-wide combined fleet of new passenger cars and light trucks 
reach at least 35 mpg not later than MY 2020, and standards for those 
years were also required to ``increase ratably.'' \56\ For model years 
after 2020, standards must be set at the maximum feasible level.\57\
---------------------------------------------------------------------------

    \56\ 49 U.S.C. 32902(b)(2)(A) and (C). NHTSA has CAFE standards 
in place that are projected to result in industry-achieved fuel 
economy levels over 35 mpg in MY 2020. EPA typically provides 
verified final CAFE data from manufacturers to NHTSA several months 
or longer after the close of the MY in question, so the actual MY 
2020 fuel economy will not be known until well after MY 2020 has 
ended. The standards for all MYs up to and including 2020 are known 
and not at issue in this regulatory action, so these provisions are 
noted for completeness rather than immediate relevance to this final 
rule. Because neither of these requirements apply after MY 2020, 
they are not relevant to this rulemaking and will not be discussed 
further.
    \57\ 49 U.S.C. 32902(b)(2)(B).
---------------------------------------------------------------------------

1. Factors That Must Be Considered in Deciding What Levels of CAFE 
Standards are ``Maximum Feasible''
(a) Technological Feasibility
    ``Technological feasibility'' refers to whether a particular method 
of improving fuel economy can be available for commercial application 
in the model year for which a standard is being established. Thus, in 
determining the level of new standards, the agency is not limited to 
technology that is already being commercially applied at the time of 
the rulemaking. For this rulemaking, NHTSA has evaluated and considered 
all types of technologies that improve real-world fuel economy, 
although not every possible technology was expressly included in the 
analysis, as discussed in Section VI and also in Section VIII.
(b) Economic Practicability
    ``Economic practicability'' refers to whether a standard is one 
``within the

[[Page 24214]]

financial capability of the industry, but not so stringent as to'' lead 
to ``adverse economic consequences, such as a significant loss of jobs 
or the unreasonable elimination of consumer choice.'' \58\ The agency 
has explained in the past that this factor can be especially important 
during rulemakings in which the automobile industry is facing 
significantly adverse economic conditions (with corresponding risks to 
jobs). Economic practicability is a broad factor that includes 
considerations of the uncertainty surrounding future market conditions 
and consumer demand for fuel economy in addition to other vehicle 
attributes.\59\ In an attempt to evaluate the economic practicability 
of different future levels of CAFE standards (i.e., the regulatory 
alternatives considered in this rulemaking), NHTSA considers a variety 
of factors, including the annual rate at which manufacturers can 
increase the percentage of their fleet(s) that employ a particular type 
of fuel-saving technology, the specific fleet mixes of different 
manufacturers, assumptions about the cost of the standards to 
consumers, and consumers' valuation of fuel economy, among other 
things, including, in part, safety.
---------------------------------------------------------------------------

    \58\ 67 FR 77015, 77021 (Dec. 16, 2002).
    \59\ See, e.g., Center for Auto Safety v. NHTSA (``CAS''), 793 
F.2d 1322 (D.C. Cir. 1986) (Administrator's consideration of market 
demand as component of economic practicability found to be 
reasonable); Public Citizen v. NHTSA, 848 F.2d 256 (D.C. Cir. 1988) 
(Congress established broad guidelines in the fuel economy statute; 
agency's decision to set lower standard was a reasonable 
accommodation of conflicting policies).
---------------------------------------------------------------------------

    It is important to note, however, that the law does not preclude a 
CAFE standard that poses considerable challenges to any individual 
manufacturer. The Conference Report for EPCA, as enacted in 1975, makes 
clear, and the case law affirms, ``a determination of maximum feasible 
average fuel economy should not be keyed to the single manufacturer 
which might have the most difficulty achieving a given level of average 
fuel economy.'' \60\ Instead, NHTSA is compelled ``to weigh the 
benefits to the nation of a higher fuel economy standard against the 
difficulties of individual automobile manufacturers.'' \61\ 
Accordingly, while the law permits NHTSA to set CAFE standards that 
exceed the projected capability of a particular manufacturer as long as 
the standard is economically practicable for the industry as a whole, 
the agency cannot simply disregard that impact on individual 
manufacturers.\62\ That said, in setting fuel economy standards, NHTSA 
does not seek to maintain competitive positions among the industry 
players, and notes that while a particular CAFE standard may pose 
difficulties for one manufacturer as being too high or too low, it may 
also present opportunities for another. NHTSA has long held that the 
CAFE program is not necessarily intended to maintain the competitive 
positioning of each particular company. Rather, it is intended to 
enhance the fuel economy of the vehicle fleet on American roads, while 
protecting motor vehicle safety and paying close attention to the 
economic risks.
---------------------------------------------------------------------------

    \60\ Center for Auto Safety v. NHTSA (``CAS''), 793 F.2d 1322, 
1352 (D.C. Cir. 1986).
    \61\ Id.
    \62\ Id. (``. . . the Secretary must weigh the benefits to the 
nation of a higher average fuel economy standard against the 
difficulties of individual automobile manufacturers.'')
---------------------------------------------------------------------------

(c) The Effect of Other Motor Vehicle Standards of the Government on 
Fuel Economy
    ``The effect of other motor vehicle standards of the Government on 
fuel economy'' involves an analysis of the effects of compliance with 
emission, safety, noise, or damageability standards on fuel economy 
capability and thus on average fuel economy. In many past CAFE 
rulemakings, NHTSA has said that it considers the adverse effects of 
other motor vehicle standards on fuel economy. It said so because, from 
the CAFE program's earliest years,\63\ the effects of such compliance 
on fuel economy capability over the history of the program have been 
negative ones. For example, safety standards that have the effect of 
increasing vehicle weight lower vehicle fuel economy capability and 
thus decrease the level of average fuel economy that the agency can 
determine to be feasible. NHTSA has considered the additional weight 
that it estimates would be added in response to new safety standards 
during the rulemaking timeframe. NHTSA has also accounted for EPA's 
``Tier 3'' standards for criteria pollutants in its estimates of 
technology effectiveness.\64\
---------------------------------------------------------------------------

    \63\ 42 FR 63184, 63188 (Dec. 15, 1977). See also 42 FR 33534, 
33537 (Jun. 30, 1977).
    \64\ See Section VI, below.
---------------------------------------------------------------------------

    The NPRM also discussed how EPA's CO2 standards for 
light-duty vehicles and California's Advanced Clean Cars program fit 
into NHTSA's consideration of ``the effect of other motor vehicle 
standards of the Government on fuel economy.'' The agencies note that 
on September 19, 2019, to ensure One National Program for automobile 
fuel economy and carbon dioxide emissions standards, the agencies 
finalized regulatory text related to preemption of State tailpipe 
CO2 standards and Zero Emission Vehicle (ZEV) mandates under 
EPCA and partial withdrawal of a waiver previously provided to 
California under the Clean Air Act.\65\ This final rule's impact on 
State programs--including California's--will therefore be somewhat 
different from the NPRM's consideration. In the interest of brevity, 
this preamble will hold further discussion of that point, along with 
responses to comments received, until Section VIII.
---------------------------------------------------------------------------

    \65\ 84 FR 51310 (Sept. 27, 2019).
---------------------------------------------------------------------------

(d) The Need of the United States To Conserve Energy
    ``The need of the United States to conserve energy'' means ``the 
consumer cost, national balance of payments, environmental, and foreign 
policy implications of our need for large quantities of petroleum, 
especially imported petroleum.'' \66\ Environmental implications 
principally include changes in emissions of carbon dioxide and criteria 
pollutants and air toxics. Prime examples of foreign policy 
implications are energy independence and security concerns.
---------------------------------------------------------------------------

    \66\ 42 FR 63184, 63188 (1977).
---------------------------------------------------------------------------

(1) Consumer Costs and Fuel Prices
    Fuel for vehicles costs money for vehicle owners and operators. All 
else equal (and this is an important qualification), consumers benefit 
from vehicles that need less fuel to perform the same amount of work. 
Future fuel prices are a critical input into the economic analysis of 
potential CAFE standards because they determine the value of fuel 
savings both to new vehicle buyers and to society, the amount of fuel 
economy that the new vehicle market is likely to demand in the absence 
of new standards, and they inform NHTSA about the consumer cost of the 
nation's need for large quantities of petroleum. In this final rule, 
NHTSA's analysis relies on fuel price projections estimated using the 
version of NEMS used for the U.S. Energy Information Administration's 
(EIA) Annual Energy Outlook for 2019.\67\ Federal government agencies 
generally use EIA's price projections in their assessment of future 
energy-related policies.
---------------------------------------------------------------------------

    \67\ The analysis for the proposal relied on fuel price 
projections from AEO 2017; the difference in the projections is 
discussed in Section VI.
---------------------------------------------------------------------------

(2) National Balance of Payments
    Historically, the need of the United States to conserve energy has 
included consideration of the ``national balance of payments'' because 
of concerns that importing large amounts of oil created a

[[Page 24215]]

significant wealth transfer to oil-exporting countries and left the 
U.S. economically vulnerable.\68\ As recently as 2009, nearly half of 
the U.S. trade deficit was driven by petroleum,\69\ yet this concern 
has largely lain fallow in more recent CAFE actions, in part because 
other factors besides petroleum consumption have since played a bigger 
role in the U.S. trade deficit.\70\ Given significant recent increases 
in U.S. oil production and corresponding decreases in oil imports, this 
concern seems likely to remain fallow for the foreseeable future.\71\ 
Increasingly, changes in the price of fuel have come to represent 
transfers between domestic consumers of fuel and domestic producers of 
petroleum rather than gains or losses to foreign entities.
---------------------------------------------------------------------------

    \68\ See, e.g., 42 FR 63184, 63192 (Dec. 15, 1977) (``A major 
reason for this need [to reduce petroleum consumption] is that the 
importation of large quantities of petroleum creates serious balance 
of payments and foreign policy problems. The United States currently 
spends approximately $45 billion annually for imported petroleum. 
But for this large expenditure, the current large U.S. trade deficit 
would be a surplus.'')
    \69\ See ``Today in Energy: Recent improvements in petroleum 
trade balance mitigate U.S. trade deficit,'' U.S. Energy Information 
Administration (Jul. 21, 2014), available at https://www.eia.gov/todayinenergy/detail.php?id=17191.
    \70\ See, e.g., Nida [Ccedil]akir Melek and Jun Nie, ``What 
Could Resurging U.S. Energy Production Mean for the U.S. Trade 
Deficit,'' Mar. 7, 2018, Federal Reserve Bank of Kansas City. 
Available at https://www.kansascityfed.org/publications/research/mb/articles/2018/what-could-resurging-energy-production-mean. The 
authors state that ``The decline in U.S. net energy imports has 
prevented the total U.S. trade deficit from widening further. . . . 
In 2006, petroleum accounted for about 16 percent of U.S. goods 
imports and about 3 percent of U.S. goods exports. By the end of 
2017, the share of petroleum in total goods imports declined to 8 
percent, while the share in total goods exports almost tripled, 
shrinking the U.S. petroleum trade deficit. Had the petroleum trade 
deficit not improved, all else unchanged, the total U.S. trade 
deficit would likely have been more than 35 percent wider by the end 
of 2017.''
    \71\ For an illustration of recent increases in U.S. production, 
see, e.g., `U.S. crude oil and liquid fuels production,'' Short-Term 
Energy Outlook, U.S. Energy Information Administration (Aug. 2019), 
available at http://www.eia.gov/outlooks/steo/images/Fig16.png. EIA 
noted in April 2019 that ``Annual U.S. crude oil production reached 
a record level of 10.96 million barrels per day (b/d) in 2018, 1.6 
million b/d (17%) higher than 2017 levels. In December 2018, monthly 
U.S. crude oil production reached 11.96 million b/d, the highest 
monthly level of crude oil production in U.S. history. U.S crude oil 
production has increased significantly over the past 10 years, 
driven mainly by production from tight rock formations using 
horizontal drilling and hydraulic fracturing. EIA projects that U.S. 
crude oil production will continue to grow in 2019 and 2020, 
averaging 12.3 million b/d and 13.0 million b/d, respectively.'' 
``Today in Energy: U.S. crude oil production grew 17% in 2018, 
surpassing the previous record in 1970,'' EIA, Apr. 9, 2019. 
Available at http://www.eia.gov/todayinenergy/detail.php?id=38992.
---------------------------------------------------------------------------

    As flagged in the NPRM, some commenters raised concerns about 
potential economic consequences for automaker and supplier operations 
in the U.S. due to disparities between CAFE standards at home and their 
counterpart fuel economy/efficiency and CO2 standards 
abroad. NHTSA finds these concerns more relevant to technological 
feasibility and economic practicability considerations than to the 
national balance of payments. The discussion in Section VIII below 
addresses this topic in more detail.
(3) Environmental Implications
    Higher fleet fuel economy can reduce U.S. emissions of various 
pollutants by reducing the amount of oil that is produced and refined 
for the U.S. vehicle fleet, but can also increase emissions by reducing 
the cost of driving, which can result in more vehicle miles traveled 
(i.e., the rebound effect). Thus, the net effect of more stringent CAFE 
standards on emissions of each pollutant depends on the relative 
magnitude of both its reduced emissions in fuel refining and 
distribution and increases in its emissions from vehicle use. Fuel 
savings from CAFE standards also necessarily results in lower emissions 
of CO2, the main greenhouse gas emitted as a result of 
refining, distributing, and using transportation fuels. Reducing fuel 
consumption directly reduces CO2 emissions because the 
primary source of transportation-related CO2 emissions is 
fuel combustion in internal combustion engines.
    NHTSA has considered environmental issues, both within the context 
of EPCA and the context of the National Environmental Policy Act 
(NEPA), in making decisions about the setting of standards since the 
earliest days of the CAFE program. As courts of appeal have noted in 
three decisions stretching over the last 20 years,\72\ NHTSA defined 
``the need of the United States to conserve energy'' in the late 1970s 
as including, among other things, environmental implications. In 1988, 
NHTSA included climate change concepts in its CAFE notices and prepared 
its first environmental assessment addressing that subject.\73\ It 
cited concerns about climate change as one of its reasons for limiting 
the extent of its reduction of the CAFE standard for MY 1989 passenger 
cars.\74\ Since then, NHTSA has considered the effects of reducing 
tailpipe emissions of CO2 in its fuel economy rulemakings 
pursuant to the need of the United States to conserve energy by 
reducing petroleum consumption.
---------------------------------------------------------------------------

    \72\ CAS, 793 F.2d 1322, 1325 n. 12 (D.C. Cir. 1986); Public 
Citizen, 848 F.2d 256, 262-63 n. 27 (D.C. Cir 1988) (noting that 
``NHTSA itself has interpreted the factors it must consider in 
setting CAFE standards as including environmental effects''); CBD, 
538 F.3d 1172 (9th Cir. 2007).
    \73\ 53 FR 33080, 33096 (Aug. 29, 1988).
    \74\ 53 FR 39275, 39302 (Oct. 6, 1988).
---------------------------------------------------------------------------

(4) Foreign Policy Implications
    U.S. consumption and imports of petroleum products can impose 
additional costs (i.e., externalities) on the domestic economy that are 
not reflected in the market price for crude petroleum or in the prices 
paid by consumers for petroleum products such as gasoline. NHTSA has 
said previously that these costs can include (1) higher prices for 
petroleum products resulting from the effect of U.S. oil demand on 
world oil prices, (2) the risk of disruptions to the U.S. economy 
caused by sudden increases in the global price of oil and its resulting 
impact on fuel prices faced by U.S. consumers, and (3) expenses for 
maintaining the strategic petroleum reserve (SPR) to provide a response 
option should a disruption in commercial oil supplies threaten the U.S. 
economy, to allow the U.S. to meet part of its International Energy 
Agency obligation to maintain emergency oil stocks, and to provide a 
national defense fuel reserve.\75\ Higher U.S. consumption of crude oil 
or refined petroleum products increases the magnitude of these external 
economic costs, thus increasing the true economic cost of supplying 
transportation fuels above the resource costs of producing them. 
Conversely, reducing U.S. consumption of crude oil or refined petroleum 
products (by reducing motor fuel use) can reduce these external costs.
---------------------------------------------------------------------------

    \75\ While the U.S. maintains a military presence in certain 
parts of the world to help secure global access to petroleum 
supplies, that is neither the primary nor the sole mission of U.S. 
forces overseas. Additionally, the scale of oil consumption 
reductions associated with CAFE standards would be insufficient to 
alter any existing military missions focused on ensuring the safe 
and expedient production and transportation of oil around the globe. 
See the FRIA's discussion on energy security for more information on 
this topic.
---------------------------------------------------------------------------

    While these costs are considerations, the United States has 
significantly increased oil production capabilities in recent years, to 
the extent that the U.S. is currently producing enough oil to satisfy 
nearly all of its energy needs and is projected to continue to do so 
(or even become a net energy exporter in the near future).\76\ This has 
added stable new supply to the global oil market, which ameliorates the 
U.S.' need to

[[Page 24216]]

conserve energy from a security perspective even given that oil is a 
global commodity. The agencies discuss this issue in more detail in 
Section VIII below.
---------------------------------------------------------------------------

    \76\ See AEO 2019, at 14 (``In the Reference case, the United 
States becomes a net exporter of petroleum liquids after 2020 as 
U.S. crude oil production increases and domestic consumption of 
petroleum products decreases.''). Available at https://www.eia.gov/outlooks/aeo/pdf/aeo2019.pdf.
---------------------------------------------------------------------------

(2) Factors That NHTSA Is Prohibited From Considering
    EPCA states that in determining the level at which it should set 
CAFE standards for a particular model year, NHTSA may not consider the 
ability of manufacturers to take advantage of several EPCA provisions 
that facilitate compliance with CAFE standards and thereby can reduce 
their costs of compliance.\77\ As discussed further below, NHTSA cannot 
consider compliance credits that manufacturers earn by exceeding the 
CAFE standards and then use to achieve compliance in years in which 
their measured average fuel economy falls below the standards. NHTSA 
also cannot consider the use of alternative fuels by dual-fueled 
vehicles (such as plug-in hybrid electric vehicles) nor the 
availability of dedicated alternative fuel vehicles (such as battery 
electric or hydrogen fuel cell vehicles) in any model year. EPCA 
encourages the production of alternative fuel vehicles by specifying 
that their fuel economy is to be determined using a special calculation 
procedure that results in those vehicles being assigned a higher fuel 
economy level than they actually achieve. For non-statutory incentives 
that NHTSA developed by regulation, NHTSA does not consider these 
incentives subject to the EPCA prohibition on considering 
flexibilities. These topics will be addressed further in Section VIII 
below.
---------------------------------------------------------------------------

    \77\ 49 U.S.C. 32902(h).
---------------------------------------------------------------------------

(3) Other Considerations in Determining Maximum Feasible CAFE Standards
    NHTSA historically has interpreted EPCA's statutory factors as 
including consideration for potential adverse safety consequences in 
setting CAFE standards. Courts have consistently recognized that this 
interpretation is reasonable. As courts have recognized, ``NHTSA has 
always examined the safety consequences of the CAFE standards in its 
overall consideration of relevant factors since its earliest rulemaking 
under the CAFE program.'' \78\ The courts have consistently upheld 
NHTSA's implementation of EPCA in this manner.\79\ Thus, in evaluating 
what levels of stringency would result in maximum feasible standards, 
NHTSA assesses the potential safety impacts and considers them in 
balancing the statutory considerations and to determine the maximum 
feasible level of the standards.\80\ Many commenters addressed the 
NPRM's analysis of safety impacts; those comments will be summarized 
and responded to in Section VI.D.2 and also in each agency's discussion 
in Section VIII.
---------------------------------------------------------------------------

    \78\ Competitive Enterprise Institute v. NHTSA, 901 F.2d 107, 
120 n. 11 (D.C. Cir. 1990) (``CEI-I'') (citing 42 FR 33534, 33551 
(Jun. 30, 1977).
    \79\ See, e.g., Competitive Enterprise Institute v. NHTSA, 956 
F.2d 321, 322 (D.C. Cir. 1992) (``CEI-II'') (in determining the 
maximum feasible fuel economy standard, ``NHTSA has always taken 
passenger safety into account,'' citing CEI-I, 901 F.2d at 120 n. 
11); Competitive Enterprise Institute v. NHTSA, 49 F.3d 481, 483-83 
(D.C. Cir. 1995) (same); Center for Biological Diversity v. NHTSA, 
538 F.3d 1172, 1203-04 (9th Cir. 2008) (upholding NHTSA's analysis 
of vehicle safety issues with weight in connection with the MYs 
2008-2011 light truck CAFE rulemaking).
    \80\ NHTSA stated in the NPRM that ``While we discuss safety as 
a separate consideration, NHTSA also considers safety as closely 
related to, and in some circumstances a subcomponent of, economic 
practicability. On a broad level, manufacturers have finite 
resources to invest in research and development. Investment into the 
development and implementation of fuel saving technology necessarily 
comes at the expense of investing in other areas such as safety 
technology. On a more direct level, when making decisions on how to 
equip vehicles, manufacturers must balance cost considerations to 
avoid pricing further consumers out of the market. As manufacturers 
add technology to increase fuel efficiency, they may decide against 
installing new safety equipment to reduce cost increases. And as the 
price of vehicles increase beyond the reach of more consumers, such 
consumers continue to drive or purchase older, less safe vehicles. 
In assessing practicability, NHTSA also considers the harm to the 
nation's economy caused by highway fatalities and injuries.'' 83 FR 
at 43209 (Aug. 24, 2018). Many comments were received on this issue, 
which will be discussed further in Section VIII below.
---------------------------------------------------------------------------

    The above sections explain what Congress thought was important 
enough to codify when it directed each agency to regulate, and begin to 
explain how the agencies have interpreted those directions over time 
and in this final rule. The next section looks more closely at the 
interplay between Congress's direction to the agencies and the aspects 
of the market that these regulations affect, as follows.

IV. Purpose of Analytical Approach Considered as Part of Decision-
Making

A. Relationship of Analytical Approach to Governing Law

    Like the NPRM, today's final rule is supported by extensive 
analysis of potential impacts of the regulatory alternatives under 
consideration. Below, Section VI reviews the analytical approach, 
Section VII summarizes the results of the analysis, and Section VIII 
explains how the final standards--informed by this analysis--fulfill 
the agencies' statutory obligations. Accompanying today's notice, a 
final Regulatory Impact Analysis (FRIA) and, for NHTSA's consideration, 
a final Environmental Impact Analysis (FEIS), together provide a more 
extensive and detailed enumeration of related methods, estimates, 
assumptions, and results. The agencies' analysis has been constructed 
specifically to reflect various aspects of governing law applicable to 
CAFE and CO2 standards, and has been expanded and improved 
in response to comments received to the NPRM and based on additional 
work by the agencies. The analysis aided the agencies in implementing 
their statutory obligations, including the weighing of competing 
considerations, by reasonably informing the agencies about the 
estimated effects of choosing different regulatory alternatives.
    The agencies' analysis makes use of a range of data (i.e., 
observations of things that have occurred), estimates (i.e., things 
that may occur in the future), and models (i.e., methods for making 
estimates). Two examples of data include (1) records of actual odometer 
readings used to estimate annual mileage accumulation at different 
vehicle ages and (2) CAFE compliance data used as the foundation for 
the ``analysis fleet'' containing, among other things, production 
volumes and fuel economy levels of specific configurations of specific 
vehicle models produced for sale in the U.S. Two examples of estimates 
include (1) forecasts of future GDP growth used, with other estimates, 
to forecast future vehicle sales volumes and (2) the ``retail price 
equivalent'' (RPE) factor used to estimate the ultimate cost to 
consumers of a given fuel-saving technology, given accompanying 
estimates of the technology's ``direct cost,'' as adjusted to account 
for estimated ``cost learning effects'' (i.e., the tendency that it 
will cost a manufacturer less to apply a technology as the manufacturer 
gains more experience doing so).
    The agencies' analysis makes use of several models, some of which 
are actually integrated systems of multiple models. As discussed in the 
NPRM, the agencies' analysis of CAFE and CO2 standards 
involves two basic elements: First, estimating ways each manufacturer 
could potentially respond to a given set of standards in a manner that 
considers potential consumer response; and second, estimating various 
impacts of those responses. Estimating manufacturers' potential 
responses involves simulating manufacturers' decision-making processes 
regarding the year-by-year application of fuel-saving technologies to 
specific vehicles. Estimating impacts involves calculating resultant 
changes in new vehicle costs, estimating a

[[Page 24217]]

variety of costs (e.g., for fuel) and effects (e.g., CO2 
emissions from fuel combustion) occurring as vehicles are driven over 
their lifetimes before eventually being scrapped, and estimating the 
monetary value of these effects. Estimating impacts also involves 
consideration of the response of consumers--e.g., whether consumers 
will purchase the vehicles and in what quantities. Both of these basic 
analytical elements involve the application of many analytical inputs.
    The agencies' analysis uses the CAFE Model to estimate 
manufacturers' potential responses to new CAFE and CO2 
standards and to estimate various impacts of those responses. The model 
may be characterized as an integrated system of models. For example, 
one model estimates manufacturers' responses, another estimates 
resultant changes in total vehicle sales, and still another estimates 
resultant changes in fleet turnover (i.e., scrappage). The CAFE model 
makes use of many inputs, values of which are developed outside of the 
model and not by the model. For example, the model applies fuel prices; 
it does not estimate fuel prices. The model does not determine the form 
or stringency of the standards; instead, the model applies inputs 
specifying the form and stringency of standards to be analyzed and 
produces outputs showing effects of manufacturers working to meet those 
standards, which become the basis for comparing between different 
potential stringencies.
    The agencies also use EPA's MOVES model to estimate ``tailpipe'' 
(a.k.a. ``vehicle'' or ``downstream'') emission factors for criteria 
pollutants,\81\ and use four DOE and DOE-sponsored models to develop 
inputs to the CAFE model, including three developed and maintained by 
DOE's Argonne National Laboratory. The agencies use the DOE Energy 
Information Administration's (EIA's) National Energy Modeling System 
(NEMS) to estimate fuel prices,\82\ and use Argonne's Greenhouse gases, 
Regulated Emissions, and Energy use in Transportation (GREET) model to 
estimate emissions rates from fuel production and distribution 
processes.\83\ DOT also sponsored DOE/Argonne to use Argonne's 
Autonomie full-vehicle modeling and simulation system to estimate the 
fuel economy impacts for roughly a million combinations of technologies 
and vehicle types.84 85 Section VI.B.3, below, and the 
accompanying final RIA document details of the agencies' use of these 
models. In addition, as discussed in the final EIS accompanying today's 
notice, DOT relied on a range of climate and photochemical models to 
estimate impacts on climate, air quality, and public health. The EIS 
discusses and documents the use of these models.
---------------------------------------------------------------------------

    \81\ See https://www.epa.gov/moves. Today's final rule used 
version MOVES2014b, available at https://www.epa.gov/moves/latest-version-motor-vehicle-emission-simulator-moves.
    \82\ See https://www.eia.gov/outlooks/aeo/info_nems_archive.php. 
Today's final rule uses fuel prices estimated using the Annual 
Energy Outlook (AEO) 2019 version of NEMS (see https://www.eia.gov/outlooks/aeo/data/browser/#/?id=3-AEO2019&cases=ref2019&sourcekey=0).
    \83\ Information regarding GREET is available at https://greet.es.anl.gov/index.php. Today's notice uses the 2018 version of 
GREET.
    \84\ As part of the Argonne simulation effort, individual 
technology combinations simulated in Autonomie were paired with 
Argonne's BatPAC model to estimate the battery cost associated with 
each technology combination based on characteristics of the 
simulated vehicle and its level of electrification. Information 
regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
    \85\ In addition, the impact of engine technologies on fuel 
consumption, torque, and other metrics was characterized using GT 
POWER simulation modeling in combination with other engine modeling 
that was conducted by IAV Automotive Engineering, Inc. (IAV). The 
engine characterization ``maps'' resulting from this analysis were 
used as inputs for the Autonomie full-vehicle simulation modeling. 
Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
---------------------------------------------------------------------------

    As further explained in the NPRM,\86\ to prepare for analysis 
supporting the proposal, DOT expanded the CAFE model to address EPA 
statutory and regulatory requirements through a year-by-year simulation 
of how manufacturers could comply with EPA's CO2 standards, 
including:
---------------------------------------------------------------------------

    \86\ 83 FR 42986, 43003 (Aug. 24, 2018).
---------------------------------------------------------------------------

     Calculation of vehicle models' CO2 emission 
rates before and after application of fuel-saving (and, therefore, 
CO2-reducing) technologies;
     Calculation of manufacturers' fleet average CO2 
emission rates;
     Calculation of manufacturers' fleet average CO2 
emission rates under attribute-based CO2 standards;
     Accounting for adjustments to average CO2 
emission rates reflecting reduction of air conditioner refrigerant 
leakage;
     Accounting for the treatment of alternative fuel vehicles 
for CO2 compliance;
     Accounting for production ``multipliers'' for PHEVs, BEVs, 
compressed natural gas (CNG) vehicles, and fuel cell vehicles (FCVs);
     Accounting for transfer of CO2 credits between 
regulated fleets; and
     Accounting for carried-forward (a.k.a. ``banked'') 
CO2 credits, including credits from model years earlier than 
modeled explicitly.
    As further discussed in the NPRM, although EPA had previously 
developed a vehicle simulation tool (``ALPHA'') and a fleet compliance 
model (``OMEGA''), and had applied these in prior actions, having 
considered the facts before the Agency in 2018, EPA determined that, 
``it is reasonable and appropriate to use DOE/Argonne's model for full-
vehicle simulation, and to use DOT's CAFE model for analysis of 
regulatory alternatives.'' \87\
---------------------------------------------------------------------------

    \87\ 83 FR 42986, 43000 (Aug. 24, 2018).
---------------------------------------------------------------------------

    As discussed below and in Section VI.B.3, some commenters--some 
citing deliberative EPA staff communications during NPRM development, 
and one submitting comments by a former EPA staff member closely 
involved in the origination of the above-mentioned OMEGA model--took 
strong exception to EPA's decision to rely on DOE/Argonne and DOT-
originated models as the basis for analysis informing EPA's decisions 
regarding CO2 standards. Some commenters argued that the EPA 
Administrator must consider exclusively models and analysis originating 
with EPA staff, and that to do otherwise would be arbitrary and 
capricious. As explained below (and as explained in the NPRM), it is 
reasonable for the Administrator to consider analysis and information 
produced from many sources, including, in this instance, the DOE/
Argonne and DOT models. The Administrator has the discretion to 
determine what information reasonably and appropriately informs 
decisions regarding emissions standards. Some commenters conflated 
models with decisions, suggesting that the former mechanically 
determine the latter. The CAA authorizes the EPA Administrator, not a 
model, to make decisions about emissions standards, just as EPCA 
provides similar authority to the Secretary. Models produce analysis, 
the results of which help to inform decisions. However, in making such 
decisions, the Administrator may and should consider other relevant 
information beyond the outputs of any models--including public 
comment--and, in all cases, must exercise judgment in establishing 
appropriate standards.
    Some commenters conflated models with inputs and/or with results of 
the modeling. All of the models mentioned above rely on inputs, 
including not only data (i.e., facts), but also estimates (inputs about 
the future are estimates, not data). Given these inputs, the models 
produce estimates--ultimately, the agencies' reported estimates of the 
potential impacts of standards under

[[Page 24218]]

consideration. In other words, inputs do not define models; models use 
inputs. Therefore, disagreements about inputs do not logically extend 
to disagreements about models. Similarly, while models determine 
resulting outputs, they do so based on inputs. Therefore, disagreements 
about results do not necessarily imply disagreements about models; they 
may merely reflect disagreements about inputs. With respect to the 
Administrator's decisions regarding models underlying today's analysis, 
comments regarding inputs, therefore, are more appropriately addressed 
separately, which is done so below in Section VI.
    The EPA Administrator's decision to continue relying on the DOE/
Argonne Autonomie tool and DOT CAFE model rather than on the 
corresponding tools developed by EPA staff is informed by consideration 
of comments on results and on technical aspects of the models 
themselves. As discussed below, some commenters questioned specific 
aspects of the CAFE model's simulation of manufacturer's potential 
responses to CO2 standards. Considering these comments, the 
CAFE model applied in the final rule's analysis includes some revisions 
and updates. For example, the ``effective cost'' metric used to select 
among available opportunities to apply fuel-saving technologies now 
uses a ``cost per credit'' metric rather than the metric used for the 
NPRM. Also, the model's representation of sales ``multipliers'' EPA has 
included for CNG vehicles, PHEVs, BEVs, and FCVs reflects current EPA 
regulations or, as an input-selectable option, an alternative approach 
under consideration. On the other hand, some commenters questioning the 
CAFE model's approach to some CO2 program features appear to 
ignore the fact that prior analysis by EPA (using EPA's OMEGA) model 
likewise did not account for the same program features. For example, 
some stakeholders took issue with the CAFE model's approach to 
accounting for banked CO2 credits and, in particular, 
credits banked prior to the model years accounted for explicitly in the 
analysis. In the course of updating the basis for analysis fleet from 
model year 2016 to model year 2017, the agencies have since updated 
corresponding inputs. However, even though the ability to carry forward 
credits impacts outcomes, EPA's OMEGA model used in previous 
rulemakings never attempted to account for credit banking and, indeed, 
lacking a year-by-year structure, cannot account for credit banking. 
Therefore, at least with respect to this important CO2 
program flexibility, the CAFE model provides a more complete and 
realistic basis for estimating actual impacts of new CO2 
standards.
    For its part, NHTSA remains confident that the combination of the 
Autonomie and CAFE models remains the best available for CAFE 
rulemaking analysis, and notes, as discussed below, that even the 
environmental group coalition stated that the CAFE model is aligned 
with EPCA requirements.\88\ In late 2001, after Congress discontinued 
an extended series of budget ``riders'' prohibiting work on CAFE 
standards, NHTSA and the DOT Volpe Center began development of a 
modeling system appropriate for CAFE rulemaking analysis, because other 
available models were not designed with this purpose in mind, and 
lacked capabilities important for CAFE rulemakings. For example, 
although NEMS had procedures to account for CAFE standards, those 
procedures did not provide the ability to account for specific 
manufacturers, as is especially relevant to the statutory requirement 
that NHTSA consider the economic practicability of any new CAFE 
standards. Also, as early as the first rulemaking making use of this 
early CAFE model, commenters stressed the importance of product 
redesign schedules, leading developers to introduce procedures to 
account for product cadence. In the 2003 notice regarding light truck 
standards for MYs 2005-2007, NHTSA stated that ``we also changed the 
methodology to recognize that capital costs require employment of 
technologies for several years, rather than a single year. . . . In our 
view, this makes the Volpe analysis more consistent with the [manually 
implemented] Stage analysis and better reflects actual conditions in 
the automotive industry.'' \89\ Since that time, NHTSA and the Volpe 
Center have significantly refined the CAFE model with each of 
rulemaking. For example, for the 2006 rulemaking regarding standards 
for MYs 2008-2011 light trucks, NHTSA introduced the ability to account 
for attribute-based standards, account for the social cost of 
CO2 emissions, estimate stringencies at which net benefits 
would be maximized, and perform probabilistic uncertainty analysis 
(i.e., Monte Carlo simulation).\90\ For the 2009 rulemaking regarding 
standards for MY 2011 passenger cars and light trucks, we introduced 
the ability to account for attribute-based passenger car standards, and 
the ability to apply ``synergy factors'' to estimate how some 
technology pairings impact fuel consumption,\91\ For the 2010 
rulemaking regarding standards for MYs 2012-2016, we introduced 
procedures to account for FFV credits, and to account for product 
planning as a multiyear consideration.\92\ For the 2012 rulemaking 
regarding standards for MYs 2017-2025, we introduced several new 
procedures, such as (1) accounting for electricity used to charge 
electric vehicles (EVs) and plug-in hybrid electric vehicles (PHEVs), 
(2) accounting for use of ethanol blends in flexible-fuel vehicles 
(FFVs), (3) accounting for costs (i.e., ``stranded capital'') related 
to early replacement of technologies, (4) accounting for previously-
applied technology when determining the extent to which a manufacturer 
could expand use of the technology, (5) applying technology-specific 
estimates of changes in consumer value, (6) simulating the extent to 
which manufacturers might utilize EPCA's provisions regarding 
generation and use of CAFE credits, (7) applying estimates of fuel 
economy adjustments (and accompanying costs) reflecting increases in 
air conditioner efficiency, (8) reporting privately-valued benefits, 
(9) simulating the extent to which manufacturers might voluntarily 
apply technology beyond levels needed for compliance with CAFE 
standards, and (10) estimating changes in highway fatalities 
attributable to any applied reductions in vehicle mass.\93\ Also for 
the 2012 rulemaking, we began making use of Autonomie to estimate fuel 
consumption impacts of different combinations of technologies, using 
these estimates to specify inputs to the CAFE model.\94\ In 2016, 
providing analyses for both the draft TAR regarding light-duty CAFE 
standards and the final rule regarding fuel consumption standards for 
heavy-duty pickup trucks and vans, we greatly expanded the agency's use 
of Autonomie-based full vehicle simulations and introduced the ability 
to simulate compliance with attribute-based standards for heavy-duty 
pickups and vans.\95\ And, as discussed at length in the NPRM and 
below, for this rulemaking, we have, among other things, refined 
procedures to account for impacts on highway travel and safety,

[[Page 24219]]

added procedures to simulate compliance with CO2 standards, 
refined procedures to account for compliance credits, and added 
procedures to account for impacts on sales, scrappage, and employment. 
We have also significantly revised the model's graphical user interface 
(GUI) in order to make the model easier to operate and understand. Like 
any model, both Autonomie and the CAFE model benefit from ongoing 
refinement. However, NHTSA is confident that this combination of models 
produces a more realistic characterization of the potential impacts of 
new standards than would another combination of available models. Some 
stakeholders, while commenting on specific aspects of the inputs, 
models, and/or results, commended the agencies' exclusive reliance on 
the DOE/Argonne Autonomie tool and DOT CAFE model. With respect to 
CO2 standards, these stakeholders noted not only technical 
reasons to use these models rather than the EPA models, but also other 
reasons such as efficiency, transparency, and ease with which outside 
parties can exercise models and replicate the agencies' analysis. These 
comments are discussed below and in Section VI.
---------------------------------------------------------------------------

    \88\ Environmental group coalition, NHTSA-2018-0067-12000, 
Appendix A, at 24-25.
    \89\ 68 FR at 16885 (Apr. 7, 2003).
    \90\ 71 FR at 17566 et seq. (Apr. 6, 2006).
    \91\ 74 FR at 14196 et seq. (Mar. 30, 3009).
    \92\ 75 FR at 25599 et seq. (May 7, 2010).
    \93\ 77 FR 63009 et seq. (Oct. 15, 2012).
    \94\ 77 FR at 62712 et seq. (Oct. 15, 2012).
    \95\ 81 FR at 73743 et seq. (Oct. 25, 2016); Draft TAR, 
available at Docket No. NHTSA-2016-0068-0001, Chapter 13.
---------------------------------------------------------------------------

    Nevertheless, some comments regarding the model's handling of CAFE 
and/or CO2 standards, and some comments regarding the 
model's estimation of resultant impacts, led the agencies to make 
changes to specific aspects of the model. Comments on and changes to 
the inputs and model are discussed below and in Section VI; results are 
discussed in Section VII and in the accompanying RIA; and the meaning 
of results in the context of the applicable statutory requirements is 
discussed in Section VIII.
    As explained, the analysis is designed to reflect a number of 
statutory and regulatory requirements applicable to CAFE and tailpipe 
CO2 standard setting. EPCA contains a number of requirements 
governing the scope and nature of CAFE standard setting. Among these, 
some have been in place since EPCA was first signed into law in 1975, 
and some were added in 2007, when Congress passed EISA and amended 
EPCA. The CAA, as discussed elsewhere, provides EPA with very broad 
authority under Section 202(a), and does not contain EPCA/EISA's 
prescriptions. In the interest of harmonization, however, EPA has 
adopted some of the EPCA/EISA requirements into its tailpipe 
CO2 regulations, and NHTSA, in turn, has created some 
additional flexibilities by regulation not expressly envisioned by 
EPCA/EISA in order to harmonize better with some of EPA's programmatic 
decisions. EPCA/EISA requirements regarding the technical 
characteristics of CAFE standards and the analysis thereof include, but 
are not limited to, the following, and the analysis reflects these 
requirements as summarized:
    Corporate Average Standards: 49 U.S.C. 32902 requires standards 
that apply to the average fuel economy levels achieved by each 
corporation's fleets of vehicles produced for sale in the U.S.\96\ CAA 
Section 202(a) does not preclude the EPA Administrator from expressing 
CO2 standards as de facto fleet average requirements, and 
EPA has adopted a similar approach in the interest of harmonization. 
The CAFE Model, used by the agencies to conduct the bulk of today's 
analysis, calculates the CAFE and CO2 levels of each 
manufacturer's fleets based on estimated production volumes and 
characteristics, including fuel economy levels, of distinct vehicle 
models that could be produced for sale in the U.S.
---------------------------------------------------------------------------

    \96\ This differs from safety standards and traditional 
emissions standards, which apply separately to each vehicle. For 
example, every vehicle produced for sale in the U.S. must, on its 
own, meet all applicable federal motor vehicle safety standards 
(FMVSS), but no vehicle produced for sale must, on its own, federal 
fuel economy standards. Rather, each manufacturer is required to 
produce a mix of vehicles that, taken together, achieve an average 
fuel economy level no less than the applicable minimum level.
---------------------------------------------------------------------------

    Separate Standards for Passenger Cars and Light Trucks: 49 U.S.C. 
32902 requires the Secretary of Transportation to set CAFE standards 
separately for passenger cars and light trucks. CAA Section 202(a) does 
not preclude the EPA Administrator from specifying CO2 
standards separately for passenger cars and light trucks, and EPA has 
adopted a similar approach. The CAFE Model accounts separately for 
passenger cars and light trucks, including differentiated standards and 
compliance.
    Attribute-Based Standards: 49 U.S.C. 32902 requires the Secretary 
of Transportation to define CAFE standards as mathematical functions 
expressed in terms of one or more vehicle attributes related to fuel 
economy. This means that for a given manufacturer's fleet of vehicles 
produced for sale in the U.S. in a given regulatory class and model 
year, the applicable minimum CAFE requirement (i.e., the numerical 
value of the requirement) is computed based on the applicable 
mathematical function, and the mix and attributes of vehicles in the 
manufacturer's fleet. In the 2012 final rule that first established 
CO2 standards, EPA also adopted an attribute-based standard 
under its broad CAA Section 202(a) authority. The CAFE Model accounts 
for such functions and vehicle attributes explicitly.
    Separately Defined Standards for Each Model Year: 49 U.S.C. 32902 
requires the Secretary to set CAFE standards (separately for passenger 
cars and light trucks) at the maximum feasible levels in each model 
year. CAA Section 202(a) allows EPA to establish CO2 
standards separately for each model year, and EPA has chosen to do so 
for this final rule, similar to the approach taken in the previous 
light-duty vehicle CO2 standard-setting rules. The CAFE 
Model represents each model year explicitly, and accounts for the 
production relationships between model years.\97\
---------------------------------------------------------------------------

    \97\ For example, a new engine first applied to given vehicle 
model/configuration in model year 2020 will most likely be ``carried 
forward'' to model year 2021 of that same vehicle model/
configuration, in order to reflect the fact that manufacturers do 
not apply brand-new engines to a given vehicle model every single 
year.
---------------------------------------------------------------------------

    Separate Compliance for Domestic and Imported Passenger Car Fleets: 
49 U.S.C. 32904 requires the EPA Administrator to determine CAFE 
compliance separately for each manufacturers' fleets of domestic 
passenger cars and imported passenger cars, which manufacturers must 
consider as they decide how to improve the fuel economy of their 
passenger car fleets. CAA 202(a) does not preclude the EPA 
Administrator from determining compliance with CO2 standards 
separately for a manufacturer's domestic and imported car fleets, but 
EPA did not include such a distinction in either the 2010 or 2012 final 
rules, and EPA did not propose or ask for comment on taking such an 
approach in the proposal. The CAFE Model is able to account explicitly 
for this requirement when simulating manufacturers' potential responses 
to CAFE standards, but combines any given manufacturer's domestic and 
imported cars into a single fleet when simulating that manufacturer's 
potential response to CO2 standards.
    Minimum CAFE Standards for Domestic Passenger Car Fleets: 49 U.S.C. 
32902 requires that domestic passenger car fleets achieve CAFE levels 
no less than 92 percent of the industry-wide average level required 
under the applicable attribute-based CAFE standard, as projected by the 
Secretary at the time the standard is promulgated. CAA 202(a) does not 
preclude the EPA Administrator from correspondingly requiring that 
domestic passenger car fleets achieve CO2 levels no greater 
than 108.7 percent (1/0.92 = 1.087) of the projected industry-wide 
average CO2

[[Page 24220]]

requirement under the attribute-based standard, but the GHG program 
that EPA designed in the 2010 and 2012 final rules did not include such 
a distinction, and EPA did not propose or seek comment on such an 
approach in the proposal. The CAFE Model is able to account explicitly 
for this requirement for CAFE standards, and sets this requirement 
aside for CO2 standards.
    Civil Penalties for Noncompliance: 49 U.S.C. 32912 prescribes a 
rate (in dollars per tenth of a mpg) at which the Secretary is to levy 
civil penalties if a manufacturer fails to comply with a CAFE standard 
for a given fleet in a given model year, after considering available 
credits. Some manufacturers have historically demonstrated a 
willingness to treat CAFE noncompliance as an ``economic'' choice, 
electing to pay civil penalties rather than achieving full numerical 
compliance across all fleets. The CAFE Model calculates civil penalties 
for CAFE shortfalls and provides means to estimate that a manufacturer 
might stop adding fuel-saving technologies once continuing to do so 
would be effectively more ``expensive'' (after accounting for fuel 
prices and buyers' willingness to pay for fuel economy) than paying 
civil penalties. In contrast, the CAA does not authorize the EPA 
Administrator to allow manufacturers to sell noncompliant fleets and 
instead only pay civil penalties; manufacturers who choose to pay civil 
penalties for CAFE compliance tend to employ EPA's more-extensive 
programmatic flexibilities to meet tailpipe CO2 emissions 
standards. Thus, the CAFE Model does not allow civil penalty payment as 
an option for CO2 standards.
    Dual-Fueled and Dedicated Alternative Fuel Vehicles: For purposes 
of calculating CAFE levels used to determine compliance, 49 U.S.C. 
32905 and 32906 specify methods for calculating the fuel economy levels 
of vehicles operating on alternative fuels to gasoline or diesel 
through MY 2020. After MY 2020, methods for calculating alternative 
fuel vehicle (AFV) fuel economy are governed by regulation. The CAFE 
Model is able to account for these requirements explicitly for each 
vehicle model. However, 49 U.S.C. 32902 requires that maximum feasible 
CAFE standards be set in a manner that does not presume manufacturers 
can respond by producing new dedicated alternative fuel vehicle (AFV) 
models. The CAFE model can be run in a manner that excludes the 
additional application of dedicated AFV technologies in model years for 
which maximum feasible standards are under consideration. As allowed 
under NEPA for analysis appearing in EISs informing decisions regarding 
CAFE standards, the CAFE Model can also be run without this analytical 
constraint. CAA 202(a) does not preclude the EPA Administrator adopting 
analogous provisions, but EPA has instead opted through regulation to 
``count'' dual- and alternative fuel vehicles on a CO2 basis 
(and through MY 2026, to set aside emissions from electricity 
generation). The CAFE model accounts for this treatment of dual- and 
alternative fuel vehicles when simulating manufacturers' potential 
responses to CO2 standards. For natural gas vehicles, both 
dedicated and dual-fueled, EPA is establishing a multiplier of 2.0 for 
model years 2022-2026.
    Creation and Use of Compliance Credits: 49 U.S.C. 32903 provides 
that manufacturers may earn CAFE ``credits'' by achieving a CAFE level 
beyond that required of a given fleet in a given model year, and 
specifies how these credits may be used to offset the amount by which a 
different fleet falls short of its corresponding requirement. These 
provisions allow credits to be ``carried forward'' and ``carried back'' 
between model years, transferred between regulated classes (domestic 
passenger cars, imported passenger cars, and light trucks), and traded 
between manufacturers. However, these provisions also impose some 
specific statutory limits. For example, CAFE compliance credits can be 
carried forward a maximum of five model years and carried back a 
maximum of three model years. Also, EPCA/EISA caps the amount of credit 
that can be transferred between passenger car and light truck fleets, 
and prohibits manufacturers from applying traded or transferred credits 
to offset a failure to achieve the applicable minimum standard for 
domestic passenger cars. The CAFE Model explicitly simulates 
manufacturers' potential use of credits carried forward from prior 
model years or transferred from other fleets.\98\ 49 U.S.C. 32902 
prohibits consideration of manufacturers' potential application of CAFE 
compliance credits when setting maximum feasible CAFE standards. The 
CAFE Model can be operated in a manner that excludes the application of 
CAFE credits after a given model year. CAA 202(a) does not preclude the 
EPA Administrator adopting analogous provisions. EPA has opted to limit 
the ``life'' of compliance credits from most model years to 5 years, 
and to limit borrowing to 3 years, but has not adopted any limits on 
transfers (between fleets) or trades (between manufacturers) of 
compliance credits. The CAFE Model is able to account for the absence 
of limits on transfers of CO2 standards. Insofar as the CAFE 
model can be exercised in a manner that simulates trading of 
CO2 compliance credits, such simulations treat trading as 
unlimited.\99\ EPA has considered manufacturers' ability to use credits 
as part of its decisions on these final standards, and the CAFE model 
is now able to account for that.
---------------------------------------------------------------------------

    \98\ As explained in Section VI, the CAFE Model does not 
explicitly simulate the potential that manufacturers would carry 
CAFE or CO2 credits back (i.e., borrow) from future model 
years, or acquire and use CAFE compliance credits from other 
manufacturers. At the same time, because EPA has elected to not 
limit credit trading, the CAFE Model can be exercised in a manner 
that simulates unlimited (a.k.a. ``perfect'') CO2 
compliance credit trading throughout the industry (or, potentially, 
within discrete trading ``blocs''). The agencies believe there is 
significant uncertainty in how manufacturers may choose to employ 
these particular flexibilities in the future: for example, while it 
is reasonably foreseeable that a manufacturer who over-complies in 
one year may ``coast'' through several subsequent years relying on 
those credits rather than continuing to make technology 
improvements, it is harder to assume with confidence that 
manufacturers will rely on future technology investments (that may 
not pan out as expected, as if market demand for ``target-beater'' 
vehicles is lower than expected) to offset prior-year shortfalls, or 
whether/how manufacturers will trade credits with market competitors 
rather than making their own technology investments. Historically, 
carry-back and trading have been much less utilized than carry-
forward, for a variety of reasons including higher risk and 
preference not to ``pay competitors to make fuel economy 
improvements we should be making'' (to paraphrase one manufacturer), 
although the agencies recognize that carry-back and trading are used 
more frequently when standards require more technology application 
than manufacturers believe their markets will bear. Given the 
uncertainty just discussed, and given also the fact that the 
agencies have yet to resolve some of analytical challenges 
associated with simulating use of these flexibilities, the agencies 
consider borrowing and trading to involve sufficient risk that it is 
prudent to support today's decisions with analysis that sets aside 
the potential that manufacturers could come to depend widely on 
borrowing and trading. While compliance costs in real life may be 
somewhat different from what is modeled today as a result of this 
analytical decision, that is broadly true no matter what, and the 
agencies do not believe that the difference would be so great that 
it would change the policy outcome.
    \99\ To avoid making judgments (that would invariably turn out 
to be at least somewhat incorrect) about possible future trading 
activity, the model simulates trading by combining all manufacturers 
into a single entity, so that the most cost-effective choices are 
made for the fleet as a whole.
---------------------------------------------------------------------------

    Statutory Basis for Stringency: 49 U.S.C. 32902 requires the 
Secretary to set CAFE standards at the maximum feasible levels, 
considering technological feasibility, economic practicability, the 
need of the Nation to conserve energy, and the impact of other 
government standards. EPCA/EISA authorizes the Secretary to interpret

[[Page 24221]]

these factors, and as the Department's interpretation has evolved, 
NHTSA has continued to expand and refine its qualitative and 
quantitative analysis. For example, as discussed below in Section 
VI.B.3, the Autonomie simulations reflect the agencies' judgment that 
it would not be economically practicable for a manufacturer to 
``split'' an engine shared among many vehicle model/configurations into 
a myriad of versions each optimized to a single vehicle model/
configuration. Also responding to evolving interpretation of these 
EPCA/EISA factors, the CAFE Model has been expanded to address 
additional impacts in an integrated manner. For example, the CAFE Model 
version used for the NPRM analysis included the ability to estimate 
impacts on labor utilization internally, rather than as an external 
``off model'' or ``post processing'' analysis. In addition, NEPA 
requires the Secretary to issue an EIS that documents the estimated 
impacts of regulatory alternatives under consideration. The EIS 
accompanying today's notice documents changes in emission inventories 
as estimated using the CAFE model, but also documents corresponding 
estimates--based on the application of other models documented in the 
EIS, of impacts on the global climate, on tropospheric air quality, and 
on human health. Regarding CO2 standards, CAA 202(a) 
provides general authority for the establishment of motor vehicle 
emissions standards, and the final rule's analysis, like that 
accompanying the agencies' proposal, addresses impacts relevant to the 
EPA Administrator's decision making, such as technological feasibility, 
air quality impacts, costs to industry and consumers, and lead time 
necessary for compliance.
    Other Factors: Beyond these statutory requirements applicable to 
DOT and/or EPA are a number of specific technical characteristics of 
CAFE and/or CO2 regulations that are also relevant to the 
construction of today's analysis. These are discussed at greater length 
in Section II.F. For example, EPA has defined procedures for 
calculating average CO2 levels, and has revised procedures 
for calculating CAFE levels, to reflect manufacturers' application of 
``off-cycle'' technologies that increase fuel economy (and reduce 
CO2 emissions) in ways not reflected by the long-standing 
test procedures used to measure fuel economy. Although too little 
information is available to account for these provisions explicitly in 
the same way that the agencies have accounted for other technologies, 
the CAFE Model does include and makes use of inputs reflecting the 
agencies' expectations regarding the extent to which manufacturers may 
earn such credits, along with estimates of corresponding costs. 
Similarly, the CAFE Model includes and makes use of inputs regarding 
credits EPA has elected to allow manufacturers to earn toward 
CO2 levels (not CAFE) based on the use of air conditioner 
refrigerants with lower global warming potential (GWP), or on the 
application of technologies to reduce refrigerant leakage. In addition, 
EPA has elected to provide that through model year 2021, manufacturers 
may apply ``multipliers'' to plug-in hybrid electric vehicles, 
dedicated electric vehicles, fuel cell vehicles, and hydrogen vehicles, 
such that when calculating a fleet's average CO2 levels (not 
CAFE), the manufacturer may, for example, ``count'' each electric 
vehicle twice. The CAFE Model accounts for these multipliers, based on 
either current regulatory provisions or on alternative approaches. 
Although these are examples of regulatory provisions that arise from 
the exercise of discretion rather than specific statutory mandate, they 
can materially impact outcomes. Section VI.B explains in greater detail 
how today's analysis addresses them.

Benefits of Analytical Approach

    The agencies' analysis of CAFE and CO2 standards 
involves two basic elements: First, estimating ways each manufacturer 
could potentially respond to a given set of standards in a manner that 
considers potential consumer response; and second, estimating various 
impacts of those responses. Estimating manufacturers' potential 
responses involves simulating manufacturers' decision-making processes 
regarding the year-by-year application of fuel-saving technologies to 
specific vehicles. Estimating impacts involves calculating resultant 
changes in new vehicle costs, estimating a variety of costs (e.g., for 
fuel) and effects (e.g., CO2 emissions from fuel combustion) 
occurring as vehicles are driven over their lifetimes before eventually 
being scrapped, and estimating the monetary value of these effects. 
Estimating impacts also involves consideration of the response of 
consumers--e.g., whether consumers will purchase the vehicles and in 
what quantities. Both of these basic analytical elements involve the 
application of many analytical inputs.
    As mentioned above, the agencies' analysis uses the CAFE model to 
estimate manufacturers' potential responses to new CAFE and 
CO2 standards and to estimate various impacts of those 
responses. DOT's Volpe National Transportation Systems Center (often 
simply referred to as the ``Volpe Center'') develops, maintains, and 
applies the model for NHTSA. NHTSA has used the CAFE model to perform 
analyses supporting every CAFE rulemaking since 2001, and the 2016 
rulemaking regarding heavy-duty pickup and van fuel consumption and 
CO2 emissions also used the CAFE model for analysis.\100\
---------------------------------------------------------------------------

    \100\ While both agencies used the CAFE Model to simulate 
manufacturers' potential responses to standards, some model inputs 
differed EPA's and DOT's analyses, and EPA also used the EPA MOVES 
model to calculate resultant changes in emissions inventories. See 
81 FR 73478, 73743 (Oct. 25, 2016).
---------------------------------------------------------------------------

    NHTSA recently arranged for a formal peer review of the model. In 
general, reviewers' comments strongly supported the model's conceptual 
basis and implementation, and commenters provided several specific 
recommendations. The agency agreed with many of these recommendations 
and has worked to implement them wherever practicable. Implementing 
some of the recommendations would require considerable further 
research, development, and testing, and will be considered going 
forward. For a handful of other recommendations, the agency disagreed, 
often finding the recommendations involved considerations (e.g., other 
policies, such as those involving fuel taxation) beyond the model 
itself or were based on concerns with inputs rather than how the model 
itself functioned. A report available in the docket for this rulemaking 
presents peer reviewers' detailed comments and recommendations, and 
provides DOT's detailed responses.\101\
---------------------------------------------------------------------------

    \101\ Docket No. NHTSA-2018-0067-0055.
---------------------------------------------------------------------------

    As also mentioned above, the agencies use EPA's MOVES model to 
estimate tailpipe emission factors, use DOE/EIA's NEMS to estimate fuel 
prices,\102\ and use Argonne's GREET model to estimate downstream 
emissions rates.\103\ DOT also sponsored DOE/Argonne to use the 
Autonomie full-vehicle modeling and simulation tool to estimate the 
fuel economy impacts for roughly a million

[[Page 24222]]

combinations of technologies and vehicle types.104 105
---------------------------------------------------------------------------

    \102\ See https://www.eia.gov/outlooks/aeo/info_nems_archive.php. Today's notice uses fuel prices estimated 
using the Annual Energy Outlook (AEO) 2019 version of NEMS (see 
https://www.eia.gov/outlooks/archive/aeo19/ and https://www.eia.gov/outlooks/aeo/data/browser/#/?id=3-AEO2019&cases=ref2019&sourcekey=0).
    \103\ Information regarding GREET is available at https://greet.es.anl.gov/index.php. Availability of NEMS is discussed at 
https://www.eia.gov/outlooks/aeo/info_nems_archive.php. Today's 
notice uses fuel prices estimated using the AEO 2019 version of 
NEMS.
    \104\ As part of the Argonne simulation effort, individual 
technology combinations simulated in Autonomie were paired with 
Argonne's BatPAC model to estimate the battery cost associated with 
each technology combination based on characteristics of the 
simulated vehicle and its level of electrification. Information 
regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
    \105\ Furthermore, the impact of engine technologies on fuel 
consumption, torque, and other metrics was characterized using GT 
POWER simulation modeling in combination with other engine modeling 
that was conducted by IAV Automotive Engineering, Inc. (IAV). The 
engine characterization ``maps'' resulting from this analysis were 
used as inputs for the Autonomie full-vehicle simulation modeling. 
Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
---------------------------------------------------------------------------

    EPA developed two models after 2009, referred to as the ``ALPHA'' 
and ``OMEGA'' models, which provide some of the same capabilities as 
the Autonomie and CAFE models. EPA applied the OMEGA model to conduct 
analysis of tailpipe CO2 emissions standards promulgated in 
2010 and 2012, and the ALPHA and OMEGA models to conduct analysis 
discussed in the above-mentioned 2016 Draft TAR and Proposed and 2017 
Initial Final Determinations regarding standards beyond 2021. In an 
August 2017 notice, the agencies requested comments on, among other 
things, whether EPA should use alternative methodologies and modeling, 
including DOE/Argonne's Autonomie full-vehicle modeling and simulation 
tool and DOT's CAFE model.\106\
---------------------------------------------------------------------------

    \106\ 82 FR 39551, 39553 (Aug. 21, 2017).
---------------------------------------------------------------------------

    Having reviewed comments on the subject and having considered the 
matter fully, the agencies have determined it is reasonable and 
appropriate to use DOE/Argonne's model for full-vehicle simulation, and 
to use DOT's CAFE model for analysis of regulatory alternatives. EPA 
interprets Section 202(a) of the CAA as giving the agency broad 
discretion in how it develops and sets CO2 emissions 
standards for light-duty vehicles. Nothing in Section 202(a) mandates 
that EPA use any specific model or set of models for analysis of 
potential CO2 standards for light-duty vehicles. EPA weighs 
many factors when determining appropriate levels for CO2 
standards, including the cost of compliance (see Section 202(a)(2)), 
lead time necessary for compliance (id.), safety (see NRDC v. EPA, 655 
F.2d 318, 336 n. 31 (D.C. Cir. 1981)) and other impacts on 
consumers,\107\ and energy impacts associated with use of the 
technology.\108\ Using the CAFE model allows consideration of a number 
of factors. The CAFE model explicitly evaluates the cost of compliance 
for each manufacturer, each fleet, and each model year; it accounts for 
lead time necessary for compliance by directly incorporating estimated 
manufacturer production cycles for every vehicle in the fleet, ensuring 
that the analysis does not assume vehicles can be redesigned to 
incorporate more technology without regard to lead time considerations; 
it provides information on safety effects associated with different 
levels of standards and information about many other impacts on 
consumers, and it calculates energy impacts (i.e., fuel saved or 
consumed) as a primary function, besides being capable of providing 
information about many other factors within EPA's broad CAA discretion 
to consider.
---------------------------------------------------------------------------

    \107\ Since its earliest Title II regulations, EPA has 
considered the safety of pollution control technologies. See 45 FR 
14496, 14503 (1980).
    \108\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-624 
(D.C. Cir. 1998) (ordinarily permissible for EPA to consider factors 
not specifically enumerated in the Act).
---------------------------------------------------------------------------

    Because the CAFE model simulates a wide range of actual constraints 
and practices related to automotive engineering, planning, and 
production, such as common vehicle platforms, sharing of engines among 
different vehicle models, and timing of major vehicle redesigns, the 
analysis produced by the CAFE model provides a transparent and 
realistic basis to show pathways manufacturers could follow over time 
in applying new technologies, which helps better assess impacts of 
potential future standards. Furthermore, because the CAFE model also 
accounts fully for regulatory compliance provisions (now including 
CO2 compliance provisions), such as adjustments for reduced 
refrigerant leakage, production ``multipliers'' for some specific types 
of vehicles (e.g., PHEVs), and carried-forward (i.e., banked) credits, 
the CAFE model provides a transparent and realistic basis to estimate 
how such technologies might be applied over time in response to CAFE or 
CO2 standards.
    There are sound reasons for the agencies to use the CAFE model 
going forward in this rulemaking. First, the CAFE and CO2 
fact analyses are inextricably linked. Furthermore, the analysis 
produced by the CAFE model and DOE/Argonne's Autonomie addresses the 
agencies' analytical needs. The CAFE model provides an explicit year-
by-year simulation of manufacturers' application of technology to their 
products in response to a year-by-year progression of CAFE standards 
and accounts for sharing of technologies and the implications for 
timing, scope, and limits on the potential to optimize powertrains for 
fuel economy. In the real world, standards actually are specified on a 
year-by-year basis, not simply some single year well into the future, 
and manufacturers' year-by-year plans involve some vehicles ``carrying 
forward'' technology from prior model years and some other vehicles 
possibly applying ``extra'' technology in anticipation of standards in 
ensuing model years, and manufacturers' planning also involves applying 
credits carried forward between model years. Furthermore, manufacturers 
cannot optimize the powertrain for fuel economy on every vehicle model 
configuration--for example, a given engine shared among multiple 
vehicle models cannot practicably be split into different versions for 
each configuration of each model, each with a slightly different 
displacement. The CAFE model is designed to account for these real-
world factors.
    Considering the technological heterogeneity of manufacturers' 
current product offerings, and the wide range of ways in which the many 
fuel economy-improving/CO2 emissions-reducing technologies 
included in the analysis can be combined, the CAFE model has been 
designed to use inputs that provide an estimate of the fuel economy 
achieved for many tens of thousands of different potential combinations 
of fuel-saving technologies. Across the range of technology classes 
encompassed by the analysis fleet, today's analysis involves more than 
a million such estimates. While the CAFE model requires no specific 
approach to developing these inputs, the National Academy of Sciences 
(NAS) has recommended, and stakeholders have commented, that full-
vehicle simulation provides the best balance between realism and 
practicality. DOE/Argonne has spent several years developing, applying, 
and expanding means to use distributed computing to exercise its 
Autonomie full-vehicle modeling and simulation tool over the scale 
necessary for realistic analysis of CAFE or average tailpipe 
CO2 emissions standards. This scalability and related 
flexibility (in terms of expanding the set of technologies to be 
simulated) makes Autonomie well-suited for developing inputs to the 
CAFE model.
    In addition, DOE/Argonne's Autonomie also has a long history of 
development and widespread application by a much wider range of users 
in government, academia, and industry. Many of these users apply

[[Page 24223]]

Autonomie to inform funding and design decisions. These real-world 
exercises have contributed significantly to aspects of Autonomie 
important to producing realistic estimates of fuel economy levels and 
CO2 emission rates, such as estimation and consideration of 
performance, utility, and driveability metrics (e.g., towing 
capability, shift business, frequency of engine on/off transitions). 
This steadily increasing realism has, in turn, steadily increased 
confidence in the appropriateness of using Autonomie to make 
significant investment decisions. Notably, DOE uses Autonomie for 
analysis supporting budget priorities and plans for programs managed by 
its Vehicle Technologies Office (VTO). Considering the advantages of 
DOE/Argonne's Autonomie model, it is reasonable and appropriate to use 
Autonomie to estimate fuel economy levels and CO2 emission 
rates for different combinations of technologies as applied to 
different types of vehicles.
    Commenters have also suggested that the CAFE model's graphical user 
interface (GUI) facilitates others' ability to use the model quickly--
and without specialized knowledge or training--and to comment 
accordingly.\109\ For the NPRM, NHTSA significantly expanded and 
refined this GUI, providing the ability to observe the model's real-
time progress much more closely as it simulates year-by-year compliance 
with either CAFE or CO2 standards.\110\ Although the model's 
ability to produce realistic results is independent of the model's GUI, 
the CAFE model's GUI appears to have facilitated stakeholders' 
meaningful review and comment during the comment period.
---------------------------------------------------------------------------

    \109\ From Docket Number EPA-HQ-OAR-2015-0827, see Comment by 
Global Automakers, Docket ID EPA-HQ-OAR-2015-0827-9728, at 34.
    \110\ The updated GUI provides a range of graphs updated in real 
time as the model operates. These graphs can be used to monitor fuel 
economy or CO2 ratings of vehicles in manufacturers' 
fleets and to monitor year-by-year CAFE (or average CO2 
ratings), costs, avoided fuel outlays, and avoided CO2-
related damages for specific manufacturers and/or specific fleets 
(e.g., domestic passenger car, light truck). Because these graphs 
update as the model progresses, they should greatly increase users' 
understanding of the model's approach to considerations such as 
multiyear planning, payment of civil penalties, and credit use.
---------------------------------------------------------------------------

    The question of whether EPA's actions should consider and be 
informed by analysis using non-EPA-staff-developed modeling tools has 
generated considerable debate over time. Even prior to the NPRM, 
certain commenters had argued that EPA could not consider, in setting 
tailpipe CO2 emissions standards, any information derived 
from non-EPA-staff-developed modeling. Many of the pre-NPRM concerns 
focused on inputs used by the CAFE model for prior rulemaking 
analyses.111 112 113 Because inputs are exogenous to any 
model, they do not determine whether it would be reasonable and 
appropriate for EPA to use NHTSA's model for analysis. Other concerns 
focused on certain characteristics of the CAFE model that were 
developed to align the model better with EPCA and EISA. The model has 
been revised to accommodate both EPCA/EISA and CAA analysis, as 
explained further below. Some commenters also argued that use of any 
models other than ALPHA and OMEGA for CAA analysis would constitute an 
arbitrary and capricious delegation of EPA's decision-making authority 
to NHTSA, if NHTSA models are used for analysis instead.\114\ As 
discussed above, the CAFE Model--as with any model--is used to provide 
analysis, and does not result in decisions. Decisions are made by EPA 
in a manner that is informed by modeling outputs, sensitivity cases, 
public comments, any many other pieces of information.
---------------------------------------------------------------------------

    \111\ For example, EDF previously stated that ``the data that 
NHTSA needs to input into its model is sensitive confidential 
business information that is not transparent and cannot be 
independently verified, . . .'' and it claimed ``the OMEGA model's 
focus on direct technological inputs and costs--as opposed to 
industry self-reported data--ensures the model more accurately 
characterizes the true feasibility and cost effectiveness of 
deploying greenhouse gas reducing technologies.'' EDF, EPA-HQ-OAR-
2015-0827-9203, at 12. These statements are not correct, as nothing 
about either the CAFE or OMEGA model either obviates or necessitates 
the use of CBI to develop inputs.
    \112\ As another example, CARB previously stated that ``another 
promising technology entering the market was not even included in 
the NHTSA compliance modeling'' and that EPA assumes a five-year 
redesign cycle, whereas NHTSA assumes a six to seven-year cycle.'' 
CARB, EPA-HQ-OAR-2015-0827-9197, at 28. Though presented as 
criticisms of the models, these comments--at least with respect to 
the CAFE model--actually concern model inputs. NHTSA did not agree 
with CARB about the commercialization potential of the engine 
technology in question (``Atkinson 2'') and applied model inputs 
accordingly. Also, rather than applying a one-size-fits-all 
assumption regarding redesign cadence, NHTSA developed estimates 
specific to each vehicle model and applied these as model inputs.
    \113\ As another example, NRDC has argued that EPA should not 
use the CAFE model because it ``allows manufacturers to pay civil 
penalties in lieu of meeting the standards, an alternative 
compliance pathway currently allowed under EISA and EPCA.'' NRDC, 
EPA-HQ-OAR-2015-0827-9826, at 37. While the CAFE model can simulate 
civil penalty payment, NRDC's comment appears to overlook the fact 
that this result depends on model inputs; the inputs can easily be 
specified such that the CAFE model will set aside civil penalty 
payment as an alternative to compliance.
    \114\ See, e.g., CBD et al., NHTSA-2018-0067-12057, at 9.
---------------------------------------------------------------------------

    Comments responding to the NPRM's use of the CAFE model and 
Autonomie rather than also (for CO2 standards) ALPHA and 
OMEGA were mixed. For example, the environmental group coalition stated 
that the CAFE model is aligned with EPCA requirements,\115\ but also 
argued (1) that EPA is legally prohibited from ``delegat[ing] technical 
decision-making to NHTSA;'' \116\ (2) that ``EPA must exercise its 
technical and scientific expertise'' to develop CO2 
standards and ``Anything less is an unlawful abdication of EPA's 
statutory responsibilities;'' \117\ (3) that EPA staff is much more 
qualified than DOT staff to conduct analysis relating to standards and 
has done a great deal of work to inform development of standards; \118\ 
(4) that ``The Draft TAR and 2017 Final Determination relied 
extensively on use of sophisticated EPA analytic tools and 
methodologies,'' i.e., the ``peer reviewed simulation model ALPHA,'' 
``the agency's vehicle teardown studies,'' and the ``peer-reviewed 
OMEGA model to make reasonable estimates of how manufacturers could add 
technologies to vehicles in order to meet a fleet-wide [CO2 
emissions] standard;'' \119\ (5) that NHTSA had said in the MYs 2012-
2016 rulemaking that the Volpe [CAFE] model was developed to support 
CAFE rulemaking and incorporates features ``that are not appropriate 
for use by EPA in setting [tailpipe CO2] standards;'' \120\ 
(6) allegations that some EPA staff had disagreed with aspects of the 
NPRM analysis and had requested that ``EPA's name and logo should be 
removed from the DOT-NHTSA Preliminary Regulatory Impact Analysis 
document'' and stated that ``EPA is relying upon the technical analysis 
performed by DOT-NHTSA for the [NPRM];'' \121\ (7) that EPA had 
developed ``a range of relevant new analysis'' that the proposal 
``failed to consider,'' including ``over a dozen 2017 and 2018 EPA peer 
reviewed SAE articles;'' \122\ (8) that EPA's OMEGA modeling undertaken 
during NPRM development ``found costs half that of NHTSA's findings,'' 
``Yet NHTSA did not correct the errors in its modeling and analysis, 
and the published proposal drastically overestimates the cost of 
complying . . . .;'' \123\ (9) that some EPA staff had requested that 
the technology ``HCR2'' be included in the NPRM analysis, ``Yet NHTSA 
overruled

[[Page 24224]]

EPA and omitted the technology;'' \124\ (10) that certain EPA staff had 
initially ``rejected use of the CAFE model for development of the 
proposed [tailpipe CO2] standards;'' \125\ (11) that there 
are ``many specific weaknesses of the modeling results derived in this 
proposal through use of the Volpe and Autonomie models'' and that the 
CAFE model is ``not designed in accordance with'' Section 202(a) of the 
CAA because (A) EPA ``is not required to demonstrate that standards are 
set at the maximum feasible level year-by-year,'' (B) because EPCA 
``preclude[s NHTSA] from considering vehicles powered by fuels other 
than gas or diesel'' and EPA is not similarly bound, and (C) because 
the CAFE model assumed that the value of an overcompliance credit 
equaled $5.50, the value of a CAFE penalty.\126\ Because of all of 
these things, the environmental group coalition stated that the 
proposal was ``unlawful'' and that ``Before proceeding with this 
rulemaking, EPA must consider all relevant materials including these 
excluded insights, perform its own analysis, and issue a reproposal to 
allow for public comment.'' \127\
---------------------------------------------------------------------------

    \115\ Environmental group coalition, NHTSA-2018-0067-12000, 
Appendix A, at 24-25.
    \116\ Id. at 12.
    \117\ Id. at 14.
    \118\ Id. at 15-17.
    \119\ Id. at 17.
    \120\ Id. at 18.
    \121\ Id. at 19.
    \122\ Id. at 20.
    \123\ Id. at 21.
    \124\ Id. at 21-22.
    \125\ Id. at 23.
    \126\ Id. at 24-25.
    \127\ Id. at 27.
---------------------------------------------------------------------------

    Some environmental organizations and States contracted for external 
technical analyses augmenting general comments such as those summarized 
above. EDF engaged a consultant, Richard Rykowski, for a detailed 
review of the agencies' analysis.\128\ Among Mr. Rykowski's comments, a 
few specifically involve differences between these two models. Mr. 
Rykowski recommended NHTSA's CAFE model replace its existing 
``effective cost'' metric (used to compare available options to add 
specific technologies to specific vehicles) with a ``ranking factor'' 
used for the same purpose. As discussed below in Section VI.A, the 
model for today's final rule adopts this recommendation. He also states 
that (1) ``EPA has developed a better way to isolate and reject cost 
ineffective combinations of technologies . . . [and] includes only 
these 50 or so technology combinations in their OMEGA model runs;'' (2) 
``NHTSA's arbitrary and rigid designation of leader-follower vehicles 
for engine, transmission and platform level technologies 
unrealistically slows the rollout of technology into the new vehicle 
fleet;'' (3) ``the Volpe Model is not capable of reasonably simulating 
manufacturers' ability to utilize CO2 credits to smooth the 
introduction of technology throughout their vehicle line-up;'' and (4) 
``the Volpe Model is not designed to reflect the use of these [A/C 
leakage] technologies and refrigerants.'' \129\
---------------------------------------------------------------------------

    \128\ EDF, NHTSA-2018-0067-12108, Appendix B. See also EPA, Peer 
Review of the Optimization Model for Reducing Emissions of 
Greenhouse Gases from Automobiles (OMEGA) and EPA's Response to 
Comments, EPA-420-R-09-016, September 2009.
    \129\ EDF, op. cit., at 73-75.
---------------------------------------------------------------------------

    Mr. Rogers's analysis focuses primarily on the agencies' published 
analysis, but mentions that some engine ``maps'' (estimates--used as 
inputs to full vehicle simulation--of engine fuel consumption under a 
wide range of engine operating conditions) applied in Autonomie show 
greater fuel consumption benefits of turbocharging than those applied 
previously by EPA to EPA's ALPHA model, and these benefits could have 
caused NHTSA's CAFE model to estimate an unrealistically great tendency 
toward turbocharged engines (rather than high compression ratio 
engines).\130\ Mr. Rogers also presents alternative examples of year-
by-year technology application to specific vehicle models, contrasting 
these with year-by-year results from the agencies' NPRM analysis, 
concluding that ``that the use of logical, unrestricted technology 
pathways, with incremental benefits supported by industry-accepted 
vehicle simulation and dynamic system optimization and calibration, 
together with publicly-defensible costs, allows cost-effective 
solutions to achieve target fuel economy levels which meet MY 2025 
existing standards.'' \131\
---------------------------------------------------------------------------

    \130\ Roush Industries, NHTSA-2018-0067-11984, at 17-21.
    \131\ Roush Industries, NHTSA-2018-0067-11984, at 17-30.
---------------------------------------------------------------------------

    Mr. Duleep's analysis also focuses primarily on the agencies' 
published analysis, but does mention that (1) ``the Autonomie modeling 
assumes no engine change when drag and rolling resistance reductions 
are implemented, as well as no changes to the transmission gear ratios 
and axle ratios, . . . [but] the EPA ALPHA model adjusts for this 
effect;'' (2) ``baseline differences in fuel economy [between two 
manufacturers' different products using similar technologies] are 
carried for all future years and this exaggerates the differences in 
technology adoption requirements and costs between manufacturers; (3) 
``assumptions [that most technology changes are best applied as part of 
a vehicle redesign or freshening] result in unnecessary distortion in 
technology paths and may bias results of costs for different 
manufacturers;'' and (4) that for the sample results shown for the 
Chevrolet Equinox ``the publicly available EPA lumped parameter model 
(which was used to support the 2016 rulemaking) and 2016 TAR cost data 
. . . results in an estimate of attaining 52.2 mpg for a cost of $2110, 
which is less than half the cost estimated in the PRIA.'' \132\
---------------------------------------------------------------------------

    \132\ H-D Systems, op. cit., at 48, et seq.
---------------------------------------------------------------------------

    Beyond these comments regarding differences between EPA's models 
and the Argonne and DOT models applied for the NPRM, these and other 
technical reviewers had many specific comments about the agencies' 
analysis for the NPRM, and these comments are discussed in detail below 
in Section VI.B.
    Manufacturers, on the other hand, supported the agencies' use of 
Autonomie and the CAFE model rather than, in EPA's case, the ALPHA and 
OMEGA models. Expressly identifying the distinction between models and 
model inputs, Global Automakers stated that:

    The agencies provided a new, updated analysis based on the most 
up-to-date data, using a proven and long-developed modeling tool, 
known as the Volpe model, and offering numerous options to best 
determine the right regulatory and policy path for ongoing fuel 
efficiency improvements in our nation. Now, all stakeholders have an 
opportunity to come to the table as part of the public process to 
provide input, data, and information to help shape the final 
rule.\133\
---------------------------------------------------------------------------

    \133\ Global Automakers, NHTSA-2018-0067-12032, at 2.
---------------------------------------------------------------------------

    This NPRM's use of a single model to evaluate alternative 
scenarios for both programs provides consistency in the technical 
analysis, and Global Automakers supports the Volpe model's use as it 
has proven to be a transparent and user-friendly option in this 
current analysis. The use of the Volpe model has allowed for a broad 
range of stakeholders, with varying degrees of technical expertise, 
to review the data inputs to provide feedback on this proposed rule. 
The Volpe model's accompanying documentation has historically 
provided a clear explanation of all sources of input and constraints 
critical to a transparent modeling process. Other inputs have come 
from modeling that is used widely by other sources, specifically the 
Autonomie model, allowing for a robust validation, review and 
reassessment.\134\
---------------------------------------------------------------------------

    \134\ Global Automakers, NHTSA-2018-0067-12032, Attachment A, at 
A-12.

    The Alliance commented, similarly, that ``at least at this time, 
NHTSA's modeling systems are superior to EPA's'' and ``as such, we 
support the Agencies' decision to use NHTSA's modeling tools for this 
rulemaking and recommend that both Agencies continue on this path. We 
encourage Agencies to work together to provide input to the single 
common set of tools.'' \135\
---------------------------------------------------------------------------

    \135\ Alliance, NHTSA-2018-0067-12073, at 134.

---------------------------------------------------------------------------

[[Page 24225]]

    Regarding the agencies' use of Argonne's Autonomie model rather 
than EPA's ALPHA model, the Alliance commented that (1) ``the benefits 
of virtually all technologies and their synergistic effects are now 
determined with full vehicle simulations;'' (2) ``vehicle categories 
have been increased to 10 to better recognize the range of 0-60 
performance characteristics within each of the 5 previous categories, 
in recognition of the fact that many vehicles in the baseline fleet 
significantly exceeded the previously assumed 0-60 performance metrics. 
This provides better resolution of the baseline fleet and more accurate 
estimates of the benefits of technology. . . .;'' (3) ``new 
technologies (like advanced cylinder deactivation) are included, while 
unproven combinations (like Atkinson engines with 14:1 compression, 
cooled EGR, and cylinder deactivation in combination) have been 
removed;'' (4) ``Consistent with the recommendation of the National 
Academy of Sciences and manufacturers, gradeability has been included 
as a performance metric used in engine sizing. This helps prevent the 
inclusion of small displacement engines that are not commercially 
viable and that would artificially inflate fuel savings;'' (5) ``the 
Alliance believes NHTSA's tools (Autonomie/Volpe) are superior to EPA's 
(APLHA[sic]/LPM/OMEGA). This is not surprising since NHTSA's tools have 
had a significant head start in development. . . .'' (6) ``the 
Autonomie model was developed at Argonne National Lab with funding from 
the Department of Energy going back to the PNGV (Partnership for Next 
Generation Vehicles) program in the 1990s. Autonomie was developed from 
the start to address the complex task of combining 2 power sources in a 
hybrid powertrain. It is a physics-based, forward looking, vehicle 
simulator, fully documented with available training,'' and (7) ``EPA's 
ALPHA model is also a physics-based, forward looking, vehicle 
simulator. However, it has not been validated or used to simulate 
hybrid powertrains. The model has not been documented with any 
instructions making it difficult for users outside of EPA to run and 
interpret the model.'' \136\
---------------------------------------------------------------------------

    \136\ Id. at 135.
---------------------------------------------------------------------------

    Regarding the use of NHTSA's CAFE model rather than EPA's OMEGA 
model, the Alliance stated that (1) NHTSA's model appropriately 
differentiate between domestic and imported automobiles; (2) in NHTSA's 
model, ``dynamic estimates of vehicle sales and scrappage in response 
to price changes replace unrealistic static sales and scrappage 
numbers;'' (3) NHTSA's model ``has new capability to perform 
[CO2 emissions] analysis with [tailpipe CO2] 
program flexibilities;'' (4) ``the baseline fleet [used in NHTSA's 
model] has been appropriately updated based on both public and 
manufacturer data to reflect the technologies already applied, 
particularly tire rolling resistance;'' and (5) ``some technologies 
have been appropriately restricted. For example, low rolling resistance 
tires are no longer allowed on performance vehicles, and aero 
improvements are limited to maximum levels of 15% for trucks and 10% 
for minivans.'' \137\ The Alliance continued, noting that ``NHTSA's 
Volpe model also predates EPA's OMEGA model. More importantly, the new 
Volpe model considers several factors that make its results more 
realistic.'' \138\ As factors leading the Volpe model to produce 
results that are more realistic than those produced by OMEGA, the 
Alliance commented that (1) ``The Volpe model includes estimates of the 
redesign and refresh schedules of vehicles based on historical trends, 
whereas the OMEGA model uses a fixed, and too short, time interval 
during which all vehicles are assumed to be fully redesigned. . . .;'' 
(2) ``The Volpe model allows users to phase-in technology based on year 
of availability, platform technology sharing, phase-in caps, and to 
follow logical technology paths per vehicle. . . .;'' (3) ``The Volpe 
model produces a year-by year analysis from the baseline model year 
through many years in the future, whereas the OMEGA model only analyzes 
a fixed time interval. . . .;'' (4) ``The Volpe model recognizes that 
vehicles share platforms, engines, and transmissions, and that 
improvements to any one of them will likely extend to other vehicles 
that use them'' whereas ``The OMEGA model treats each vehicle as an 
independent entity. . . .;'' (5) ``The Volpe model now includes sales 
and scrappage effects;'' and (6) ``The Volpe model is now capable of 
analyzing for CAFE and [tailpipe CO2] compliance, each with 
unique program restrictions and flexibilities.'' \139\ The Alliance 
also incorporated by reference concerns it raised regarding EPA's 
OMEGA-based analysis supporting EPA's proposed and prior final 
determinations.\140\
---------------------------------------------------------------------------

    \137\ Id. at 134.
    \138\ Id. at 135.
    \139\ Id. at 135-136.
    \140\ Id. at 136.
---------------------------------------------------------------------------

    The Alliance further stated that ``For all of the above reasons and 
to avoid duplicate efforts, the Alliance recommends that the Agencies 
continue to use DOT's Volpe and Autonomie modeling system, rather than 
continuing to develop two separate systems. EPA has demonstrated 
through supporting Volpe model code revisions and by supplying engine 
maps for use in the Autonomie model that their expertise can be 
properly represented in the rulemaking process without having to 
develop separate or new tools.'' \141\
---------------------------------------------------------------------------

    \141\ Id. at 136.
---------------------------------------------------------------------------

    Some individual manufacturers provided comments supporting and 
elaborating on the above comments by Global Automakers and the 
Alliance. For example, FCA commented that ``the modeling performed by 
the agencies should illuminate the differences between the CAFE and 
[tailpipe CO2 emissions] programs. This cannot be 
accomplished when each agency is using different tools and assumptions. 
Since we believe NHTSA possesses the better set of tools, we support 
both agencies using Autonomie for vehicle modeling and Volpe (CAFE) for 
fleet modeling.'' \142\
---------------------------------------------------------------------------

    \142\ FCA, NHTSA-2018-0067-11943, at 82.
---------------------------------------------------------------------------

    Honda stated that ``The current version of the CAFE model is 
reasonably accurate in terms of technology efficiency, cost, and 
overall compliance considerations, and reflects a notable improvement 
over previous agency modeling efforts conducted over the past few 
years. We found the CAFE model's characterization of Honda's 
``baseline'' fleet--critical modeling minutiae that provide a technical 
foundation of the agencies' analysis--to be highly accurate. We commend 
NHTSA and Volpe Center staff on these updates, as well as on the 
overall transparency of the model. The model's graphical user interface 
(GUI) makes it easier to run, model functionality is thoroughly 
documented, and the use of logical, traceable input and output files 
accommodates easy tracking of results.'' \143\ Similarly, in an earlier 
presentation to the agencies, Honda included the following slide 
comparing EPA's OMEGA model to DOT's CAFE (Volpe) model, and making 
recommendations regarding future improvements to the latter: \144\
---------------------------------------------------------------------------

    \143\ Honda, EPA-HQ-OAR-2018-0283, at 21-22.
    \144\ Honda, NHTSA-2018-0067-12019, at 12.

---------------------------------------------------------------------------

[[Page 24226]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.041

    Toyota, in addition to arguing that the agencies' application of 
model inputs (e.g., an analysis fleet based on MY 2016 compliance data) 
produced more realistic results than in the draft TAR and in EPA's 
former proposed and final determinations, also stressed the importance 
of the CAFE model's year-by-year accounting for product redesigns, 
stating that this produces more realistic results than the OMEGA-based 
results shown previously by EPA:

    The modeling now better accounts for factors that limit the rate 
at which new technologies enter and then diffuse through a 
manufacturer's fleet. Bringing new or improved vehicles and 
technologies to market is a several-year, capital-intensive 
undertaking. Once new designs are introduced, a period of stability 
is required so investments can be amortized. Vehicle and technology 
introductions are staggered over time to manage limited resources. 
Agency modeling now better recognizes the inherent constraints 
imposed by realities that dictate product cadence. We agree with the 
agencies' understanding that ``the simulation of compliance actions 
that manufacturers might take is constrained by the pace at which 
new technologies can be applied in the new vehicle market,'' and we 
are encouraged to learn that ``agency modeling can now account for 
the fact that individual vehicle models undergo significant 
redesigns relatively infrequently.'' The preamble correctly notes 
that manufacturers try to keep costs down by applying most major 
changes mainly during vehicle redesigns and more modest changes 
during product refresh, and that redesign cycles for vehicle models 
can range from six to ten years, and eight to ten-years for 
powertrains. This appreciation for standard business practice 
enables the modeling to more accurately capture the way vehicles 
share engines, transmissions, and platforms. There are now more 
realistic limits placed on the number of engines and transmissions 
in a powertrain portfolio which better recognizes manufacturers must 
manage limited engineering resources and control supplier, 
production, and service costs. Technology sharing and inheritance 
between vehicle models tends to limit the rate of improvement in a 
manufacturer's fleet.\145\
---------------------------------------------------------------------------

    \145\ Toyota, NHTSA-2018-0067-12098, Attachment 1, at 3 et seq.

    These comments urging EPA to use NHTSA's CAFE model echo comments 
provided in response to a 2018 peer review of the model. While 
identifying various opportunities for improvement, peer reviewers 
expressed strong overall support for the CAFE model's technical 
approach and execution. For example, one reviewer, after offering many 
---------------------------------------------------------------------------
specific technical recommendations, concluded as follows:

    The model is impressive in its detail, and in the completeness 
of the input data that it uses. Although the model is complex, the 
reader is given a clear account of how variables are variously 
divided and combined to yield appropriate granularity and efficiency 
within the model. The model tracks well a simplified version of the 
real-world and manufacturing/design decisions. The progression of 
technology choices and cost benefit choices is clear and logical. In 
a few cases, the model simply explains a constraint, or a value 
assigned to a variable, without defending the choice of the value or 
commenting on real-world variability, but these are not substantive 
omissions. The model will lend itself well to future adaptation or 
addition of variables, technologies and pathways.\146\
---------------------------------------------------------------------------

    \146\ NHTSA, CAFE Model Peer Review, DOT HS 812 590, Available 
at https://www.nhtsa.gov/document/cafe-model-peer-review, at 250.

    Although the peer review charge focused solely on the CAFE model, 
another peer reviewer separately recommended that EPA ``consider 
opportunities for EPA to use the output from the Volpe Model in place 
of their OMEGA Model output'' \147\
---------------------------------------------------------------------------

    \147\ Id. at 287-288 and 304.
---------------------------------------------------------------------------

    More recently, in response to the NPRM, Dr. Julian Morris, an 
economist at George Washington University, commented extensively on the 
superiority of the agencies' NPRM analysis to previous analyses, 
offering the following overall assessment:

    I have assessed the plausibility of the analyses undertaken by 
NHTSA and EPA in relation to the proposed SAFE rule. I found that 
the agencies have undertaken a thorough--one might even say 
exemplary--analysis, improving considerably on earlier analyses 
undertaken by the agencies of previous rules relating to CAFE 
standards and [tailpipe CO2] emission standards. Of 
particular note, the agencies included more realistic estimates of 
the rebound effect, developed a sophisticated model of the

[[Page 24227]]

scrappage effect, and better accounted for various factors affecting 
vehicle fatality rates.\148\
---------------------------------------------------------------------------

    \148\ Morris, J., OAR-2018-0283-4028, at 6-11.

    The agencies carefully considered these and other comments 
regarding which models to apply when estimating potential impacts of 
each of the contemplated regulatory alternatives. For purposes of 
estimating the impacts of CAFE standards, even the coalition of 
environmental advocates observed that the CAFE model reflects EPCA's 
requirements. As discussed below in Section VI.A, EPCA imposes specific 
requirements not only on how CAFE standards are to be structured (e.g., 
including a minimum standard for domestic passenger cars), but also on 
how CAFE standards are to be evaluated (e.g., requiring that the 
potential to produce additional AFVs be set aside for the model years 
under consideration), and the CAFE model reflects these requirements, 
and the agencies consider the CAFE model to be the best available tool 
for CAFE rulemaking analysis. Regarding the use of Autonomie to 
construct fuel consumption (i.e., efficiency) inputs to the CAFE model, 
the agencies recognize that other vehicle simulation tools are 
available, including EPA's recently-developed ALPHA model. However, as 
also discussed in Section VI.B.3, Autonomie has a much longer history 
of development and refinement, and has been much more widely applied 
and validated. Moreover, Argonne experts have worked carefully for 
several years to develop methods for running large arrays of 
simulations expressly structured and calibrated for use in DOT's CAFE 
model. Therefore, the agencies consider Autonomie to be the best 
available tool for constructing such inputs to the CAFE model. While 
the agencies have also carefully considered potential specific model 
refinements, as well as the merits of potential changes to model inputs 
and assumptions, none of these potential refinements and input have led 
either agency to reconsider using the CAFE model and Autonomie for CAFE 
rulemaking analysis.
    With respect to estimating the impacts of CO2 standards, 
even though Argonne and the agencies have adapted Autonomie and the 
CAFE model to support the analysis of CO2 standards, 
environmental groups, California, and other States would prefer that 
EPA use the models it developed during 2009-2018 for that purpose.\149\ 
Arguments that EPA revert to its ALPHA and OMEGA models fall within 
three general categories: (1) Arguments that EPA's models would have 
selected what commenters consider better (i.e., generally more 
stringent) standards, (2) arguments that EPA's models are technically 
superior, and (3) arguments that the law requires EPA use its own 
models.
---------------------------------------------------------------------------

    \149\ The last-finalized versions of EPA's OMEGA model and ALPHA 
tools were published in 2016 and 2017, respectively.
---------------------------------------------------------------------------

    The first of these arguments--that EPA's models would have selected 
better standards--conflates the analytical tool used to inform 
decision-making with the action of making the decision. As explained 
elsewhere in this document and as made repeatedly clear over the past 
several rulemakings, the CAFE model (or, for that matter, any model) 
neither sets standards nor dictates where and how to set standards; it 
simply informs as to the potential effects of setting different levels 
of standards. In this rulemaking, EPA has made its own decisions 
regarding what CO2 standards would be appropriate under the 
CAA.
    The third of these arguments--that EPA is legally required to use 
only models developed by its own staff--is also without merit. The CAA 
does not require the agency to create or use a specific model of its 
own creation in setting tailpipe CO2 standards. The fact 
that EPA's decision may be informed by non-EPA-created models does not, 
in any way, constitute a delegation of its statutory power to set 
standards or decision-making authority.\150\ Arguing to the contrary 
would suggest, for example, that EPA's decision would be invalid 
because it relied on EIA's Annual Energy Outlook for fuel prices for 
all of its regulatory actions rather than developing its own model for 
estimating future trends in fuel prices. Yet, all Federal agencies that 
have occasion to use forecasts of future fuel prices regularly (and 
appropriately) defer to EIA's expertise in this area and rely on EIA's 
NEMS-based analysis in the AEO, even when those same agencies are using 
EIA's forecasts to inform their own decision-making. Similarly, this 
argument would mean that the agencies could not rely on work done by 
contractors or other outside consultants, which is contrary to regular 
agency practice across the entirety of the Federal Government.
---------------------------------------------------------------------------

    \150\ ``[A] federal agency may turn to an outside entity for 
advice and policy recommendations, provided the agency makes the 
final decisions itself.'' U.S. Telecom. Ass'n v. FCC, 359 F.3d 554, 
565-66 (D.C. Cir. 2004). To the extent commenters meant to suggest 
outside parties have a reliance interest in EPA using ALPHA and 
OMEGA to set standards, EPA and NHTSA do not agree a reliance 
interest is properly placed on an analytical methodology, which 
consistently evolves from rule to rule. Even if it were, all parties 
that closely examined ALPHA and OMEGA-based analyses in the past 
either also simultaneously closely examined CAFE and Autonomie-based 
analyses in the past, or were fully capable of doing so, and thus, 
should face no additional difficulty now they have only one set of 
models and inputs/outputs to examine.
---------------------------------------------------------------------------

    The specific claim here that use of the CAFE model instead of ALPHA 
and OMEGA is somehow illegitimate is similarly unpersuasive. The CAFE 
and CO2 rules have, since Massachusetts v. EPA, all been 
issued as joint rulemakings, and, thus are the result of a 
collaboration between the two agencies. This was true when the 
rulemakings used separate models for the different programs and 
continues to be true in today's final rule, where the agencies take the 
next step in their collaborative approach by now using simply one model 
to simulate both programs. In 2007, immediately following this Supreme 
Court decision, the agencies worked together toward standards for model 
years 2011-2015, and EPA made use of the CAFE model for its work toward 
possible future CO2 standards. That the agencies would need 
to continue the unnecessary and inefficient process of using two 
separate combinations of models as the joint National Program continues 
to mature, therefore, runs against the idea that the agencies, over 
time, would best combine resources to create an efficient and robust 
regulatory program. For the reasons discussed throughout today's final 
rule, the agencies have jointly determined that Autonomie and the CAFE 
model have significant technical advantages, including important 
additional features, and are therefore the more appropriate models to 
use to support both analyses.
    Further, the fact that Autonomie and CAFE models were initially 
developed by DOE/Argonne and NHTSA does not mean that EPA has no role 
in either these models or their inputs. That is, the development 
process for CAFE and CO2 standards inherently requires 
technical and policy examinations and deliberations between staff 
experts and decision-makers in both agencies. Such engagements are a 
healthy and important part of any rulemaking activity--and particularly 
so with joint rulemakings. The Supreme Court stated in Massachusetts v. 
EPA that, ``The two obligations [to set CAFE standards under EPCA and 
to set tailpipe CO2 emissions standards under the CAA] may 
overlap, but there is no reason to think the two agencies cannot both 
administer their obligations and yet

[[Page 24228]]

avoid inconsistency.'' \151\ When agency experts consider analytical 
issues and agency decision-makers decide on policy, which is informed 
(albeit not dictated) by the outcome of that work, they are working 
together as the Court appears to have intended in 2007, even if 
legislators' intentions have varied in the decades since EPCA and the 
CAA have been in place.\152\ Regulatory overlap necessarily involves 
deliberation, which can lead to a more balanced, reasonable, and 
improved analyses, and better regulatory outcomes. It did here. The 
existence of deliberation is not per se evidence of unreasonableness, 
even if some commenters believe a different or preferred policy outcome 
would or should have resulted.\153\
---------------------------------------------------------------------------

    \151\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007).
    \152\ For example, when wide-ranging amendments to the CAA were 
being debated, S. 1630 contained provisions that, if enacted, would 
have authorized automotive CO2 emissions standards and 
prescribed specific average levels to be achieved by 1996 and 2000. 
In a letter to Senators, then-Administrator William K. Reilly noted 
that the Bill ``requires for the first time control of emissions of 
carbon dioxide; this is essentially a requirement to improve fuel 
efficiency'' and outlined four reasons the H.W. Bush Administration 
opposed the requirement, including that ``it is inappropriate to add 
this very complex issue to the Clean Air Act which is already full 
of complicated and controversial issues.'' Reilly, W., Letter to 
U.S. Senators (January 26, 1990). The CAA amendments ultimately 
signed into law did not contain these or any other provisions 
regarding regulation of CO2 emissions.
    \153\ See, e.g., U.S. House of Representatives, Committee on 
Oversight and Government Reform, Staff Report, 112th Congress, ``A 
Dismissal of Safety, Choice, and Cost: The Obama Administration's 
New Auto Regulations,'' August 10, 2012, at 19-21 and 33-34.
---------------------------------------------------------------------------

    Over the 44 years since EPCA established the requirement for CAFE 
standards, NHTSA, EPA and DOE career staff have discussed, collaborated 
on, and debated engineering, economic, and other aspects of CAFE 
regulation, through focused meetings and projects, informal exchanges, 
publications, conferences and workshops, and rulemakings.
    Part of this expanded exchange has involved full vehicle 
simulation. While tools such as PSAT (the DOE-sponsored simulation tool 
that predated Autonomie) were in use prior to 2007, including for 
discrete engineering studies supporting inputs to CAFE rulemaking 
analyses, these tools' information and computing requirements were such 
that NHTSA had determined (and DOE and EPA had concurred) that it was 
impractical to more fully integrate full vehicle simulation into 
rulemaking analyses. Since that time, computing capabilities have 
advanced dramatically, and the agencies now agree that such integration 
of full vehicle simulation--such as the large-scale exercise of 
Autonomie to produce inputs to the CAFE Model--can make for more robust 
CAFE and CO2 rulemaking analysis. This is not to say, 
though, that experts always agree on all methods and inputs involved 
with full vehicle simulation. Differences in approach and inputs lead 
to differences in results. For example, compared to other publicly 
available tools that can be practicably exercised at the scale relevant 
to fleetwide analysis needed for CAFE and CO2 rulemaking 
analysis, DOE/Argonne's Autonomie model is more advanced, spans a wider 
range of fuel-saving technologies, and represents them in more specific 
detail, leaving fewer ``gaps'' to be filled with other models (risking 
inconsistencies and accompanying errors). These differences discussed 
in greater detail below in Section VI.B.3. Perhaps most importantly, 
the CAFE model considers fuel prices in determining both which 
technologies are applied and the total amount of technology applied, in 
the case where market forces demand fuel economy levels in excess of 
the standards. While OMEGA can apply technology in consideration of 
fuel prices, OMEGA will apply technology to reach the same level of 
fuel economy (or CO2 emissions) if fuel prices are 3, 5, or 
20 dollars, which violates the SAB's requirement that the analysis 
``account for [. . .] future fuel prices .'' \154\ Furthermore, it 
produces a counterintuitive result. If fuel prices become exorbitantly 
high, we would expect consumers to place an emphasis on additional fuel 
efficiency as the potential for extra fuel savings is tremendous.
---------------------------------------------------------------------------

    \154\ See SAB Report 10 (``Constructing each of the scenarios is 
challenging and involve extensive scientific, engineering, and 
economic uncertainties. Projecting the baseline requires the 
agencies to account for a wide range of variables including: The 
number of new vehicles sold, future fuel prices,. . . .'').
---------------------------------------------------------------------------

    Moreover, DOE has for many years used Autonomie (and its precursor 
model, PSAT) to produce analysis supporting fuel economy-related 
research and development programs involving billions of dollars of 
public investment, and NHTSA's CAFE model with inputs from DOE/
Argonne's Autonomie model has produced analysis supporting rulemaking 
under the CAA. In 2015, EPA proposed new tailpipe CO2 
standards for MY 2021-2027 heavy-duty pickups and vans, finalizing 
those standards in 2016. Supporting the NPRM and final rule, EPA relied 
on analysis implemented by NHTSA using NHTSA's CAFE model, and NHTSA 
used inputs developed by DOE/Argonne using DOE/Argonne's Autonomie 
model. CBD questioned this history, asserting that, ``EPA conducted a 
separate analysis using a different iteration of the CAFE model rather 
than rely on the version which NHTSA used, again resulting and parallel 
but corroborative modeling results.'' \155\ CBD's comment 
mischaracterizes EPA's actual use of the CAFE Model. As explained in 
the final rule, EPA's ``Method B'' analysis was developed as follows:
---------------------------------------------------------------------------

    \155\ CBD, et al., 2018-0067-12000, Appendix A, at 27.

    In Method B, the CAFE model from the NPRM was used to project a 
pathway the industry could use to comply with each regulatory 
alternative, along with resultant impacts on per-vehicle costs. 
However, the MOVES model was used to calculate corresponding changes 
in total fuel consumption and annual emissions for pickups and vans 
in Method B. Additional calculations were performed to determine 
corresponding monetized program costs and benefits.\156\
---------------------------------------------------------------------------

    \156\ 81 FR 73478, 73506-07 (October 25, 2016).

    In other words, a version of NHTSA's CAFE Model was used to perform 
the challenging part of the analysis--that is, the part that involves 
accounting for manufacturers' fleets, accounting for available fuel-
saving technologies, accounting for standards under consideration, and 
estimating manufacturers' potential responses to new standards--EPA's 
MOVES model was used to perform ``downstream'' calculations of fuel 
consumption and tailpipe emissions, and used spreadsheets to calculate 
even more straightforward calculations of program costs and benefits. 
While some stakeholders perceive these differences as evidencing a 
meaningfully independent approach, in fact, the EPA staff's analysis 
was, at its core, wholly dependent on NHTSA's CAFE Model, and on that 
model's use of Autonomie simulations.
    Given the above, the only remaining argument for EPA to revert to 
its previously-developed models rather than relying on Autonomie and 
the CAFE model would be that the former are so technically superior to 
the latter that even model refinements and input changes cannot lead 
Autonomie and the CAFE model to produce appropriate and reasonable 
results for CO2 rulemaking analysis. As discussed below, 
having considered a wide range of technical differences, the agencies 
find that the Autonomie and CAFE models currently provide the best 
analytical combination for CAFE and tailpipe CO2 emissions 
rulemaking analysis. As discussed

[[Page 24229]]

below in Section VI.B.3, Autonomie not only has a longer and wider 
history of development and application, but also DOE/Argonne's 
interaction with automakers, supplier and academies on continuous bases 
had made individual sub-models and assumptions more robust. Argonne has 
also been using research from DOE's Vehicle Technology Office (VTO) at 
the same time to make continuous improvements in Autonomie.\157\ Also, 
while Autonomie uses engine maps as inputs, and EPA developed engine 
maps that could have been used for today's analysis, EPA declined to do 
so, and those engine maps were only used in a limited capacity for 
reasons discussed below in Section VI.C.1.
---------------------------------------------------------------------------

    \157\ U.S. DOE Benefits & Scenario Analysis publications is 
available at https://www.autonomie.net/publications/fuel_economy_report.html. Last accessed November 14, 2019.
---------------------------------------------------------------------------

    As also discussed below in Section VI.A.4, the CAFE model accounts 
for some important CO2 provisions that EPA's OMEGA model 
cannot account for. For example, the CAFE model estimates the potential 
that any given manufacturer might apply CO2 compliance 
credits it has carried forward from some prior model year. While one 
commenter, Mr. Rykowski, takes issue with how the CAFE model handles 
credit banking, he does not acknowledge that EPA's OMEGA model, lacking 
a year-by-year representation of compliance, is altogether incapable of 
accounting for the earning and use of banked compliance credits. Also, 
although Mr. Rykowski's comments regarding A/C leakage and refrigerants 
are partially correct insofar as the CAFE model does not account for 
leakage-reducing technologies explicitly, the comment is as applicable 
to OMEGA as it is to the CAFE model and, in any event, data regarding 
which vehicles have which leakage-reducing technologies was not 
available for the MY 2016 fleet. Nevertheless, as discussed in Section 
VI.A.4, NHTSA has refined the CAFE model's accounting for the cost of 
leakage reduction technologies.
    The agencies have also considered Mr. Rykowski's comments 
suggesting that using OMEGA would be preferable because, rather than 
selecting from hundreds of thousands of potential combinations of 
technologies, OMEGA includes only the ``50 or so'' combinations that 
EPA has already determined to be cost-effective. The ``better way'' of 
making this determination is also effectively a model, but the 
separation of this model from OMEGA is, as evidenced by manufacturers' 
comments, obfuscatory, especially in terms of revealing how specific 
vehicle model/configurations initial engineering properties are aligned 
with specific initial technology combinations. By using a full set of 
technology combinations, the CAFE model makes very clear how each 
vehicle model/configuration is assigned to a specific initial 
combination and, hence, how subsequently fuel consumption and cost 
changes are accounted for. Moreover, EPA's separation of ``thinning'' 
process from OMEGA's main compliance simulation makes sensitivity 
analysis difficult to implement, much less follow. The agencies find, 
therefore, that the CAFE model's approach of retaining a full set of 
vehicle simulation results throughout the compliance simulation to be 
more realistic (e.g., more capable of reflecting manufacturer- and 
vehicle-specific factors), more responsive to changes in model inputs 
(e.g., changes to fuel prices, which could impact the relative 
attractiveness of different technologies), more transparent, and more 
amenable to independent corroboration the agencies' analysis.
    Regarding comments by Messrs. Duleep, Rogers, and Rykowski 
suggesting that the CAFE model, by tying most technology application to 
planned vehicle redesigns and freshening, is too restrictive, the 
agencies disagree. As illustrated by manufacturers' comments cited 
above, as reinforced by both extensive product planning information 
provided to the agencies, and as further reinforced by extensive 
publicly available information, manufacturers tend to not make major 
changes to a specific vehicle model/configuration in one model year, 
and then make further major changes to the same vehicle model/
configuration the next model year. There is ample evidence that 
manufacturers strive to avoid such discontinuity, complexity, and 
waste, and in the agencies' view, while it is impossible to represent 
every manufacturer's decision-making process precisely and with 
certainty, the CAFE model's approach of using estimated product design 
schedules provides a realistic basis for estimating what manufacturers 
could practicably do. Also, the relevant inputs are simply inputs to 
the CAFE model, and if it is actually more realistic to assume that a 
manufacturer can change major technology on all of its products every 
year, the CAFE model can easily be operated with every model year 
designated as a redesign year for every product, but as discussed 
throughout this document, the agencies consider this to be extremely 
unrealistic. While this means the CAFE model can be run without a year-
by-year representation that carries forward technologies between model 
years, doing so would be patently unrealistic (as reflected in some 
stakeholders' comments in 2002 on the first version of the CAFE model). 
Conversely, the OMEGA model cannot be operated in a way that accounts 
for what the agencies consider to be very real product planning 
considerations.
    However, having also considered Mr. Rykowski's comments about the 
CAFE model's ``effective cost'' metric, and having conducted side-by-
side testing documented in the accompanying FRIA, the agencies are 
satisfied that an alternative ``cost per credit'' metric is also a 
reasonable metric to use for estimating how manufacturers might 
selected among available options to add specific fuel-saving 
technologies to specific vehicles.\158\ Therefore, NHTSA has revised 
the CAFE model accordingly, as discussed below in Section VI.A.4.
---------------------------------------------------------------------------

    \158\ As discussed in the FRIA, results vary with model inputs, 
among manufacturers, and across model years, but compared to the 
NPRM's ``effective cost'' metric, the ``cost per credit'' metric 
appears to more frequently produce less expensive solutions than 
more expensive solutions, at least when simulating compliance with 
CO2 standards. Differences are more mixed when simulating 
compliance with CAFE standards, and even when simulating compliance 
with CO2 standards, results simulating ``perfect'' trading of 
CO2 compliance credits are less intuitive when the ``cost 
per credit metric.'' Nevertheless, and while less expensive 
solutions are not necessarily ``optimal'' solutions (e.g., if 
gasoline costs $7 per gallon and electricity is free, expensive 
electrification could be optimal), the agencies consider it 
reasonable to apply the ``cost per credit'' metric for the analysis 
supporting today's rulemaking.
---------------------------------------------------------------------------

    Section VI.C.1 also addresses Mr. Rogers's comments on engine maps 
used as estimates to full vehicle simulation. In any event, because 
engine maps are inputs to full vehicle modeling and simulation, the 
relative merits of specific maps provide no basis to prefer one vehicle 
simulation modeling system over another. Similarly, Section VI.B.3 also 
addresses Mr. Duleep's comments preferring EPA's prior approach, using 
ALPHA, of effectively assuming that a manufacturer would incur no 
additional cost by reoptimizing every powertrain to extract the full 
fuel economy potential of even the smallest incremental changes to 
aerodynamic drag and tire rolling resistance. Mr. Duleep implies that 
Autonomie is flawed because the NPRM analysis did not apply Autonomie 
in a way that makes such assumptions. The agencies discuss powertrain 
sizing and calibration in Section VI.B.3, and note here that such 
assumptions are not inherent to

[[Page 24230]]

Autonomie; like engine maps, these are inputs to full vehicle 
simulation. Therefore, neither of these comments by Mr. Rogers and Mr. 
Duleep lead the agencies to find reason not to use Autonomie.
    None of this is to say that Autonomie and the CAFE model as 
developed and applied for the NPRM left no room for improvement. In the 
NPRM and RIA, the agencies discussed plans to continue work in a range 
of specific technical areas, and invited comment on all aspects of the 
analysis. As discussed below in Chapter VI, the agencies received 
extensive comment on the published model, inputs, and analysis, both in 
response to the NPRM and, for newly-introduced modeling capabilities 
(estimation of sales, scrappage, and employment effects), in response 
to additional peer review conducted in 2019. The agencies have 
carefully considered these comments, refined various specific technical 
aspects of the CAFE model (like the ``effective cost'' metric mentioned 
above), and have also updated inputs to both Autonomie and the CAFE 
model. Especially given these refinements and updates, as discussed 
throughout this rule, EPA maintains that for CO2 rulemaking 
analysis, Autonomie and the CAFE model have advantages that warrant 
relying on them rather than on EPA's ALPHA and OMEGA models. Some 
examples of such advantages include: A longer history of ongong 
development and application for rulemaking, including by EPA; 
documentation and model operation stakeholders have found to be 
comparatively clear and enabling of independent replication of agency 
analyses; a mechanism to explicitly reflect the fact that 
manufacturers' product decisions are likely to be informed by fuel 
prices; better integration of various model functions, enabling 
efficient sensitivity analysis; and an annual time step that makes it 
possible to conduct report results on both a calendar year and model 
year basis, to estimate accruing impacts on vehicle sales and 
scrappage, and to account for the fact that not every vehicle can be 
designed in every model year; and other advantages discussed throughout 
today's notice. Therefore, recognizing that models inform but do not 
make regulatory decisions, EPA has elected to rely solely on the 
Autonomie and CAFE models to produce today's analysis of regulatory 
alternatives for CO2 standards.
    The following sections provide a brief technical overview of the 
CAFE model, including changes NHTSA made to the model since 2012, and 
differences between the current analysis, the analysis for the 2016 
Draft TAR and for the 2017 Proposed Determination/2018 Final 
Determination, and the 2018 NPRM, before discussing inputs to the model 
and then diving more deeply into how the model works. For more 
information on the latter topic, see the CAFE model documentation, 
available in the docket for this rulemaking and on NHTSA's website.
1. What assumptions have changed since the 2012 final rule?
    Any analysis of regulatory actions that will be implemented several 
years in the future, and whose benefits and costs accrue over decades, 
requires a large number of assumptions. Over such time horizons, many, 
if not most, of the relevant assumptions in such an analysis are 
inevitably uncertain.\159\ The 2012 CAFE/CO2 rule considered 
regulatory alternatives for model years through MY 2025 (17 model years 
after the 2008 market information that formed the basis of the 
analysis) that accrued costs and benefits into the 2060s. Not only was 
the new vehicle market in 2025 unlikely to resemble the market in 2008, 
but so, too, were fuel prices. It is natural, then, that each 
successive CAFE/CO2 analysis should update assumptions to 
reflect better the current state of the world and the best current 
estimates of future conditions.\160\ However, beyond the issue of 
unreliable projections about the future, a number of agency assertions 
have proven similarly problematic. In fact, Securing America's Future 
Energy (SAFE) stated in their comments on the NPRM:
---------------------------------------------------------------------------

    \159\ As often stated, ``It's difficult to make predictions, 
especially about the future.'' See, e.g., https://quoteinvestigator.com/2013/10/20/no-predict/.
    \160\ See, e.g., 77 FR 62785 (Oct. 15, 2012) (``If EPA initiates 
a rulemaking [to revise standards for MYs 2022-2025], it will be a 
joint rulemaking with NHTSA. . . . NHTSA's development of its 
proposal in that later rulemaking will include the making of 
economic and technology analyses and estimates that are appropriate 
for those model years and based on then-current information.'').

    Although the agencies argue ``circumstances have changed'' and 
``analytical methods and inputs have been updated,'' a thorough 
analysis should provide a side-by-side comparison of the changing 
circumstances, methods, and inputs used to arrive at this 
determination . . . They represent a rapid, dramatic departure from 
the agencies' previous analyses, without time for careful review and 
consideration.\161\
---------------------------------------------------------------------------

    \161\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 39.

    We describe in detail (below) the changes to critical assumptions, 
perspectives, and modeling techniques that have created substantive 
differences between the current analysis and the analysis conducted in 
2012 to support the final rule. To the greatest extent possible, we 
have calculated the impacts of these changes on the 2012 analysis.
a) The Value of Fuel Savings
    The value of fuel savings associated with the preferred alternative 
in the 2012 final rule is primarily a consequence of two assumptions: 
\162\ The fuel price forecast and the assumed growth in fuel economy in 
the baseline alternative against which savings are measured. Therefore, 
as the value of fuel savings accounted for nearly 80 percent of the 
total benefits of the 2012 rule, each of these assumptions is 
consequential. With a lower fuel price projection and an expectation 
that new vehicle buyers respond to fuel prices, the 2012 rule would 
have shown much smaller fuel savings attributable to the more stringent 
standards. Projected fuel prices are considerably lower today than in 
2012, the agencies now understand new vehicle buyers to be at least 
somewhat responsive to fuel prices, and the agencies have therefore 
updated corresponding model inputs to produce an analysis the agencies 
consider to be more realistic.
---------------------------------------------------------------------------

    \162\ The value of fuel savings is also affected by the rebound 
effect assumption, assumed lifetime VMT accumulation, and the 
simulated penetration of alternative fuel technologies. However, 
each of these ancillary factors is small compared to the impact of 
the two factors discussed in this subsection.
---------------------------------------------------------------------------

    The first of these assumptions, fuel prices, was simply an artifact 
of the timing of the rule. Following recent periodic spikes in the 
national average gasoline price and continued volatility after the 
great recession, the fuel price forecast then produced by EIA (as part 
of AEO 2011) showed a steady march toward historically high, sustained 
gasoline prices in the United States. However, the actual series of 
fuel prices has skewed much lower. As it has turned out, the observed 
fuel price in the years between the 2012 final rule and this rule has 
frequently been lower than the ``Low Oil Price'' sensitivity case in 
the 2011 AEO, even when adjusted for inflation. The following graph 
compares fuel prices underlying the 2012 final rule to fuel prices 
applied in the analysis reported in today's notice, expressing both 
projections in 2010 dollars. The differences are clear and significant:

[[Page 24231]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.042

    The discrepancy in fuel prices is important to the discussion of 
differences between the current rule and the 2012 final rule, because 
that discrepancy leads in turn to differences in analytical outputs and 
thus to differences in what the agencies consider in assessing what 
levels of standards are reasonable, appropriate, and/or maximum 
feasible. As an example, the agencies discuss in Sections VI.D.3 
Simulating Environmental Impacts of Regulatory Alternatives and 
VIII.A.3 EPA's Conclusion that the Final CO2 Standards are 
Appropriate and Reasonable that fuel price projections from the 2012 
rule were one assumption, among others, that could have led to 
overestimates of the health benefits that resulted from reducing 
criteria pollutant emissions. Yet the agencies caution readers not to 
interpret this discrepancy as a reflection of negligence on the part of 
the agencies, or on the part of EIA. Long-term predictions are 
challenging and the fuel price projections in the 2012 rule were within 
the range of conventional wisdom at the time. However, it does suggest 
that fuel economy and tailpipe CO2 regulations set almost 
two decades into the future are vulnerable to surprises, in some ways, 
and reinforces the value of being able to adjust course when critical 
assumptions are proven inaccurate. This value was codified in 
regulation when EPA bound itself to the mid-term evaluation process as 
part of the 2012 final rule.\163\
---------------------------------------------------------------------------

    \163\ See 40 CFR 86-1818-12(h).
---------------------------------------------------------------------------

    To illustrate this point clearly, substituting the current (and 
observed) fuel price forecast for the forecast used in the 2012 final 
rule creates a significant difference in the value of fuel savings. 
Even under identical discounting methods (see Section 2, below), and 
otherwise identical inputs in the 2012 version of the CAFE Model, the 
current (and historical) fuel price forecast reduces the value of fuel 
savings by $150 billion--from $525 billion to $375 billion (in 2009 
dollars).
    The second assumption employed in the 2012 (as well as the 2010) 
final rule, that new vehicle fuel economy never improves unless 
manufacturers are required to increase fuel economy in the new vehicle 
market by increasingly stringent regulations, is more problematic. 
Despite the extensive set of recent academic studies showing, as 
discussed in Section VI.D.1.a)(2), that consumers value at least some 
portion, and in some studies nearly all, of the potential fuel savings 
from higher levels of fuel economy at the time they purchase vehicles, 
the agencies assumed in past rulemakings that buyers of new vehicles 
would never purchase, and manufacturers would never supply, vehicles 
with higher fuel economy than those in the baseline (MY 2016 in the 
2012 analysis), regardless of technology cost or prevailing fuel prices 
in future model years. In calendar year 2025, the 2012 final rule 
assumed gasoline would cost nearly $4.50/gallon in today's dollars, and 
continue to rise in subsequent years. Even recognizing that higher 
levels of fuel economy would be achieved under the augural/existing 
standards than without them, the assertion that fuel economy and 
CO2 emissions would not improve beyond 2016 levels in the 
presence of nearly $5/gallon gasoline is not supportable. This is 
highlighted by the observed increased consumer demand for higher-fuel-
economy vehicles during the gas price spike of 2008, when average U.S. 
prices briefly broke $4/gallon. In the 2012 final rule, this 
assumption--that fuel economy and emissions would never improve absent 
regulation--created a persistent gap in fuel economy between

[[Page 24232]]

the baseline and augural standards that grew to 13 mpg (at the industry 
average, across all vehicles) by MY 2025. In the 2016 Draft TAR, 
NHTSA's analysis included the assumption that manufacturers would 
deploy, and consumers would demand, any technology that recovered its 
own cost in the first year of ownership through avoided fuel costs. 
However, in both the Draft TAR and the Proposed and Final Determination 
documents, EPA's analysis assumed that the fuel economy levels achieved 
to reach compliance with MY 2021 standards would persist indefinitely, 
regardless of fuel prices or technology costs.
    By substituting the conservative assumption that consumers are 
willing to purchase fuel economy improvements that pay for themselves 
with avoided fuel expenditures over the first 2.5 years \164\ 
(identical to the assumption in this final rule's central analysis) the 
gap in industry average fuel economy between the baseline and augural 
scenarios narrows from 13 mpg in 2025 to 6 mpg in 2025. As a corollary, 
acknowledging that fuel economy would continue to improve in the 
baseline under the fuel price forecast used in the final rule erodes 
the value of fuel savings attributable to the preferred alternative. 
While each gallon is still worth as much as was assumed in 2012, fewer 
gallons are consumed in the baseline due to higher fuel economy levels 
in new vehicles. In particular, the number of gallons saved by the 
preferred alternative selected in 2012 drops from about 180 billion to 
50 billion once we acknowledge the existence of even a moderate market 
for fuel economy.\165\ The value of fuel savings is similarly eroded, 
as higher fuel prices lead to correspondingly higher demand for fuel 
economy even in the baseline--reducing the value of fuel savings from 
$525 billion to $190 billion.
---------------------------------------------------------------------------

    \164\ Greene, D.L. and Welch, J.G., ``Impacts of fuel economy 
improvements on the distribution of income in the U.S.,'' Energy 
Policy, Volume 122, November 2018, pp. 528-41 (``Four nationwide 
random sample surveys conducted between May 2004 and January 2013 
produced payback period estimates of approximately three years, 
consistent with the manufacturers' perceptions.'') (The 2018 article 
succeeds Greene and Welch's 2017 publication titled ``The Impact of 
Increased Fuel Economy for Light-Duty Vehicles on the Distribution 
of Income in the U.S.: A Retrospective and Prospective Analysis,'' 
Howard H. Baker Jr. Center for Public Policy, March 2017, which 
Consumers Union, CFA, and ACEEE comments include as Attachment 4, 
Docket NHTSA-2018-0067-11731).
    \165\ Readers should note that this is not an estimate of the 
total amount of fuel that will be consumed or not consumed by the 
fleet as a whole, but simply the amount of fuel that will be 
consumed or not consumed as a direct result of the regulation. As 
illustrated in Section VII, light-duty vehicles in the U.S. would 
continue to consume considerable quantities of fuel and emit 
considerable quantities of CO2 even under the baseline/
augural standards, and agencies' analysis shows that the standards 
finalized today will likely increase fuel consumption and 
CO2 emissions by a small amount.
---------------------------------------------------------------------------

    The magnitude of the fuel economy improvement in the baseline is a 
consequence of both the fuel prices assumed in the 2012 rule (already 
discussed as being higher than both subsequent observed prices and 
current projections) and the assumed technology costs. In 2012, a 
number of technologies were assumed to have negative incremental 
costs--meaning that applying those technologies to existing vehicles 
would both improve their fuel economy and reduce the cost to produce 
them. Asserting that the baseline would experience no improvement in 
fuel economy without regulation is equivalent to asserting that 
manufacturers, despite their status as profit maximizing entities, 
would not apply these cost-saving technologies unless forced to do so 
by regulation. While this issue is discussed in greater detail in 
Section VI.B the combination of inexpensive (or free) technology and 
high fuel prices created a logically inconsistent perspective in the 
2012 rule--where consumers never demanded additional fuel economy, 
despite high fuel costs, and manufacturers never supplied additional 
fuel economy, despite the availability of inexpensive (or cost saving) 
technology to do so.
    Many commenters on earlier rules supported the assumption that fuel 
economy would not improve at all in the absence of standards. In fact, 
some commenters still support this position. For example, EDF commented 
to the NPRM that, ``NHTSA set the Volpe model to project that, with 
CAFE standards remaining flat at MY 2020 levels through MY 2026, 
automakers would over-comply with the MY 2020 standards by 9 grams/mile 
of CO2 for cars and 15 g/mi of CO2 for light 
trucks during the 2029-2032 timeframe, plus 1%/year improvements beyond 
MY 2032. This assumption unreasonably obscures the impacts of the 
rollback and is not reflective of historical compliance performance.'' 
\166\
---------------------------------------------------------------------------

    \166\ EDF, NHTSA-2018-0067-11996, Comments to DEIS, at 4.
---------------------------------------------------------------------------

    EDF is mistaken in two different ways: (1) By acknowledging the 
existence of a well-documented market for fuel economy, rather than 
erroneously inflating the benefits associated with increasing 
standards, this assumption serves to isolate the benefits actually 
attributable to each regulatory alternative, and (2) it is, indeed, 
reflective of historical compliance performance. While the agencies 
rely on the academic literature (and comments from companies that build 
and sell automobiles) to defend the assertion that a market for fuel 
economy exists, the industry's historical CAFE compliance performance 
is a matter of public record.\167\ As shown in Figure IV-3, Figure IV-
4, and Figure IV-5 for more than a decade, the industry average CAFE 
has exceeded the standard for each regulatory class--by several mpg 
during periods of high fuel prices.
---------------------------------------------------------------------------

    \167\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed 10/08/2019.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24233]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.043


[[Page 24234]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.044

BILLING CODE 4910-59-C
    While this rulemaking has shown the impact of deviations from the 
2012 rule assumptions individually, these two assumptions affect the 
value of fuel savings jointly. Replacing the fuel price forecast with 
the observed historical and current projected prices, and including any 
technology that pays for itself in the first 2.5 years of ownership 
through avoided fuel expenditures, reduces the value of fuel savings 
from $525 billion in the 2012 rule to $140 billion, all else equal. 
Interestingly, this reduction in the value of fuel savings is smaller 
than the result when assuming only that the desired payback period is 
nonzero. While it may seem counterintuitive, it is entirely consistent.
    The number of gallons saved under the preferred alternative is 
actually higher when modifying both assumptions, compared to only 
modifying the payback period. Updating both assumptions leads to about 
100 billion gallons saved by the preferred alternative in 2012, 
compared to only 50 billion from changing only the payback period, and 
180 billion in the 2012 analysis. This occurs because the fuel economy 
in the baseline is lower when updating both the fuel price and the 
payback period--the gap between the augural standards and the baseline 
grows to 9 mpg, rather than only 6 mpg when updating only the payback 
period. Despite the existence of inexpensive

[[Page 24235]]

technology in both cases, with lower fuel prices there are fewer 
opportunities to apply technology that will pay back quickly. As a 
consequence, the number of gallons saved by the preferred alternative 
in 2012 increases--but each gallon saved is worth less because the 
price of fuel is lower.
b) Technology Cost
    While the methods used to identify cost-effective technologies to 
improve fuel economy in new vehicles have continuously evolved since 
2012 (as discussed further in Section IV.B.1), as have the estimated 
cost of individual technologies, the inclusion of a market response in 
all scenarios (including the baseline) has changed the total technology 
cost associated with a given alternative. As also discussed in Section 
VI.B, acknowledging the existence of a market for fuel economy leads to 
continued application of the most cost-effective technologies in the 
baseline--and in other less stringent alternatives--up to the point at 
which there are no remaining technologies whose cost is fully offset by 
the value of fuel saved in the first 30 months of ownership. The 
application of this market-driven technology has implications for fuel 
economy levels under lower stringencies (as discussed earlier), but 
also for the incremental technology cost associated with more stringent 
alternatives. As lower stringency alternatives (including the 2012 
baseline) accrue more technology, the incremental cost of more 
stringent alternatives decreases.
    By including a modest market for fuel economy, and preserving all 
other assumptions from the 2012 final rule, the incremental cost of 
technology attributable to the preferred alternative decreases from 
about $140 billion to about $72 billion. This significant reduction in 
technology cost is somewhat diminished by the associated reduction in 
the value of fuel savings (a decrease of $385 billion) when 
acknowledging the existence of a market for fuel economy. Another 
consequence of these changes is that the incremental cost of fuel 
economy technology is responsive to fuel price, as it should be. Under 
higher prices (as were assumed in 2012), consumers demand higher fuel 
economy in the new vehicle market. Under lower prices (as have occurred 
since the 2012 rule) consumers demand less fuel economy than would have 
been consistent with the fuel price assumptions in 2012.\168\ Including 
a market response in the analysis ensures that, in each case, the cost 
of fuel economy technology within an alternative is consistent with 
those assumptions. Using the same fuel price forecast that supports 
this rule, and the same estimate of market demand for fuel economy, the 
incremental cost of technology in the preferred alternative would rise 
back up to about $110 billion.
---------------------------------------------------------------------------

    \168\ This is why dozens of studies examining the ability of 
fuel taxes (and carbon taxes, which produce the same result for 
transportation fuels) to reduce CO2 emissions have found 
cost-effective opportunities available for those pricing mechanisms.
---------------------------------------------------------------------------

c) The Social Cost of Carbon (SCC) Emissions
    As discussed extensively in the NPRM, the agencies' perspective 
regarding the social cost of carbon has narrowed in focus. While the 
2012 final rule considered the net present value of global damages 
resulting from carbon emitted by vehicles sold in the U.S. between MY 
2009 and MY 2025, the NPRM (and this final rule) consider only those 
damages that occur to the United States and U.S. territories. As a 
result of this change in perspective, the value of estimated damages 
per-ton of carbon is correspondingly smaller. Had the 2012 final rule 
utilized the same perspective on the social cost of carbon, the 
benefits associated with the preferred alternative would have been 
about $11 billion, rather than $53 billion. However, the savings 
associated with carbon damages are a consequence of both the assumed 
cost per-ton of damages and the number of gallons of fuel saved. As 
discussed above, the gallons saved in the 2012 final rule were likely 
inflated as a result of both fuel price forecasts and the assumption 
that no market exists for fuel economy improvements. Correcting the 
estimate of gallons saved from the preferred alternative in the 2012 
rule and considering only the domestic social cost of carbon further 
reduces the savings in carbon damages to $6 billion.
d) Safety Neutrality
    In the 2012 final rule, the agencies showed a ``safety neutral'' 
compliance solution; that is, a compliance solution that produced no 
net increase in on-road fatalities for MYs 2017-2025 vehicles as a 
result of technology changes associated with the preferred alternative. 
In practice, safety neutrality was achieved by expressly limiting the 
availability of mass reduction technology to only those vehicles whose 
usage causes fewer fatalities with decreased mass. This result was 
discussed as one possible solution, where manufacturers chose 
technology solutions that limited the amount of mass reduction applied, 
and concentrated the application on vehicles that improve the safety of 
other vehicles on the roads (primarily by reducing the mass 
differential in collisions). However, it implicitly assumed that each 
and every manufacturer would leave cost-effective technologies unused 
on entire market segments of vehicles in order to preserve a safety 
neutral outcome at the fleet level for a given model year (or set of 
model years) whose useful lives stretched out as far as the 2060s. 
Removing these restrictions tells a different story.
    When mass reduction technology, determined in the model to be a 
cost-effective solution (particularly in later model years, when more 
advanced levels of mass reduction were expected to be possible), is 
unrestricted in its application, the 2012 version of the CAFE Model 
chooses to apply it to vehicles in all segments. This has a small 
effect on technology costs, increasing compliance costs in the earliest 
years of the program by a couple billion dollars, and reducing 
compliance costs for MYs 2022--2025 by a couple billion dollars. 
However, the impact on safety outcomes is more pronounced.
    Also starting with the model and inputs used for the 2012 final 
rule (and, as an example, focusing on that rule's 2008-based market 
forecast), removing the restrictions on the application of mass 
reduction technology results in an additional 3,400 fatalities over the 
full lives of MYs 2009-2025 vehicles in the baseline,\169\ and another 
6,900 fatalities over those same vehicle lives under the preferred 
alternative. The result, a net increase of 3,500 fatalities under the 
preferred alternative relative to the baseline, also produces a net 
social cost of $18 billion. The agencies' current treatment of both 
mass reduction technology, which can greatly improve the effectiveness 
of certain technology packages by reducing road load, and estimated 
fatalities and now account for both general exposure (omitted in the 
2012 final rule modeling) and fatality risk by age of the vehicle, 
further changes the story around mass reduction technology application 
for compliance and its relationship to on-road safety.
---------------------------------------------------------------------------

    \169\ Relative to the continuation of vehicle mass from the 2008 
model year carried forward into the future.
---------------------------------------------------------------------------

2. What methods have changed since the 2012 final rule?
    Simulating how manufacturers might respond to CAFE/CO2 
standards

[[Page 24236]]

requires information about existing products being offered for sale, as 
well as information about the costs and effectiveness of technologies 
that could be applied to those vehicles to bring the fleets in which 
they reside into compliance with a given set of standards. Following 
extensive additional work and consideration since the 2012 analysis, 
both agencies now use the CAFE Model to simulate these compliance 
decisions. This has several practical implications which are discussed 
in greater detail in Section VI.A. Briefly, this change represents a 
shift toward including a number of real-world production constraints--
such as component sharing across a manufacturer's portfolio--and 
product cadence, where only a subset of vehicles in a given model year 
are redesigned (and thus eligible to receive fuel economy technology). 
Furthermore, the year-by-year accounting ensures a continuous evolution 
of a manufacturer's product portfolio that begins with the market data 
of an initial model year (model year 2017, in this analysis) and 
continues through the last year for which compliance is simulated. 
Finally, the modeling approach has migrated from one that relied on the 
simple product of single values to estimate technology effectiveness to 
a model that relies on full vehicle simulation to determine the 
effectiveness of any combination of fuel economy technologies. The 
combination of these changes has greatly improved the realism of 
simulated vehicle fuel economy for combinations of technologies across 
vehicle systems and classes.
    In addition to these changes to the portions of the analysis that 
represent the supply of fuel economy (by manufacturer, fleet, and model 
year) in the new vehicle market, this analysis contains changes to the 
representation of consumer demand for fuel economy. One such measure 
was discussed above--the notion that consumers will demand some amount 
of fuel economy improvement over time, consistent with technology costs 
and fuel prices. However, another deviation from the 2012 final rule 
analysis reflects overall demand for new vehicles. Across ten 
alternatives, ranging from the baseline (freezing future standards at 
2016 levels) to scenarios that increased stringency by seven percent 
per year, from 2017 through 2025, the 2012 analysis showed no response 
in new vehicle sales, down to the individual model level. This implied 
that, regardless of changes to vehicle cost or attributes driven by 
stringency increases, no fewer (or possibly more) units of any single 
model would be sold in any year, in any alternative. Essentially, that 
analysis asserted that the new vehicle market does not respond, in any 
way, to average new vehicle prices across the alternatives--regardless 
of whether the incremental cost is $1,600/vehicle (as it was estimated 
to be under the preferred alternative) or nearly $4,000/vehicle (as it 
was in under the 7 percent alternative). Both the NPRM and this final 
rule, while not employing pricing models or full consumer choice models 
to address differentiated demand within brands or manufacturer 
portfolios, have incorporated a modeled sales response that seeks to 
quantify what was not quantified in previous rulemakings.
    An important accounting method has also changed since the 2012 
final rule was published. At the time of that rule, the agencies used 
an approach to discounting that combined attributes of a private 
perspective and a social perspective in their respective benefit cost 
analyses. This approach was logically inconsistent, and further 
reinforced some of the exaggerated estimates of fuel savings, social 
benefits (from reduced externalities), and technology costs described 
above. The old method discounted the value of all incremental 
quantities, whether categorized as benefits or costs, to the model year 
of the vehicle to which they accrued. This approach is largely 
acceptable for use in a private benefit cost analysis, where the costs 
and benefits accrue to the buyer of a new vehicle (in the case of this 
policy) who weighs their discounted present values at the time of 
purchase. However, the private perspective would not include any costs 
or benefits that are external to the buyer (e.g., congestion or the 
social cost of carbon emissions). For an analysis that compares 
benefits and costs from the social perspective, attempting to estimate 
the relative value of a policy to all of society rather than just 
buyers of new vehicles, this approach is more problematic.
    The discounting approach in the 2012 final rule was particularly 
distortionary for a few reasons. The fact that benefits and costs 
occurred over long time periods in the 2012 rule, and the standards 
isolated the most aggressive stringency increases in the latter years 
of the program, served to allow benefits that occurred in 2025 (for 
example) to enter the accounting without being discounted, provided 
that they accrued to the affected vehicles during their first year of 
ownership. In a setting where numerous inputs (e.g., fuel price and 
social cost of carbon) increase over time, benefits were able to grow 
faster than the discount rate in some cases--essentially making them 
infinite. The interpretation of discounting for externalities was 
equally problematic. For example, the discounting approach in the 2012 
final rule would have counted a ton of CO2 not emitted in CY 
2025 in multiple ways, despite the fact that the social cost of carbon 
emissions was inherently tied to the calendar year in which the 
emissions occurred. Were the ton avoided by a MY 2020 vehicle, which 
would have been five years old in CY 2025, the value of that ton would 
have been the social cost of carbon times 0.86, but would have been 
undiscounted if that same ton had been saved by a MY 2025 vehicle in 
its initial year of usage.
    This approach was initially updated in the 2016 Draft TAR to be 
consistent with common economic practice for benefit-cost analysis, and 
this analysis continues that approach. In the social perspective, all 
benefits and costs are discounted back to the decision year based on 
the calendar year in which they occur. Had the agencies utilized such 
an approach in the 2012 final rule, net benefits would have been 
reduced by about 20 percent, from $465 billion to $374 billion--not 
accounting for any of the other adjustments discussed above.
3. How have conditions changed since the 2012 final rule was published?
    The 2012 final rule relied on market and compliance information 
from model year 2008 to establish standards for model years 2017-2025. 
However, in the intervening years, both the market and the industry's 
compliance positions have evolved. The industry has undergone a 
significant degree of change since the MY 2008 fleet on which the 
2012FR was based. Entire brands (Pontiac, Oldsmobile, Saturn, Hummer, 
Mercury, etc.) and companies (Saab, Suzuki, Lotus) have exited the U.S. 
market, while others (most notably Tesla) have emerged. Several dozen 
nameplates have been retired and dozens of other created in that time. 
Overall, the industry has offered a diverse set of vehicle models that 
have generally higher fuel economy than the prior generation, and an 
ever-increasing set of alternative fuel powertrains.
    As Table IV-1 shows, alternative powertrains have steadily 
increased under CAFE/CO2 regulations. Under the standards 
between 2011 and 2018, the number of electric vehicle offerings in the 
market has increased from 1 model to 57 models (inclusive of all plug-
in vehicles that are rated for use on the highway), and hybrids (like 
the Toyota Prius) have increased from 20 models to

[[Page 24237]]

43 models based on data from DOE's Alternative Fuels Data Center. Fuel 
efficient diesel vehicles have similarly been on the rise in that 
period, more than doubling the number of offerings. Flexible fuel 
vehicles (FFVs), capable of operating on both gasoline and E85 remain 
readily available in the market, but have been excluded from the table 
due to both their lower fuel economy and demonstrated consumer 
reluctance to operate FFVs on E85. They have historically been used to 
improve a manufacturer's compliance position, rather than other 
alternative fuel systems that reduce fuel consumption and save buyers 
money.
[GRAPHIC] [TIFF OMITTED] TR30AP20.045

    Not only have alternative powertrain options proliferated since the 
2012 FR, the average fuel economy of new vehicles within each body 
style has increased. However, the more dramatic effect may lie in the 
range of fuel economies available within each body style. Figure IV-6 
shows the distribution of new vehicle fuel economy (in miles per gallon 
equivalent) by body style for MYs 2008, 2016, and 2020 (simulated). 
Each box represents the 25th and 75th percentiles, where 25 and 75 
percent of new models offered are less fuel efficient than that level. 
Not only has the median fuel economy improved (the median shows the 
point at which 50 percent of new models are less efficient) under the 
CAFE/CO2 programs, but the range of available fuel economies 
(determined by the length of the boxes and their whiskers) has 
increased as well. For example, the 25th percentile of pickup truck 
fuel economy in 2020 is expected to be significantly more efficient 
than 75 percent of the pickups offered in 2008. In MY 2008, there were 
only a few SUVs offered with rated fuel economies above 34MPG. By MY 
2020 almost half of the SUVs offered will have higher fuel economy 
ratings--with almost 20 percent of offerings exceeding 40MPG.
    The improvement in passenger car styles has been no less dramatic. 
As with the other styles, the range of available fuel economies has 
increased under the CAFE/CO2 programs and the distribution 
of available fuel economies skewed higher--with 40 percent of MY 2020 
models exceeding 40MPG. The attribute-based standards are designed to 
encourage manufacturers to improve vehicle fuel economy across their 
portfolios, and they have clearly done so. Not only have the higher 
ends of the distributions increased, the lower ends in all body styles 
have improved as well, where the least fuel efficient 25 percent of 
vehicles offered in MY 2016 (and simulated in 2020) are more fuel 
efficient than the most efficient 25 percent of vehicles offered in MY 
2008.
BILLING CODE 4910-59-P

[[Page 24238]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.046

BILLING CODE 4910-59-C
    Some commenters have argued that consumers will be harmed by any 
set of standards lower than the baseline (augural) standards because 
buyers of new vehicles will be forced to spend more on fuel than they 
would have under the augural standards. However, as Figure IV-6 shows, 
the range of fuel economies available in the new market is already 
sufficient to suit the needs of buyers who desire greater fuel economy 
rather than interior volume or some other attributes. Full size pickup 
trucks are now available with smaller turbocharged engines paired with 
8 and 10-speed transmissions and some mild electrification. Buyers 
looking to transport a large family can choose to purchase a plug-in 
hybrid minivan. There were 57 electric models available in 2018, and 
hybrid powertrains are no longer limited to compact cars (as they once 
were). Buyers can choose hybrid SUVs with all-wheel and four-wheel 
drive. While these kinds of highly efficient options were largely 
absent from some body styles in MY 2008, this is no longer the case. 
Given that high-MPG vehicles are widely available, consumers must also 
value other vehicle attributes (e.g., acceleration and load-carrying 
capacity) that can can also be improved with the same technologies that 
can be used to improve fuel economy.
---------------------------------------------------------------------------

    \170\ Circles represent specific outlying vehicle models.
---------------------------------------------------------------------------

    Manufacturers have accomplished a portfolio-wide improvement by 
improving the combustion efficiency of engines (through direct 
injection and

[[Page 24239]]

turbocharging), migrating from four and five speed transmissions to 8 
and 10 speed transmissions, and electrifying to varying degrees. All of 
this has increased both production costs and fuel efficiency during a 
period of economic expansion and low energy prices. While the vehicles 
offered for sale have increased significantly in efficiency between MY 
2008 and MY 2020, the sales-weighted average fuel economy has achieved 
less improvement. Despite stringency increases of about five percent 
(year-over-year) between 2012 and 2016, the sales-weighted average fuel 
economy increased marginally. Figure IV-7 shows an initial increase in 
average new vehicle fuel economy (the heavy solid line, shown in mpg as 
indicated on the left y axis), followed by relative stagnation as fuel 
prices (the light dashed lines, shown in dollars per gallon as 
indicated on the right y axis) fell and remained low.\171\ It is worth 
noting that average new vehicle fuel economy observed a brief spike 
during the year that the Tesla Model 3 was introduced (as a consequence 
of strong initial sales volumes, as pre-orders were satisfied, and fuel 
economy ratings that are significantly higher than the industry 
average), and settled around 27.5 MPG in Fall 2019. Average fuel 
economy receded further over the next several months to 26.6 MPG in 
February 2020.\172\
---------------------------------------------------------------------------

    \171\ Ward's Automotive, https://www.wardsauto.com/industry/fuel-economy-index-shows-slow-improvement-april. Last accessed Dec. 
13, 2019.
    \172\ Ward's Automotive, https://wardsintelligence.informa.com/WI964622/Fuel-Economy-Slightly-Down-in-February. Last accessed Mar. 
9, 2020.
[GRAPHIC] [TIFF OMITTED] TR30AP20.047

    In their NPRM comments, manufacturers expressed concern that CAFE 
standards had already increased to the point where the price increases 
necessary to recoup manufacturers' increased costs for providing 
further increases in fuel economy outweigh the value of fuel 
savings.173 174 The agencies do not agree that this point 
has already been reached by previous stringency increases, but 
acknowledge the reality of diminishing marginal returns to improvements 
in fuel economy. A driver with a 40MPG vehicle uses about 300 gallons 
of fuel per year. Increasing the fuel economy of that vehicle to 50MPG, 
a 25 percent increase, would likely be over $1000 in additional 
technology cost. However, that driver would only save 25 percent of 
their annual fuel consumption, or 75 gallons out of 300 gallons. Even 
at $3/gallon, higher than the current national average, that represents 
$225 per year in fuel savings. That means that the buyer's $1000 
investment in additional fuel economy pays back in just under 4.5 years 
(undiscounted). The agencies' respective programs have created greater 
access to high MPG vehicles in all classes and encouraged the 
proliferation of alternative fuels and powertrains. But if the value of 
the fuel savings is insufficient to motivate buyers to invest in ever 
greater levels of fuel economy, manufacturers will face challenges in 
the market.
---------------------------------------------------------------------------

    \173\ NHTSA-2018-0067-12064-25.
    \174\ NHTSA-2018-0067-12073-2.
---------------------------------------------------------------------------

    While Figure IV-3 through Figure IV-5 illustrate the trends in 
historical CAFE compliance for the entire industry, the figures contain 
another relevant fact. After several consecutive years of increasing 
standards, the achieved and required levels converge. When the 
standards began increasing again for passenger cars in 2011, the prior 
year had industry CAFE levels 5.6 mpg and 7.7 mpg in excess of their 
standards for domestic cars and imported cars, respectively. Yet, by 
2016, the consecutive year-over-year increases had eroded the levels of 
over-compliance. Light trucks similarly exceeded their standard prior 
to increasing standards, which began in 2005. Yet, by 2011, after 
several consecutive years of stringency increases, the industry light-
truck average CAFE was merely compliant with the rising standard.
    This is largely due to the fact that stringency requirements have 
increased at a faster rate than achieved fuel

[[Page 24240]]

economy levels for several years. The attribute-based standards took 
effect in 2011 for all regulatory classes, although light truck CAFE 
standards had been increasing since 2005. Since 2004, light truck 
stringency has increased an average of 2.7 percent per year, while 
light truck's compliance fuel economy has increased by an average of 
1.7 percent over the same period.\175\ For the passenger classes, a 
similar story unfolds over a shorter period of time. Year over year 
stringency increases have averaged 4.7 percent per year for domestic 
cars (though increases in the first two years were about 8 percent--
with lower subsequent increases), but achieved fuel economy increases 
averaged only 2.2 percent per year over the same period. Imported 
passenger cars were similar to domestic cars, with average annual 
stringency increases of 4.4 percent but achieved fuel economy levels 
increasing an average of only 1.4 percent per year from 2011 through 
2017. Given that each successive percent increase in stringency is 
harder to achieve than the previous one, long-term discrepancies 
between required and achieved year-over-year increases cannot be offset 
indefinitely with existing credit banks, as they have been so far.
---------------------------------------------------------------------------

    \175\ Both the standards and these calculations are defined in 
consumption space--gallons per mile--which also translates directly 
into CO2 based on the carbon content of the fuel 
consumed.
---------------------------------------------------------------------------

    With the fuel price increases fresh in the minds of consumers, and 
the great recession only recently passed, the CAFE stringency increases 
that began in 2011 (and subsequent CAFE/CO2 stringency 
increases after EPA's program was first enforced in MY 2012) had 
something of a head start. As Figure IV-3 through Figure IV-5 
illustrate, the standards were not binding in MY 2011--even 
manufacturers that had historically paid civil penalties were earning 
credits for overcompliance. It took two years of stringency increase to 
catch up to the CAFE levels already present in MY 2011. However, seven 
consecutive years of increases for passenger cars and a decade of 
increases for light trucks has changed the credit situation. Figure IV-
8 shows CAFE credit performance for regulated fleets--the solid line 
represents the number of fleets generating shortfalls and the dashed 
line represents the number of fleets earning credits in each model 
year.
[GRAPHIC] [TIFF OMITTED] TR30AP20.048

    Fewer than half as many fleets earned surplus credits for over-
compliance in MY 2017 compared to MY 2011--and this trend is 
persistent. The story varies from one manufacturer to another, but it 
seems sufficient to state the obvious--when the agencies conducted the 
analysis to establish standards through MY 2025 back in 2012, most (if 
not all) manufacturers had healthy credit positions. That is no longer 
the case, and each successive increase requires many fleets to not only 
achieve the new level from the resulting increase, but to resolve 
deficits from the prior year as well. The large sums of credits, which 
last five years under both programs, have allowed most manufacturers to 
resolve shortfalls. But the light truck fleet, in particular, has a 
dwindling supply of credits available for purchase or trade. The 
CO2 program has a provision that allows credits earned 
during the early years of over-compliance to be applied through MY 
2021. This has reduced the compliance burden in the last several years, 
as intended, but will not mitigate the compliance challenges some OEMs 
would face if the baseline standards remained in place and energy 
prices persisted at current levels.
BILLING CODE 4910-59-P

[[Page 24241]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.049

BILLING CODE 4910-59-C
    Table IV-2 shows the credits earned by each manufacturer over 
time.\176\ As the table shows, when the agencies considered future 
standards in 2012, most manufacturers were earning credits in at least 
one fleet. However, the bold values show years with deficits and even 
some manufacturers who started out in strong positions, such as Ford's 
passenger car fleet, have seen growing deficits in recent years. While

[[Page 24242]]

the initial banks for early-action years eases the burden of 
CO2 compliance for many OEMs, the year-to-year compliance 
story is similar to CAFE, see Table IV-3.
---------------------------------------------------------------------------

    \176\ MY 2017 values represent estimated earned credits based on 
MY 2017 final compliance data.
[GRAPHIC] [TIFF OMITTED] TR30AP20.050

    Credit position and shortfall rates clearly illustrate 
manufacturers' fleet performance relative to the standards. Recognizing 
that manufacturers plan compliance over several model years at any 
given time, sporadic shortfalls may not be evidence of undue 
difficulty, but sustained, widespread, growing shortfalls should 
probably be viewed as evidence that standards previously believed to be 
manageable might no longer be so. While NHTSA is prohibited by statute 
from considering availability of credits (and thus, size of credit 
banks) in determining maximum feasible standards, it does consider 
shortfalls as part of its determination. EPA has no such prohibition 
under the CAA and is free to consider both credits and shortfalls.
    These increasing credit shortfalls are occurring at a time that the 
industry is deploying more technology than the agencies anticipated 
when establishing future standards in 2012. The agencies' projections 
of transmission technologies were mixed. While the agencies expected 
the deployment of 8-speed transmissions to about 25 percent of the 
market by MY 2018, transmissions with eight or more gears account for 
almost 30 percent of the market. However, the agencies projected no CVT 
transmissions in future model years, instead projecting high 
penetration of DCTs. However, CVTs currently make up more than 20 
percent of new transmissions. The tradeoff between advanced engines and 
electrification was also underestimated. While the agencies projected 
penetration rates of turbocharged engines that are higher than we've 
observed in the market (45 percent compared to 30 percent), the 
estimated penetration of electric technologies was significantly lower. 
The agencies projected a couple percent of strong hybrids--which we've 
seen--but virtually no PHEVs or EVs. While the volumes of those 
vehicles are still only a couple percent of the new vehicle market, 
they are heavily credited under both programs and can significantly 
improve compliance positions even at smaller volumes. Even lower-level 
electrification technologies, like stop-start systems, are 
significantly more prevalent than we anticipated (stop-start systems 
were projected to be in about 2 percent of the market, compared to over 
20 percent in the 2018 fleet). Despite technology deployment that is 
comparable to 2012 projections, and occasionally more aggressive, 
passenger car and light truck fleets have slightly lower fuel economy 
than projected. As fleet volumes have shifted along the footprint 
curve, the standards have decreased as well (relative to the 
expectation in 2012), but less so. While compliance deficits have been 
modest, they have been accompanied by record sales for several years. 
This has not only depleted existing credit banks, but created 
significant shortfalls that may be more difficult to overcome if sales 
recede from record levels.

V. Regulatory Alternatives Considered

    Agencies typically consider regulatory alternatives in proposals as 
a way of evaluating the comparative effects of different potential ways 
of accomplishing their desired goal. NEPA

[[Page 24243]]

requires agencies (in this case, NHTSA, but not EPA) to compare the 
potential environmental impacts of their proposed actions to those of a 
reasonable range of alternatives. Executive Orders 12866 and 13563 and 
OMB Circular A-4 also encourage agencies to evaluate regulatory 
alternatives in their rulemaking analyses. Alternatives analysis begins 
with a ``no-action'' alternative, typically described as what would 
occur in the absence of any regulatory action. This final rule, like 
the proposal, includes a no-action alternative, described below, as 
well as seven ``action alternatives.'' The final standards may, in 
places, be referred to as the ``preferred alternative,'' which is NEPA 
parlance, but NHTSA and EPA intend ``final standards'' and ``preferred 
alternative'' to be used interchangeably for purposes of this 
rulemaking.
    In the proposal, NHTSA and EPA defined the different regulatory 
alternatives (other than the no-action alternative) in terms of 
percent-increases in CAFE and CO2 stringency from year to 
year. Percent increases in stringency referred to changes in the 
standards year over year--as in, standards that become 1 percent more 
stringent each year. Readers should recognize that those year-over-year 
changes in stringency are not measured in terms of mile per gallon or 
CO2 gram per mile differences (as in, 1 percent more 
stringent than 30 miles per gallon in one year equals 30.3 miles per 
gallon in the following year), but in terms of shifts in the footprint 
functions that form the basis for the actual CAFE and CO2 
standards (as in, on a gallon or gram per mile basis, the CAFE and 
CO2 standards change by a given percentage from one model 
year to the next). Under some alternatives, the rate of change was the 
same for both passenger cars and light trucks; under others, the rate 
of change differed. Like the no-action alternative, all of the 
alternatives considered in the proposal were more stringent than the 
preferred alternative.
    Alternatives considered in the proposal also varied in other 
significant ways. Alternatives 3 and 7 in the proposal involved a 
gradual discontinuation of CAFE and average CO2 adjustments 
reflecting the use of technologies that improve air conditioner 
efficiency or otherwise improve fuel economy under conditions not 
represented by long-standing fuel economy test procedures (off-cycle 
adjustments, described in further detail in Section IX, although the 
proposal itself would have retained these flexibilities. Commenters 
responding to the request for comment on phasing out these 
flexibilities generally supported maintaining the existing program, as 
proposed. Some commenters suggested changes to the existing program 
that were not discussed in the NPRM. Such changes would be beyond the 
scope of this rulemaking and were not considered. Section IX contains a 
more thorough summary of these comments and the issues they raise, as 
well as the agencies' responses. Consistent with the decision to retain 
these flexibilities in the final rule, alternatives reflecting their 
phase-out have not been considered in this final rule.
    Additionally, in the NPRM for this rule, EPA proposed to exclude 
the option for manufacturers partially to comply with tailpipe 
CO2 standards by generating CO2-equivalent 
emission adjustments associated with air conditioning refrigerants and 
leakage after MY 2020. This approach was proposed in the interest of 
improved harmonization between the CAFE and tailpipe CO2 
emissions programs because this optional flexibility cannot be 
available in the CAFE program.\177\ Alternatives 1 through 8 excluded 
this option. EPA requested comment ``on whether to proceed with [the] 
proposal to discontinue accounting for A/C leakage, methane emissions, 
and nitrous oxide emissions as part of the CO2 emissions 
standards to provide for better harmony with the CAFE program, or 
whether to continue to consider these factors toward compliance and 
retain that as a feature that differs between the programs.'' \178\ EPA 
stated that if EPA were to proceed with excluding A/C refrigerant 
credits as proposed, ``EPA would consider whether it is appropriate to 
initiate a new rulemaking to regulate these programs independently . . 
. .'' \179\ EPA also stated that ``[i]f the agency decides to retain 
the A/C leakage . . . provisions for CO2 compliance, it 
would likely re-insert the current A/C leakage offset and increase the 
stringency levels for CO2 compliance by the offset amounts 
described above (i.e., 13.8 g/mi equivalent for passenger cars and 17.2 
g/mi equivalent for light trucks). EPA received comments from a wide 
range of stakeholders, most of whom opposed the elimination of these 
flexibility provisions.
---------------------------------------------------------------------------

    \177\ For the CAFE program, carbon-based tailpipe emissions 
(including CO2, HC, and CO) are measured, and fuel 
economy is calculated using a carbon balance equation. EPA also uses 
carbon-based emissions (CO2, HC, and CO, the same as for 
CAFE) to calculate tailpipe CO2 for use in determining 
compliance with its standards. In addition, under the no-action 
alternative for the proposal and under all alternatives in the final 
rule, in determining compliance, EPA includes on a CO2 
equivalent basis (using Global Warming Potential (GWP) adjustment) 
A/C refrigerant leakage credits, at the manufacturer's option, and 
nitrous oxide and methane emissions. EPA also has separate emissions 
standards for methane and nitrous oxides. The CAFE program does not 
include or account for A/C refrigerant leakage, nitrous oxide and 
methane emissions because they do not impact fuel economy. Under 
Alternatives 1-8 in the proposal, the standards were closely aligned 
for gasoline powered vehicles because compliance with the fleet 
average standard for such vehicles is based on tailpipe 
CO2, HC, and CO for both programs and not emissions 
unrelated to fuel economy, although diesel and alternative fuel 
vehicles would have continued to be treated differently between the 
CAFE and CO2 programs. While such an approach would have 
significantly improved harmonization between the programs, standards 
would not have been fully aligned because of the small fraction of 
the fleet that uses diesel and alternative fuels (as described in 
the proposal, such vehicles made up approximately four percent of 
the MY 2016 fleet), as well as differences involving EPCA/EISA 
provisions EPA has not adopted, such as minimum standards for 
domestic passenger cars and limits on credit transfers between 
regulated fleets. The proposal to eliminate flexibilities associated 
with A/C refrigerants and leakage was not adopted for this final 
rule, and the reasons for and implications of that decision are 
discussed further below.
    \178\ 83 FR at 43193 (Aug. 24, 2018).
    \179\ Id. at 43194.
---------------------------------------------------------------------------

    Specifically, the two major trade organizations of auto 
manufacturers, as well as some individual automakers, supported 
retaining these provisions. Global Automakers commented that ``[a]ir 
conditioning refrigerant leakage . . . should be included for 
compliance with the EPA standards for all MYs, even if it means a 
divergence from the NHTSA standards.'' \180\ Global provides several 
detailed reasons for their comments, including that the existing 
provisions are ``. . . important to maintaining regulatory flexibility 
through real [CO2] emission reductions and would prevent the 
potential for additional bifurcated, separate programs at the state 
level.'' \181\ The Alliance similarly commented that it ``supports 
continuation of the full air conditioning refrigerant leakage credits 
under the [CO2] standards.'' \182\ Some individual

[[Page 24244]]

manufacturers, including General Motors,\183\ Fiat Chrysler,\184\ and 
BMW,\185\ also commented in support of maintaining the current A/C 
refrigerant and leakage credits.
---------------------------------------------------------------------------

    \180\ Global, NHTSA-2018-0067-12032, Appendix A at A-5.
    \181\ Id. Global also stated that excluding A/C leakage credits 
would ``. . . greatly limit the ability [of manufacturers] to select 
the most cost-effective approach for technology improvements and 
result in a costlier, separate set of regulations that actually 
relate to the overall GHG standards.'' Global also expressed concern 
that issuing separate regulations for A/C leakage could take too 
long and create a gap in which States might act to separately 
regulate or even ban refrigerants, and supported continued inclusion 
of A/C leakage and refrigerant regulation in EPA's GHG program to 
avoid risk of an ensuing patchwork. Global argued that manufacturers 
had already invested to meet the existing program, and that ``the 
proposed phase-out also creates another risk that manufacturers will 
have stranded capital in technologies that are not fully 
amortized.'' Global Automakers, EPA-HQ-OAR-2018-0283-5704, 
Attachment A, at A.43-44.
    \182\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 12. 
Alliance also expressed concern about stranded capital and risk of 
patchwork of state regulation if MAC direct credits were not 
retained in the Federal GHG program. Id. at 80-81.
    \183\ General Motors, NHTSA-2018-0067-11858, Appendix 4, at 1 
(``General Motors supports the extensive comments from the Alliance 
of Automobile Manufacturers regarding flexibility mechanisms, and 
incorporates them by reference. In particular, the Alliance cites 
the widening gap between the regulatory standards and actual 
industry-wide new vehicle average fuel economy that has become 
evident since 2016, despite the growing use of improvement `credits' 
from various flexibility mechanisms, such as off-cycle technology 
credits, mobile air conditioner efficiency credits, mobile air 
conditioner refrigerant leak reduction credits and credits from 
electrified vehicles.'')
    \184\ FCA, NHTSA-2018-0067-11943, at 8. FCA also expressed 
concern about patchwork in the absence of a federal rule. Id.
    \185\ BMW, EPA-HQ-OAR-2018-4204, at 3. BMW stated that ``Today's 
rules allow flexibilities to be used by the motor vehicle 
manufacturers for fuel saving technologies and efficiency gains 
which are not covered in the applicable test procedures. To enhance 
the future use of these technologies and to reward motor vehicle 
manufacturer's investments taken for future innovations, the 
agencies should consider the continuation of current flexibilities 
for the model years 2021 to 2026.''
---------------------------------------------------------------------------

    Auto manufacturing suppliers who addressed A/C refrigerant and 
leakage credits also generally supported retaining the existing 
provisions. MEMA commented that ``It is essential for supplier 
investment and jobs, and continuous innovation and improvements in the 
technologies that the credit programs continue and expand to broaden 
the compliance pathways. MEMA urges the agencies to continue the 
current credit and incentives programs . . . . '' \186\ DENSO also 
supported maintaining the current provisions.\187\ However, BorgWarner 
supported the proposed removal of A/C refrigerant credits ``for 
harmonization reasons,'' while encouraging EPA to regulate A/C 
refrigerants and leakage separately from the CO2 
standards.\188\
---------------------------------------------------------------------------

    \186\ MEMA, available at https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf, comment at p. 2. MEMA also expressed 
concern about stranded capital investments by suppliers and supplier 
jobs if the direct MAC credits were not available; stated that the 
credits were an important compliance flexibility and ``one of the 
highest values of any credit offered in the EPA program;'' and 
stated that ``Harmonizing the programs does not require making them 
identical or equivalent. Rather, harmonization can be achieved by 
better coordinating the two programs to the extent feasible while 
allowing each agency to implement its separate and distinct 
mandate.'' Id. at 15-16.
    \187\ DENSO, NHTSA-2018-0067-11880, at 8.
    \188\ BorgWarner, NHTSA-2018-0067-11895, at 10.
---------------------------------------------------------------------------

    The two producers of a lower GWP refrigerant, Chemours and 
Honeywell, commented extensively in support of continuing to allow A/C 
refrigerant and leakage credits for CO2 compliance, making 
both economic and legal arguments. Both Chemours and Honeywell stated 
that A/C refrigerant and leakage credits were a highly cost-effective 
way for OEMs to comply with the CO2 standards,\189\ with 
Chemours suggesting that OEM compliance strategies are based on the 
assumption that these credits will be available for CO2 
compliance \190\ and that any increase in stringency above the proposal 
effectively necessitates that the credits remain part of the 
program.\191\ Honeywell stated that all OEMs (and a variety of other 
parties) supported retaining the credits for CO2 
compliance,\192\ and Chemours, Honeywell, and CBD et al. all noted that 
OEMs are already using the credits for low GWP refrigerants in more 
than 50 percent of the MY 2018 vehicles produced for sale in the 
U.S.\193\ The American Chemistry Council also stated that the ``auto 
industry widely supports the credits, and U.S. chemical manufacturers 
are at a loss as to why EPA would propose to eliminate such a 
successful flexible compliance program.'' \194\ In response to NPRM 
statements expressing concern that the A/C refrigerant and leakage 
credits could be market distorting, both Chemours and Honeywell 
disagreed,\195\ arguing that the credits were simply a highly cost-
effective means of complying with the CO2 standards,\196\ 
and that removal of the credits at this point would, itself, distort 
the market for refrigerants. Honeywell argued that eliminating the A/C 
credit program from CO2 compliance would put the U.S. at a 
competitive disadvantage with other countries, and would risk U.S. 
jobs.\197\
---------------------------------------------------------------------------

    \189\ Chemours at 1 (``MVAC credits many times offer the `least 
cost' approach to compliance . . .'') and 9; Honeywell at 6.
    \190\ Chemours at 6-7; both Chemours and Honeywell expressed 
concern about OEM reliance on the expectation that HFC credits would 
continue to be part of the CO2 program (Chemours at 31; 
Honeywell at 16-20) and that investments in alternative refrigerants 
would be stranded (Chemours at 1, 3, 4-6; Honeywell at 2, 7-8).
    \191\ Chemours at 7.
    \192\ Honeywell at 8-11.
    \193\ Chemours at 4; Honeywell at 6-7; CBD et al. at 46-47.
    \194\ American Chemistry Council, EPA-HQ-OAR-2018-0283-1415, at 
9-10 (comments similar to Chemours and Honeywell).
    \195\ Chemours at 1; Honeywell at 13.
    \196\ Chemours at 29-30; Honeywell at 13-14.
    \197\ Honeywell at 20-21.
---------------------------------------------------------------------------

    Regarding the NPRM's statements that removing the A/C refrigerant 
and leakage credits from CO2 compliance would promote 
harmonization with the CAFE program, these commenters argued that 
harmonization was not a valid basis for that aspect of the proposal. 
Chemours, Honeywell, and CBD et al. all argued that Section 202(a) 
creates no obligation to harmonize the [CO2] program with 
the CAFE program.\198\ Chemours further argued that to the extent 
disharmonization between the programs existed, it should be addressed 
via stringency changes (i.e., reducing CAFE stringency relative to 
CO2 stringency) rather than ``dropping low-cost compliance 
options.'' \199\
---------------------------------------------------------------------------

    \198\ Chemours at 23-24; Honeywell at 11-12; CBD et al. at 47.
    \199\ Chemours at 9-11.
---------------------------------------------------------------------------

    These commenters also expressed concern that the proposal 
constituted an EPA decision not to regulate HFC emissions from motor 
vehicles at all. Commenters argued that the NPRM provided no legal 
analysis or reasoned explanation for stopping regulation of HFCs,\200\ 
and that Massachusetts v. EPA requires any final rule to regulate all 
greenhouse gases from motor vehicles and not CO2 alone,\201\ 
suggesting that there was a high likelihood of a lapse in regulation 
because EPA had not yet proposed a new way of regulating HFC 
emissions.\202\ Because the NPRM provided no specific information about 
how EPA might regulate non-CO2 emissions separately, 
commenters argued that the lack of clarity was inherently disruptive to 
OEMs.\203\ CBD et al. argued that any lapse in regulation is ``illegal 
on its face'' and that even creating a separate standard for HFC 
emissions would be ``illegal'' because it ``would increase costs to 
manufacturers and result in environmental detriment by removing any 
incentive to use the most aggressive approaches to curtail emissions of 
these highly potent GHGs.'' \204\
---------------------------------------------------------------------------

    \200\ Chemours at 1-2; Honeywell at 11.
    \201\ Chemours at 18-19; Honeywell at 14-16.
    \202\ Chemours at 6; Honeywell at 16.
    \203\ Chemours at 21; Honeywell at 16; ICCT at I-39.
    \204\ CBD et al. at 46.
---------------------------------------------------------------------------

    Environmental organizations,\205\ other NGOs, academic 
institutions, consumer organizations, and state governments \206\ also 
commented in support of continuing the existing provisions.
---------------------------------------------------------------------------

    \205\ ICCT, NHTSA-2018-0067-11741, Full Comments, at 4 
(describing ``air conditioning GHG-reduction technologies [as] 
available, cost-effective, and experiencing increased deployment by 
many companies due to the standards.''); CBD et al., Appendix A, at 
45-47.
    \206\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 120-
121; Washington State Department of Ecology, NHTSA-2018-0067-11926, 
at 6 (HFCs are an important GHG; compliance flexibility is 
important).
---------------------------------------------------------------------------

    EPA has considered its proposed approach to A/C refrigerant and 
leakage

[[Page 24245]]

credits in light of these comments. EPA believes that maintaining this 
element of its program is consistent with EPA's authority under Section 
202(a) to establish standards for reducing emissions from LDVs. Thus, 
maintaining the optional HFC credit program is appropriate. In 
addition, EPA recognizes the value of regulatory flexibility and 
compliance options, and has concluded that the advantages from 
retaining the existing A/C refrigerant/leakage credit program and 
associated offset between the CO2 and CAFE standards--in 
terms of providing for a more-comprehensive regulation of emissions 
from light-duty vehicles--outweigh the disadvantages resulting from the 
lack of harmonization.
    Regarding the comment from BorgWarner about how having a separate 
A/C refrigerant and leakage regulation would allow for better 
harmonization between the programs, the agencies accept this to be an 
accurate statement, but believe the benefits of continued refrigerant 
regulation as an option for CO2 compliance outweigh the 
problems associated with lack of harmonization with the CAFE program.
    For these reasons, EPA is not finalizing the proposed provisions, 
and is making no changes in the A/C refrigerant and leakage-related 
provisions of the current program. In light of this conclusion, EPA 
does not need to address the legal arguments made by CBD et al. and 
CARB about regulating refrigerant-related emissions separately, or 
potential lapses in regulation of refrigerant emissions while such a 
program could be developed.
    As with A/C refrigerant and leakage credits, EPA proposed to 
exclude nitrous oxide and methane from average performance calculations 
after model year 2020, thereby removing these optional program 
flexibilities. Alternatives 1 through 8 excluded this option. EPA 
sought comment on whether to remove those aspects of the program that 
allow a manufacturer to use nitrous oxide and methane emissions 
reductions for compliance with its CO2 average fleet 
standards because such a flexibility is not allowed in the NHTSA CAFE 
program, or whether to retain the flexibilities as a feature that 
differs between the programs. Further, EPA sought comment on whether to 
change the existing methane and nitrous oxide standards. Specifically, 
EPA requested information from the public on whether the existing 
standards are appropriate, or whether they should be revised to be less 
stringent or more stringent based on any updated data.
    The Alliance in its comments may have misunderstood EPA's proposal 
to mean that EPA was proposing to eliminate regulation of methane and 
nitrous oxide emissions altogether. The Alliance commented in support 
of such a proposal as they understood it, to eliminate the standards to 
provide better harmony between the two compliance programs.\207\ The 
Alliance commented that ``[n]ot only is emission of these two 
substances from vehicles a relatively minor contribution to GHG 
emissions, the Alliance has continuing concern regarding measurement 
and testing technologies for nitrous oxide.'' \208\ The Alliance 
commented further that if ``EPA decides instead to continue to regulate 
methane and nitrous oxide, the Alliance recommends that EPA re-assess 
whether the levels of the standards remain appropriate and to retain 
the current compliance flexibilities. Furthermore, in this scenario, 
the Alliance also recommends that methane and nitrous oxide standards 
be assessed as a fleet average and as the average of FTP and HFET test 
cycles.'' \209\ Several individual manufacturers submitted similar 
comments, including Ford,\210\ FCA,\211\ Volvo,\212\ and Mazda.\213\ 
Ford also commented that it does not support the proposal to maintain 
the existing N2O/CH4 standards while removing the 
program flexibilities.\214\
---------------------------------------------------------------------------

    \207\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 13.
    \208\ Id.
    \209\ Id.
    \210\ Ford, EPA-HQ-OAR-2018-0283-5691, at 4.
    \211\ FCA, NHTSA-2018-0067-11943, at 9.
    \212\ Volvo, NHTSA-2018-0067-12036, at 5.
    \213\ Mazda, NHTSA-2018-0067-11727, at 3 (``In reality, these 
emissions are at deminimis levels and have very little, if any, 
impact on global warming. So, the need to regulate these emissions 
as part of the GHG program, or separately, is unclear. Although most 
current engines can comply with the existing requirements, there are 
some existing and upcoming new technologies that may not be able to 
fully comply. These technologies can provide substantial 
CO2 reductions.'').
    \214\ Ford, at 4 (``Finally, without the ability to incorporate 
exceedances into CREE, each vehicle will need to employ hardware 
solutions if they do not comply. We do not believe it was EPA's 
intent in the original rulemaking to require additional after-
treatment, with associated cost increases, explicitly for the 
control and reduction of an insignificant contributor to GHG 
emissions. Therefore, we do not support the proposal to maintain the 
existing N2O/CH4 standards while removing the 
CREE exceedance pathway.'').
---------------------------------------------------------------------------

    The Alliance further commented that ``data from the 2016 EPA report 
on light-duty vehicle emissions supports the position that 
CH4 and N2O have minimal impact on total GHG 
emissions, reporting only 0.045 percent in exceedance of the standard. 
This new information makes it apparent that CH4 and 
N2O contribute a de minimis amount to GHG emissions. 
Additionally, gasoline CH4 and N2O performance is 
within the current standards. Finally, the main producers of 
CH4 and N2O emissions are flex fuel (E85) and 
diesel vehicles, and these vehicles have been declining in sales as 
compared to gasoline-fueled vehicles.'' \215\ The Alliance also 
commented that CH4 and N2O have minimal 
opportunities to be catalytically treated, as N2O is 
generated in the catalyst and CH4 has a low conversion 
efficiency compared to other emissions. EPA did not intend that 
additional hardware should be required to comply with the 
CH4 or N2O standards on any vehicle.'' \216\
---------------------------------------------------------------------------

    \215\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 43.
    \216\ Id. at 44.
---------------------------------------------------------------------------

    Global Automakers commented in support of continuing inclusion of 
nitrous oxide and methane emissions standards for all MYs, even if it 
means a divergence from the NHTSA standards for these program elements 
in the regulations, ``because they are complementary to EPA's program, 
and are better managed through a coordinated federal policy. They are 
also important to maintaining regulatory flexibility through real 
[CO2] emission reductions and would prevent the potential 
for additional bifurcated, separate programs at the state level.'' 
\217\ Global Automakers recommended that they remain in place per the 
existing program but continued to support that the N2O 
testing is not necessary. Global Automakers commented that it 
``strongly recommends reducing the need for N2O testing or 
eliminating these test requirements in their entirety. It should be 
sufficient to allow manufacturers to attest to compliance with the 
N2O capped standards based upon good engineering judgment, 
development testing, and correlation to NOX emissions. EPA 
could, however, maintain the option to request testing to be performed 
for new technologies only, which could have unknown impacts on 
N2O emissions.'' \218\ Hyundai \219\ and Kia \220\ submitted 
similar comments.
---------------------------------------------------------------------------

    \217\ Global, NHTSA-2018-0067-12032, at 4, 5.
    \218\ Global, Appendix A, NHTSA-2018-0067-12032, at A-44, fn. 
89.
    \219\ Hyundai, EPA-HQ-OAR-2018-0283-4411, at 7.
    \220\ Kia, EPA-HQ-OAR-2018-0283-4195, at 8-9.
---------------------------------------------------------------------------

    Others commented in support of retaining the existing program. MECA 
commented that it supports the existing standards for methane and 
nitrous oxide because catalyst technologies provided by MECA members 
that reduce these climate forcing gases are readily

[[Page 24246]]

available and cost-effective.\221\ MECA also commented that the ability 
to trade reductions in these pollutants in exchange for CO2 
gives vehicle manufacturers the flexibilities they need to comply with 
the emission limits by the most cost-effective means.\222\ CBD et al. 
commented that the alternative compliance mechanisms currently 
available in the program exist to provide cost-effective options for 
compliance, and were considered by manufacturers to be a necessary 
element of the program for certain types of vehicles.\223\ CBD et al. 
further argued that ``[e]liminating these flexibilities consequently 
imposes costs on manufacturers without discernible environmental 
benefits,'' and suggested that harmonization with the CAFE program was 
not a relevant decision factor for EPA.\224\ Several other parties 
commented generally in support of retaining the existing program for A/
C leakage credits, discussed above, and N2O and 
CH4 standards.\225\
---------------------------------------------------------------------------

    \221\ MECA, NHTSA-2018-0067-11994, at 12.
    \222\ Id.
    \223\ CBD et al. at 48.
    \224\ Id.
    \225\ Washington State Department of Ecology, NHTSA-2018-0067-
11926, at 6.
---------------------------------------------------------------------------

    After considering these comments, EPA is retaining the regulatory 
provisions related to the N2O and CH4 standards 
with no changes, specifically including the existing flexibilities that 
accompany those standards. EPA is not adopting its proposal to exclude 
nitrous oxide and methane emissions from average performance 
calculations after model year 2020 or any other changes to the program. 
The standards continue to serve their intended purpose of capping 
emissions of those pollutants and providing for more-comprehensive 
regulation of emissions from light-duty vehicles. The standards were 
intended to prevent future emissions increases, and these standards 
were generally not expected to result in the application of new 
technologies or significant costs for manufacturers using current 
vehicle designs.\226\ The program flexibilities are working as intended 
and all manufacturers are successfully complying with the standards. 
Most vehicle models are well below the standards and for those that are 
above the standards, manufacturers have used the flexibilities to 
offset exceedances with CO2 improvements to demonstrate 
compliance. EPA did not receive any data in response to its request for 
comments supporting potential alternative levels of stringency.
---------------------------------------------------------------------------

    \226\ 77 FR 62624, at 62799 (Oct 15, 2012).
---------------------------------------------------------------------------

    While the Alliance and several individual manufacturers recommended 
eliminating the standards altogether, EPA did not propose to eliminate 
the standards, but to eliminate the optional flexibilities, and 
solicited comment on adjusting the standards to be more or less 
stringent. Thus, EPA does not believe it would be appropriate to 
eliminate completely the standards in this final rule without providing 
an opportunity for comment on that idea. Furthermore, as noted above, 
EPA believes the standards are continuing to serve their intended 
purpose of capping emissions and remain appropriate. Manufacturers have 
been subject to the standards for several years, and the Alliance 
acknowledges in their comments that the exceedance of the standards, 
which is offset by manufacturers using compliance flexibilities, is 
very small and that most vehicles meet the standards. Regarding the 
Alliance comments that the standards should be based on a fleet average 
approach, EPA notes that the purpose of the standards is to cap 
emissions, not to achieve fleet-wide reductions.\227\ The fleet average 
emissions for N2O and CH4 are well below the 
numerical level of the cap standards and therefore the existing cap 
standards would not be an appropriate fleet average standard. Adopting 
a fleet average approach using the same numerical level as the 
established cap standards would not achieve the intended goal of 
capping emissions at current levels. If technologies lead to 
exceedances of the caps, automakers have the opportunity to apply 
appropriate flexibilities under the current program to achieve GHG 
emission neutrality. EPA is not aware of any manufacturer that has been 
prevented from bringing a technology to the marketplace because of the 
current cap levels or approach. EPA believes it would need to consider 
all options further, with an opportunity for public comment, before 
adopting such a significant change to the program.
---------------------------------------------------------------------------

    \227\ Relatedly, the Alliance and Global Automakers raised 
concerns in their comments regarding N2O measurement and 
testing burden. EPA did not propose any changes in testing 
requirements and at this time EPA is not adopting any changes. 
Manufacturers have been measuring N2O emissions and have 
successfully certified vehicles to the N2O standards for 
several years and EPA does not believe N2O measurement is 
an issue needing regulatory change. EPA continues to believe direct 
measurement is the best way for manufacturers to demonstrate 
compliance with the N2O standards and is more appropriate 
than an engineering statement without direct measurement.
---------------------------------------------------------------------------

    As explained above, the agencies have changed the alternatives 
considered for the final rule, partly in response to comments. The 
basic form of the standards represented by the alternatives--footprint-
based, defined by particular mathematical functions--remains the same 
and as described in the NPRM. For the EPA program, EPA has chosen in 
this final rule to retain the existing program for regulation of A/C 
refrigerant leakage, nitrous oxide, and methane emissions as part of 
the CO2 standard. This allows manufacturers to continue to 
rely on this flexibility which they describe as extremely important for 
compliance, although it results in continued differences between EPA's 
and NHTSA's programs. This approach also avoids the possibility of gaps 
in the regulation of HFCs, CH4, and N2O while EPA 
developed a different way of regulating the non-CO2 
emissions as part of or concurrent with the NPRM, and thereby allows 
EPA to continue to regulate GHE emissions from light-duty vehicles on a 
more-comprehensive basis. Thus, all alternatives considered in this 
final rule reflect inclusion of CH4, N2O, and HFC 
in EPA's overall ``CO2'' (more accurately, CO2-
equivalent, or CO2e) requirements. Besides this change, the 
alternatives considered for the final rule differ from the NPRM in two 
additional ways: First, alternatives reflecting the phase-out of the A/
C efficiency and off-cycle programs have been dropped in response to 
certain comments and in recognition of the potential real-world 
benefits of those programs. And second, the preferred alternative for 
this final rule reflects a 1.5 percent year-over-year increase for both 
passenger cars and light trucks. These changes will be discussed 
further below, following a brief discussion of the form of the 
standards.

A. Form of the Standards

    As in the CAFE and CO2 rulemakings in 2010 and 2012, 
NHTSA and EPA proposed in the NPRM to set attribute-based CAFE and 
CO2 standards defined by a mathematical function of vehicle 
footprint, which has observable correlation with fuel economy and 
vehicle emissions. EPCA, as amended by EISA, expressly requires that 
CAFE standards for passenger cars and light trucks be based on one or 
more vehicle attributes related to fuel economy and be expressed in the 
form of a mathematical function.\228\ While the CAA includes no 
specific requirements regarding CO2 regulation, EPA has 
chosen to adopt attribute-based CO2 standards consistent 
with NHTSA's EPCA/EISA requirements in the interest of harmonization 
and simplifying compliance. Such an approach is permissible under 
section 202(a) of the

[[Page 24247]]

CAA, and EPA has used the attribute-based approach in issuing standards 
under analogous provisions of the CAA. Thus, both the proposed and 
final standards take the form of fuel economy and CO2 
targets expressed as functions of vehicle footprint (the product of 
vehicle wheelbase and average track width). Section V.A.2 below 
discusses the agencies' continued reliance on footprint as the relevant 
attribute.
---------------------------------------------------------------------------

    \228\ 49 U.S.C. 32902(a)(3)(A).
---------------------------------------------------------------------------

    Under the footprint-based standards, the function defines a 
CO2 or fuel economy performance target for each unique 
footprint combination within a car or truck model type. Using the 
functions, each manufacturer thus will have a CAFE and CO2 
average standard for each year that is almost certainly unique to each 
of its fleets,\229\ based upon the footprints and production volumes of 
the vehicle models produced by that manufacturer. A manufacturer will 
have separate footprint-based standards for cars and for trucks. The 
functions are mostly sloped, so that generally, larger vehicles (i.e., 
vehicles with larger footprints) will be subject to lower CAFE mpg 
targets and higher CO2 grams/mile targets than smaller 
vehicles. This is because, generally speaking, smaller vehicles are 
more capable of achieving higher levels of fuel economy/lower levels of 
CO2 emissions, mostly because they tend not to have to work 
as hard (and therefore require as much energy) to perform their driving 
task. Although a manufacturer's fleet average standards could be 
estimated throughout the model year based on the projected production 
volume of its vehicle fleet (and are estimated as part of EPA's 
certification process), the standards to which the manufacturer must 
comply are determined by its final model year production figures. A 
manufacturer's calculation of its fleet average standards as well as 
its fleets' average performance at the end of the model year will thus 
be based on the production-weighted average target and performance of 
each model in its fleet.\230\
---------------------------------------------------------------------------

    \229\ EPCA/EISA requires NHTSA to separate passenger cars into 
domestic and import passenger car fleets whereas EPA combines all 
passenger cars into one fleet.
    \230\ As discussed in prior rulemakings, a manufacturer may have 
some vehicle models that exceed their target and some that are below 
their target. Compliance with a fleet average standard is determined 
by comparing the fleet average standard (based on the production-
weighted average of the target levels for each model) with fleet 
average performance (based on the production-weighted average of the 
performance of each model).
---------------------------------------------------------------------------

    For passenger cars, consistent with prior rulemakings, NHTSA is 
defining fuel economy targets as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.051

where:

TARGETFE is the fuel economy target (in mpg) applicable to a 
specific vehicle model type with a unique footprint combination,
a is a minimum fuel economy target (in mpg),
b is a maximum fuel economy target (in mpg),
c is the slope (in gallons per mile per square foot, or gpm, per 
square foot) of a line relating fuel consumption (the inverse of 
fuel economy) to footprint, and
d is an intercept (in gpm) of the same line.

    Here, MIN and MAX are functions that take the minimum and maximum 
values, respectively, of the set of included values. For example, 
MIN[40,35] = 35 and MAX(40, 25) = 40, such that MIN[MAX(40, 25), 35] = 
35.
    For light trucks, also consistent with prior rulemakings, NHTSA is 
defining fuel economy targets as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.052

where:

TARGETFE is the fuel economy target (in mpg) applicable to a 
specific vehicle model type with a unique footprint combination,
a, b, c, and d are as for passenger cars, but taking values specific 
to light trucks,
e is a second minimum fuel economy target (in mpg),
f is a second maximum fuel economy target (in mpg),
g is the slope (in gpm per square foot) of a second line relating 
fuel consumption (the inverse of fuel economy) to footprint, and
h is an intercept (in gpm) of the same second line.

    Although the general model of the target function equation is the 
same for each vehicle category (passenger cars and light trucks) and 
each model year, the parameters of the function equation differ for 
cars and trucks. For MYs 2020-2026, the parameters are unchanged, 
resulting in the same stringency in each of those model years.
    Mathematical functions defining the CO2 targets are 
expressed as functions that are similar, with coefficients a-h 
corresponding to those listed above.\231\ For passenger cars, EPA is 
defining CO2 targets mathematically equivalent to the 
following:
---------------------------------------------------------------------------

    \231\ EPA regulations use a different but mathematically 
equivalent approach to specify targets. Rather than using a function 
with nested minima and maxima functions, EPA regulations specify 
requirements separately for different ranges of vehicle footprint. 
Because these ranges reflect the combined application of the listed 
minima, maxima, and linear functions, it is mathematically 
equivalent and more efficient to present the targets as in this 
Section.

---------------------------------------------------------------------------
TARGETCO2 = MIN[b, MAX[a, c x FOOTPRINT + d]]

where:

TARGETCO2 is the is the CO2 target (in grams per mile, or 
g/mi) applicable to a specific vehicle model configuration,
a is a minimum CO2 target (in g/mi),
b is a maximum CO2 target (in g/mi),
c is the slope (in g/mi, per square foot) of a line relating 
CO2 emissions to footprint, and
d is an intercept (in g/mi) of the same line.

    For light trucks, CO2 targets are defined as follows:

TARGETCO2 = MIN[MIN[b, MAX[a, c x FOOTPRINT + d]], MIN[f, MAX[e, g x 
FOOTPRINT + h]]


[[Page 24248]]


where:

TARGETCO2 is the is the CO2 target (in g/mi) applicable 
to a specific vehicle model configuration,
a, b, c, and d are as for passenger cars, but taking values specific 
to light trucks,
e is a second minimum CO2 target (in g/mi),
f is a second maximum CO2 target (in g/mi),
g is the slope (in g/mi per square foot) of a second line relating 
CO2 emissions to footprint, and
h is an intercept (in g/mi) of the same second line.

    To be clear, as has been the case since the agencies began 
establishing attribute-based standards, no vehicle need meet the 
specific applicable fuel economy or CO2 targets, because 
compliance with either CAFE or CO2 standards is determined 
based on corporate average fuel economy or fleet average CO2 
emission rates. In this respect, CAFE and CO2 standards are 
unlike, for example, safety standards and traditional vehicle emissions 
standards. CAFE and CO2 standards apply to the average fuel 
economy levels and CO2 emission rates achieved by 
manufacturers' entire fleets of vehicles produced for sale in the U.S. 
Safety standards apply on a vehicle-by-vehicle basis, such that every 
single vehicle produced for sale in the U.S. must, on its own, comply 
with minimum FMVSS. Similarly, criteria pollutant emissions standards 
are applied on a per-vehicle basis, such that every vehicle produced 
for sale in the U.S. must, on its own, comply with all applicable 
emissions standards. When first mandating CAFE standards in the 1970s, 
Congress specified a more flexible averaging-based approach that allows 
some vehicles to ``under comply'' (i.e., fall short of the overall flat 
standard, or fall short of their target under attribute-based 
standards) as long as a manufacturer's overall fleet is in compliance.
    The required CAFE level applicable to a given fleet in a given 
model year is determined by calculating the production-weighted 
harmonic average of fuel economy targets applicable to specific vehicle 
model configurations in the fleet, as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.053

where:

CAFErequired is the CAFE level the fleet is required to achieve,
i refers to specific vehicle model/configurations in the fleet,
PRODUCTIONi is the number of model configuration i produced for sale 
in the U.S., and
TARGETFE,i the fuel economy target (as defined above) for model 
configuration i.

    Similarly, the required average CO2 level applicable to 
a given fleet in a given model year is determined by calculating the 
production-weighted average (not harmonic) of CO2 targets 
applicable to specific vehicle model configurations in the fleet, as 
follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.054

where:

CO2required is the average CO2 level the fleet is 
required to achieve,
i refers to specific vehicle model/configurations in the fleet,
PRODUCTIONi is the number of model configuration i produced for sale 
in the U.S., and
TARGETCO2,i is the CO2 target (as defined above) for 
model configuration i.

    Section VI.A.1 describes the advantages of attribute standards, 
generally. Section VI.A.2 explains the agencies' specific decision to 
use vehicle footprint as the attribute over which to vary stringency 
for past and current rules. Section VI.A.3 discusses the policy 
considerations in selecting the specific mathematical function. Section 
VI.A.4 discusses the methodologies used to develop current attribute-
based standards, and the agencies' current proposal to continue to do 
so for MYs 2021-2026. Section VI.A.5 discusses the methodologies used 
to reconsider the mathematical function for the proposed standards.
1. Why attribute-based standards, and what are the benefits?
    Under attribute-based standards, every vehicle model has fuel 
economy and CO2 targets, the levels of which depend on the 
level of that vehicle's determining attribute (for the MYs 2021-2026 
standards, footprint is the determining attribute, as discussed below). 
The manufacturer's fleet average CAFE performance is calculated by the 
harmonic production-weighted average of those targets, as defined 
below:
[GRAPHIC] [TIFF OMITTED] TR30AP20.055

    Here, i represents a given model \232\ in a manufacturer's 
fleet, Productioni represents the U.S. production of that model, and 
Targeti represents the target as defined by the attribute-based 
standards. This means no vehicle is required to meet its target; 
instead, manufacturers are free to balance improvements however they 
deem best within (and, given credit transfers, at least partially 
across) their fleets.

    \232\ If a model has more than one footprint variant, here each 
of those variants is treated as a unique model, i, since each 
footprint variant will have a unique target.
---------------------------------------------------------------------------

    Because CO2 is on a gram per mile basis rather a mile 
per gallon basis,

[[Page 24249]]

harmonic averaging is not necessary when calculating required 
CO2 levels:
[GRAPHIC] [TIFF OMITTED] TR30AP20.056

    The idea is to select the shape of the mathematical function 
relating the standard to the fuel economy-related attribute to reflect 
the trade-offs manufacturers face in producing more of that attribute 
over fuel efficiency (due to technological limits of production and 
relative demand of each attribute). If the shape captures these trade-
offs, every manufacturer is more likely to continue adding fuel-
efficient technology across the distribution of the attribute within 
their fleet, instead of potentially changing the attribute--and other 
correlated attributes, including fuel economy--as a part of their 
compliance strategy. Attribute-based standards that achieve this have 
several advantages.
    First, assuming the attribute is a measurement of vehicle size, 
attribute-based standards help to at least partially reduce the 
incentive for manufacturers to respond to CAFE and CO2 
standards by reducing vehicle size in ways harmful to safety, as 
compared to ``flat,'' non-attribute based standards.\233\ Larger 
vehicles, in terms of mass and/or crush space, generally consume more 
fuel and produce more carbon dioxide emissions, but are also generally 
better able to protect occupants in a crash.\234\ Because each vehicle 
model has its own target (determined by a size-related attribute), 
properly fitted attribute-based standards reduce the incentive to build 
smaller vehicles simply to meet a fleet-wide average, because smaller 
vehicles are subject to more stringent compliance targets.
---------------------------------------------------------------------------

    \233\ The 2002 NAS Report described at length and quantified the 
potential safety problem with average fuel economy standards that 
specify a single numerical requirement for the entire industry. See 
Transportation Research Board and National Research Council. 2002. 
Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) 
Standards, Washington, DC: The National Academies Press (``2002 NAS 
Report'') at 5, finding 12, available at https://www.nap.edu/catalog/10172/effectiveness-and-impact-of-corporate-average-fuel-economy-cafe-standards (last accessed June 15, 2018). Ensuing 
analyses, including by NHTSA, support the fundamental conclusion 
that standards structured to minimize incentives to downsize all but 
the largest vehicles will tend to produce better safety outcomes 
than flat standards.
    \234\ Bento, A., Gillingham, K., & Roth, K. (2017). The Effect 
of Fuel Economy Standards on Vehicle Weight Dispersion and Accident 
Fatalities. NBER Working Paper No. 23340. Available at http://www.nber.org/papers/w23340 (last accessed June 15, 2018).
---------------------------------------------------------------------------

    Second, attribute-based standards, if properly fitted, provide 
automakers with more flexibility to respond to consumer preferences 
than do single-valued standards. As discussed above, a single-valued 
standard encourages a fleet mix with a larger share of smaller vehicles 
by creating incentives for manufacturers to use downsizing the average 
vehicle in their fleet (possibly through fleet mixing) as a compliance 
strategy, which may result in manufacturers building vehicles for 
compliance reasons that consumers do not want. Under a size-related, 
attribute-based standard, reducing the size of the vehicle for 
compliance's sake is a less-viable strategy because smaller vehicles 
have more stringent regulatory targets. As a result, the fleet mix 
under such standards is more likely to reflect aggregate consumer 
demand for the size-related attribute used to determine vehicle 
targets.
    Third, attribute-based standards provide a more equitable 
regulatory framework across heterogeneous manufacturers who may each 
produce different shares of vehicles along attributes correlated with 
fuel economy.\235\ An industry-wide single-value CAFE standard imposes 
disproportionate cost burden and compliance challenges on manufacturers 
who produce more vehicles with attributes inherently correlated with 
lower fuel economy--i.e. manufacturers who produce, on average, larger 
vehicles. As discussed above, retaining flexibility for manufacturers 
to produce vehicles which respect heterogeneous market preferences is 
an important consideration. Since manufacturers may target different 
markets as a part of their business strategy, ensuring that these 
manufacturers do not incur a disproportionate share of the regulatory 
cost burden is an important part of conserving consumer choices within 
the market.
---------------------------------------------------------------------------

    \235\ 2002 NAS Report at 4-5, finding 10.
---------------------------------------------------------------------------

    Industry commenters generally supported attribute-based standards, 
while other commenters questioned their benefits. IPI argued that 
preserving the current vehicle mix was not necessarily desirable or 
necessary for consumer welfare, and suggested that some vehicle 
downsizing in the fleet might be beneficial both for safety and for 
compliance.\236\ IPI also argued that compliance credit trading would 
``help smooth out any disproportionate impacts on certain 
manufacturers'' and ``ensure that manufacturers with relatively 
efficient fleets still have an incentive to continue improving fuel 
economy (in order to generate credits)'' \237\ Similarly, citing Ito 
and Sallee, Kathryn Doolittle commented that ``. . . Ito and Sallee 
(2018) have found ABR [``attribute-based regulations''] inefficient in 
cost when juxtaposed with flat standard with compliance trading.'' 
\238\
---------------------------------------------------------------------------

    \236\ IPI, NHTSA-2018-0067-12362, at 14-15.
    \237\ IPI, NHTSA-2018-0067-12362, at 14.
    \238\ Doolittle, K, NHTSA-2018-0067-7411. See also Ito, K and 
Sallee, J. ``The Economics of Attribute-Based Regulation: Theory and 
Evidence from Fuel Economy Standards.'' The Review of Economics and 
Statistics (2018), 100(2), pp. 319-36.
---------------------------------------------------------------------------

    The agencies have considered these comments. IPI incorrectly 
characterizes the agencies' prior statements as claims that it is 
important to preserve the current vehicle mix. EPA and NHTSA have never 
claimed, and are not today claiming that it is important to preserve 
the current fleet mix. The agencies have said, and are today 
reiterating, that it is reasonable to expect that reducing the tendency 
of standards to distort the market should reduce at least part of the 
tendency of standards to reduce consumer welfare. Or, more concisely, 
it is better to work with the market than against it. Single-value (aka 
flat) CAFE standards in place from the 1970s through 2010 were clearly 
distortionary. Recognizing this, the National Academy of Sciences 
recommended in 2002 that NHTSA adopt attribute-based CAFE standards. 
NHTSA did so in 2006, for light trucks produced starting MY 2008. As 
mentioned above, in 2007, Congress codified the requirement for 
attribute-based passenger car and light truck CAFE standards. Agreeing 
with this history, premise, and motivation, EPA has also adopted 
attribute-based CO2 standards. None of this is to say the 
agencies consider it important to hold fleet mix constant. Rather, the 
agencies expect that, compared to flat standards, attribute-based 
standards can allow the market--including fleet mix--to better

[[Page 24250]]

follow its natural course, and all else equal, consumer acceptance is 
likely to be greater if the market does so.
    The agencies also disagree with comments implying that compliance 
credit trading can address all of the market distortion that flat 
standards would entail. Evidence thus far suggests that trading is 
fragmented, with some manufacturers apparently willing to trade only 
with some other specific manufacturers. The Ito and Sallee article 
cited by one commenter is a highly idealized theoretical construction, 
with the authors noting, inter alia, that their model ``assumes perfect 
competition.'' \239\ Its findings regarding comparative economic 
efficiency of flat- and attribute-based standards are, therefore, 
merely hypothetical, and the agencies find little basis in recent 
transactions to suggest the compliance credit trading market reflects 
the authors' idealized assumptions. Even if the agencies did expect 
credit trading markets to operate as in an idealized textbook example, 
basing the structure of standards on the presumption of perfect trading 
would not be appropriate. FCA commented that ``. . . when flexibilities 
are considered while setting targets, they cease to be flexibilities 
and become simply additional technology mandates,'' and the Alliance 
commented, similarly, that ``the Agencies should keep `flexibilities' 
as optional ways to comply and not unduly assume that each flexibility 
allows additional stringency of footprint-based standards.'' \240\ 
Perhaps recognizing this reality, Congress has barred NHTSA from 
considering manufacturers' ability to use compliance credits (even 
credits earned and used by the same OEM, much less credits traded 
between OEMs). As discussed further in Section VIII.A.2, EPA believes 
that while credit trading may be a useful flexibility to reduce the 
overall costs of the program, it is important to set standards in a way 
that does not rely on credit purchasing availability as a compliance 
mechanism.
---------------------------------------------------------------------------

    \239\ Ito and Sallee, op. cit., Supplemental Appendix, at A-15, 
available at https://www.mitpressjournals.org/doi/suppl/10.1162/REST_a_00704/suppl_file/REST_a_00704-esupp.pdf (accessed October 29, 
2019).
    \240\ FCA, NHTSA-2018-0067-11943, at 6; Alliance, NHTSA-2018-
0067-12073, Full Comment Set, at 40, fn. 82.
---------------------------------------------------------------------------

    Considering these comments and realities, considering EPCA's 
requirement for attribute-based CAFE standards, and considering the 
benefits of regulatory harmonization, the agencies are, again, 
finalizing attribute-based CAFE and CO2 standards rather 
than, for either program, finalizing flat standards.
Why footprint as the attribute?
    It is important that the CAFE and CO2 standards be set 
in a way that does not unnecessarily incentivize manufacturers to 
respond by selling vehicles that are less safe. Vehicle size is highly 
correlated with vehicle safety--for this reason, it is important to 
choose an attribute correlated with vehicle size (mass or some 
dimensional measure). Given this consideration, there are several 
policy and technical reasons why footprint is considered to be the most 
appropriate attribute upon which to base the standards, even though 
other vehicle size attributes (notably, curb weight) are more strongly 
correlated with fuel economy and tailpipe CO2 emissions.
    First, mass is strongly correlated with fuel economy; it takes a 
certain amount of energy to move a certain amount of mass. Footprint 
has some positive correlation with frontal surface area, likely a 
negative correlation with aerodynamics, and therefore fuel economy, but 
the relationship is less deterministic. Mass and crush space 
(correlated with footprint) are both important safety considerations. 
As discussed below and in the accompanying PRIA, NHTSA's research of 
historical crash data indicates that holding footprint constant, and 
decreasing the mass of the largest vehicles, will result in a net 
positive safety impact to drivers overall, while holding footprint 
constant and decreasing the mass of the smallest vehicles will result 
in a net decrease in fleetwide safety. Properly fitted footprint-based 
standards provide little, if any, incentive to build smaller footprint 
vehicles to meet CAFE and CO2 standards, and therefore help 
minimize the impact of standards on overall fleet safety.
    Second, it is important that the attribute not be easily 
manipulated in a manner that does not achieve the goals of EPCA or 
other goals, such as safety. Although weight is more strongly 
correlated with fuel economy than footprint, there is less risk of 
artificial manipulation (i.e., changing the attribute(s) to achieve a 
more favorable target) by increasing footprint under footprint-based 
standards than there would be by increasing vehicle mass under weight-
based standards. It is relatively easy for a manufacturer to add enough 
weight to a vehicle to decrease its applicable fuel economy target a 
significant amount, as compared to increasing vehicle footprint, which 
is a much more complicated change that typically takes place only with 
a vehicle redesign.
    Further, some commenters on the MY 2011 CAFE rulemaking were 
concerned that there would be greater potential for such manipulation 
under multi-attribute standards, such as those that also depend on 
weight, torque, power, towing capability, and/or off-road capability. 
As discussed in NHTSA's MY 2011 CAFE final rule,\241\ it is anticipated 
that the possibility of manipulation is lowest with footprint-based 
standards, as opposed to weight-based or multi-attribute-based 
standards. Specifically, standards that incorporate weight, torque, 
power, towing capability, and/or off-road capability in addition to 
footprint would not only be more complex, but by providing degrees of 
freedom with respect to more easily adjusted attributes, they could 
make it less certain that the future fleet would actually achieve the 
projected average fuel economy and CO2 levels. This is not 
to say that a footprint-based system eliminates manipulation, or that a 
footprint-based system eliminates the possibility that manufacturers 
will change vehicles in ways that compromise occupant protection, but 
footprint-based standards achieve the best balance among affected 
considerations.
---------------------------------------------------------------------------

    \241\ See 74 FR at 14359 (Mar. 30, 2009).
---------------------------------------------------------------------------

    Several stakeholders commented on whether vehicular footprint is 
the most suitable attribute upon which to base standards. IPI commented 
that ``. . . footprint-based standards may be unnecessary to respect 
consumer preferences, may negatively impact safety, and may be overall 
inefficient. Several arguments call into question the footprint-based 
approach, but a particularly important one is that large vehicles can 
impose a negative safety externality on other drivers.'' \242\ IPI 
commented, further, that the agencies should consider the relative 
merits of other vehicle attributes, including vehicle fuel type, 
suggesting that it would be more difficult for manufacturers to 
manipulate a flatter standard or one ``differentiated by fuel type.'' 
\243\ Similarly, Michalek and Whitefoot recommended ``that the agencies 
reexamine automaker response to the footprint-based standards to 
determine if adjustments should be made to avoid inducing increases to 
vehicle size.'' \244\
---------------------------------------------------------------------------

    \242\ IPI, NHTSA-2018-0067-12362, at 12.
    \243\ IPI, NHTSA-2018-0067-12362, at 13 et seq.
    \244\ Michalek, J. and Whitefoot, K., NHTSA-2018-0067-11903, at 
13.

---------------------------------------------------------------------------

[[Page 24251]]

    Conversely, ICCT commented that ``the switch to footprint-based 
CAFE and [CO2] standards has been widely credited with 
diminishing safety concerns with efficiency standards. Footprint 
standards encourage larger vehicles with wider track width, which 
reduces rollovers, and longer wheelbase, which increases the crush 
space and reduces deceleration forces for both vehicles in a two-
vehicle collision.'' \245\ Similarly, BorgWarner commented that ``the 
use of a footprint standard not only provides greater incentive for 
mass reduction, but also encourages a larger footprint for a given 
vehicle mass, thus providing increased safety for a given mass 
vehicle,'' \246\ and the Aluminum Association commented footprint based 
standards drive ``fuel-efficiency improvement across all vehicle 
classes,'' ``eliminate the incentive to shift fleet volume to smaller 
cars which has been shown to slightly decrease safety in vehicle-to-
vehicle collisions,'' and provide ``an incentive for reducing weight in 
the larger vehicles, where weight reduction is of the most benefit for 
societal safety,'' citing Ford's aluminum-intensive F150 pickup truck 
as an example.\247\ NADA urged the agencies to continue basing 
standards on vehicle footprint, as doing so ``serves both to require 
and allow OEMs to build more fuel-efficient vehicles across the 
broadest possible light-duty passenger car and truck spectrum,'' \248\ 
and UCS commented that footprint-based standards ``increase consumer 
choice, ensuring that the vehicles available for purchase in every 
vehicle class continue to get more efficient.'' \249\ Furthermore, 
regarding concerns that footprint-based standards may be susceptible to 
manipulation, the Alliance commented that ``the data above [from 
Novation Analytics] shows there are no systemic footprint increases (or 
any type of target manipulation) occurring.'' \250\ While FCA's 
comments supported this Alliance comment, FCA commented further that, 
lacking some utility-related vehicle attributes such as towing 
capability, 4-wheel-drive, and ride height, ``it is clear the footprint 
standard does not fully account for pickup truck capability and the 
components needed such as larger powertrains, greater mass and frontal 
area,'' and requested the agencies ``correct LDT standards to reflect 
the current market preference for capability over efficiency, and 
introduce mechanisms into the regulation that can adjust for efficiency 
and capability tradeoffs that footprint standards currently ignore.'' 
\251\
---------------------------------------------------------------------------

    \245\ ICCT, NHTSA-2018-0067-11741, at B-4.
    \246\ BorgWarner, NHTSA-2018-0067-11893, at 10.
    \247\ Aluminum Association, NHTSA-2018-0067-11952, at 3.
    \248\ NADA, NHTSA-2018-0067-12064, at 13.
    \249\ UCS, UCS, NHTSA-2018-0067-12039, at 46.
    \250\ Alliance, NHTSA-2018-0067-12073, at 123.
    \251\ FCA, NHTSA-2018-0067-11943, at 49.
---------------------------------------------------------------------------

    When first electing to adopt footprint-based standards, NHTSA 
carefully considered other alternatives, including vehicle mass and 
``shadow'' (overall width multiplied by overall length). Compared to 
both of these other alternatives, footprint is much less susceptible to 
gaming, because while there is some potential to adjust track width, 
wheelbase is more expensive to change, at least outside a planned 
vehicle redesign. EPA agreed with NHTSA's assessment, nothing has 
changed the relative merits of at least these three potential 
attributes, and nothing in the evolution of the fleet demonstrates that 
footprint-based standards are leading manufacturers to increase the 
footprint of specific vehicle models by more than they would in 
response to customer demand. Also, even if footprint-based standards 
are encouraging some increases in vehicle size, NHTSA continues to 
maintain, and EPA to agree, that such increases should tend to improve 
overall highway safety rather than degrading it. Regarding FCA's 
request that the agencies adopt an approach that accounts for a wider 
range of vehicle attributes related to both vehicle fuel economy and 
customer-facing vehicle utility, the agencies are concerned that doing 
so could further complicate already-complex standards and also lead to 
unintended consequences. For example, it is not currently clear how a 
multi-attribute approach would appropriately balance emphasis between 
vehicle attributes (e.g., how much relative fuel consumption should be 
attributed to, respectively, vehicle footprint, towing capacity, drive 
type, and ground clearance). Also, basing standards on, in part, ground 
clearance would encourage manufacturers to increase ride height, 
potentially increasing the frequency of vehicle rollover crashes. 
Regarding IPI's recommendation that fuel type be included as a vehicle 
attribute for attribute-based standards, the agencies note that both 
CAFE and CO2 standards already account for fuel type in the 
procedures for measuring fuel economy levels and CO2 
emission rates, and for calculating fleet average CAFE and 
CO2 levels.
    Therefore, having considered public comments on the choice of 
vehicle attributes for CAFE and CO2 standards, the agencies 
are finalizing standards that, as proposed, are defined in terms of 
vehicle footprint.
3. What mathematical function should be used to specify footprint-based 
standards?
    In requiring NHTSA to ``prescribe by regulation separate average 
fuel economy standards for passenger and non-passenger automobiles 
based on 1 or more vehicle attributes related to fuel economy and 
express each standard in the form of a mathematical function,'' EPCA/
EISA provides ample discretion regarding not only the selection of the 
attribute(s), but also regarding the nature of the function. The CAA 
provides no specific direction regarding CO2 regulation, and 
EPA has continued to harmonize this aspect of its CO2 
regulations with NHTSA's CAFE regulations. The relationship between 
fuel economy (and CO2 emissions) and footprint, though 
directionally clear (i.e., fuel economy tends to decrease and 
CO2 emissions tend to increase with increasing footprint), 
is theoretically vague, and quantitatively uncertain; in other words, 
not so precise as to a priori yield only a single possible curve.
    The decision of how to specify this mathematical function therefore 
reflects some amount of judgment. The function can be specified with a 
view toward achieving different environmental and petroleum reduction 
goals, encouraging different levels of application of fuel-saving 
technologies, avoiding any adverse effects on overall highway safety, 
reducing disparities of manufacturers' compliance burdens, and 
preserving consumer choice, among other aims. The following are among 
the specific technical concerns and resultant policy tradeoffs the 
agencies have considered in selecting the details of specific past and 
future curve shapes:
     Flatter standards (i.e., curves) increase the risk that 
both the size of vehicles will be reduced, potentially compromising 
highway safety, and reducing any utility consumers would have gained 
from a larger vehicle.
     Steeper footprint-based standards may create incentives to 
upsize vehicles, potentially oversupplying vehicles of certain 
footprints beyond what consumers would naturally demand, and thus 
increasing the possibility that fuel savings and CO2 
reduction benefits will be forfeited artificially.
     Given the same industry-wide average required fuel economy 
or CO2 standard, flatter standards tend to place greater 
compliance burdens on full-line manufacturers.
     Given the same industry-wide average required fuel economy 
or CO2

[[Page 24252]]

standard, dramatically steeper standards tend to place greater 
compliance burdens on limited-line manufacturers (depending of course, 
on which vehicles are being produced).
     If cutpoints are adopted, given the same industry-wide 
average required fuel economy, moving small-vehicle cutpoints to the 
left (i.e., up in terms of fuel economy, down in terms of 
CO2 emissions) discourages the introduction of small 
vehicles, and reduces the incentive to downsize small vehicles in ways 
that could compromise overall highway safety.
     If cutpoints are adopted, given the same industry-wide 
average required fuel economy, moving large-vehicle cutpoints to the 
right (i.e., down in terms of fuel economy, up in terms of 
CO2 emissions) better accommodates the design requirements 
of larger vehicles--especially large pickups--and extends the size 
range over which downsizing is discouraged.
4. What mathematical functions have been used previously, and why?
    Notwithstanding the aforementioned discretion under EPCA/EISA, data 
should inform consideration of potential mathematical functions, but 
how relevant data is defined and interpreted, and the choice of 
methodology for fitting a curve to that data, can and should include 
some consideration of specific policy goals. This section summarizes 
the methodologies and policy concerns that were considered in 
developing previous target curves (for a complete discussion see the 
2012 FRIA).
    As discussed below, the MY 2011 final curves followed a constrained 
logistic function defined specifically in the final rule.\252\ The MYs 
2012-2021 final standards and the MYs 2022-2025 augural standards are 
defined by constrained linear target functions of footprint, as shown 
below: \253\
---------------------------------------------------------------------------

    \252\ See 74 FR 14196, 14363-14370 (Mar. 30, 2009) for NHTSA 
discussion of curve fitting in the MY 2011 CAFE final rule.
    \253\ The right cutpoint for the light truck curve was moved 
further to the right for MYs 2017-2021, so that more possible 
footprints would fall on the sloped part of the curve. In order to 
ensure that, for all possible footprints, future standards would be 
at least as high as MY 2016 levels, the final standards for light 
trucks for MYs 2017-2021 is the maximum of the MY 2016 target curves 
and the target curves for the give MY standard. This is defined 
further in the 2012 final rule. See 77 FR 62624, at 62699-700 (Oct. 
15, 2012).
[GRAPHIC] [TIFF OMITTED] TR30AP20.057

    Here, Target is the fuel economy target applicable to vehicles 
of a given footprint in square feet (Footprint). The upper 
asymptote, a, and the lower asymptote, b, are specified in mpg; the 
reciprocal of these values represent the lower and upper asymptotes, 
respectively, when the curve is instead specified in gallons per 
mile (gpm). The slope, c, and the intercept, d, of the linear 
portion of the curve are specified as gpm per change in square feet, 
---------------------------------------------------------------------------
and gpm, respectively.

    The min and max functions will take the minimum and maximum values 
within their associated parentheses. Thus, the max function will first 
find the maximum of the fitted line at a given footprint value and the 
lower asymptote from the perspective of gpm. If the fitted line is 
below the lower asymptote it is replaced with the floor, which is also 
the minimum of the floor and the ceiling by definition, so that the 
target in mpg space will be the reciprocal of the floor in mpg space, 
or simply, a. If, however, the fitted line is not below the lower 
asymptote, the fitted value is returned from the max function and the 
min function takes the minimum value of the upper asymptote (in gpm 
space) and the fitted line. If the fitted value is below the upper 
asymptote, it is between the two asymptotes and the fitted value is 
appropriately returned from the min function, making the overall target 
in mpg the reciprocal of the fitted line in gpm. If the fitted value is 
above the upper asymptote, the upper asymptote is returned is returned 
from the min function, and the overall target in mpg is the reciprocal 
of the upper asymptote in gpm space, or b.
    In this way curves specified as constrained linear functions are 
specified by the following parameters:

a = upper limit (mpg)
b = lower limit (mpg)
c = slope (gpm per sq.ft.)
d = intercept (gpm)

    The slope and intercept are specified as gpm per sq. ft. and gpm 
instead of mpg per sq. ft. and mpg because fuel consumption and 
emissions appear roughly linearly related to gallons per mile (the 
reciprocal of the miles per gallon).
a) NHTSA in MY 2008 and MY 2011 CAFE (Constrained Logistic)
    For the MY 2011 CAFE rule, NHTSA estimated fuel economy levels by 
footprint from the MY 2008 fleet after normalization for differences in 
technology,\254\ but did not make adjustments to reflect other vehicle 
attributes (e.g., power-to-weight ratios). Starting with the 
technology-adjusted passenger car and light truck fleets, NHTSA used 
minimum absolute deviation (MAD) regression without sales weighting to 
fit a logistic form as a starting point to develop mathematical 
functions defining the standards. NHTSA then identified footprints at 
which to apply minimum and maximum values (rather than letting the 
standards extend without limit) and transposed these functions 
vertically (i.e., on a gallons-per-mile basis, uniformly downward) to 
produce the promulgated standards. In the preceding rule, for MYs 2008-
2011 light truck standards, NHTSA examined a range of potential 
functional forms, and concluded that, compared to other considered 
forms, the constrained logistic form provided the expected and 
appropriate trend (decreasing fuel economy as footprint increases), but 
avoided creating ``kinks'' the agency was concerned would provide 
distortionary incentives for vehicles with neighboring footprints.\255\
---------------------------------------------------------------------------

    \254\ See 74 FR 14196, 14363-14370 (Mar. 30, 2009) for NHTSA 
discussion of curve fitting in the MY 2011 CAFE final rule.
    \255\ See 71 FR 17556, 17609-17613 (Apr. 6, 2006) for NHTSA 
discussion of ``kinks'' in the MYs 2008-2011 light truck CAFE final 
rule (there described as ``edge effects''). A ``kink,'' as used 
here, is a portion of the curve where a small change in footprint 
results in a disproportionally large change in stringency.
---------------------------------------------------------------------------

b) MYs 2012-2016 Standards (Constrained Linear)
    For the MYs 2012-2016 rule, potential methods for specifying 
mathematical functions to define fuel economy and CO2 
standards were reevaluated. These methods were fit to the same MY 2008 
data as the MY 2011 standard. Considering these further specifications, 
the constrained logistic form, if applied to post-MY 2011 standards, 
would likely contain a steep mid-section that would provide undue 
incentive to increase the footprint of midsize passenger cars.\256\ A 
range of

[[Page 24253]]

methods to fit the curves would have been reasonable, and a minimum 
absolute deviation (MAD) regression without sales weighting on a 
technology-adjusted car and light truck fleet was used to fit a linear 
equation. This equation was used as a starting point to develop 
mathematical functions defining the standards. Footprints were then 
identified at which to apply minimum and maximum values (rather than 
letting the standards extend without limit). Finally, these 
constrained/piecewise linear functions were transposed vertically 
(i.e., on a gpm or CO2 basis, uniformly downward) by 
multiplying the initial curve by a single factor for each MY standard 
to produce the final attribute-based targets for passenger cars and 
light trucks described in the final rule.\257\ These transformations 
are typically presented as percentage improvements over a previous MY 
target curve.
---------------------------------------------------------------------------

    \256\ 75 FR at 25362.
    \257\ See generally 74 FR at 49491-96; 75 FR at 25357-62.
---------------------------------------------------------------------------

c) MYs 2017 and Beyond Standards (Constrained Linear)
    The mathematical functions finalized in 2012 for MYs 2017 and 
beyond changed somewhat from the functions for the MYs 2012-2016 
standards. These changes were made both to address comments from 
stakeholders, and to consider further some of the technical concerns 
and policy goals judged more preeminent under the increased uncertainty 
of the impacts of finalizing and proposing standards for model years 
further into the future.\258\ Recognizing the concerns raised by full-
line OEMs, it was concluded that continuing increases in the stringency 
of the light truck standards would be more feasible if the light truck 
curve for MYs 2017 and beyond was made steeper than the MY 2016 truck 
curve and the right (large footprint) cut-point was extended only 
gradually to larger footprints. To accommodate these considerations, 
the 2012 final rule finalized the slope fit to the MY 2008 fleet using 
a sales-weighted, ordinary least-squares regression, using a fleet that 
had technology applied to make the technology application across the 
fleet more uniform, and after adjusting the data for the effects of 
weight-to-footprint. Information from an updated MY 2010 fleet was also 
considered to support this decision. As the curve was vertically 
shifted (with fuel economy specified as mpg instead of gpm or 
CO2 emissions) upwards, the right cutpoint was progressively 
moved for the light truck curves with successive model years, reaching 
the final endpoint for MY 2021.
---------------------------------------------------------------------------

    \258\ The MYs 2012-2016 final standards were signed April 1st, 
2010--putting 6.5 years between its signing and the last affected 
model year, while the MYs 2017-2021 final standards were signed 
August 28th, 2012--giving just more than nine years between signing 
and the last affected final standards.
---------------------------------------------------------------------------

5. Reconsidering the Mathematical Functions for Today's Rulemaking
a) Why is it important to reconsider the mathematical functions?
    By shifting the developed curves by a single factor, it is assumed 
that the underlying relationship of fuel consumption (in gallons per 
mile) to vehicle footprint does not change significantly from the model 
year data used to fit the curves to the range of model years for which 
the shifted curve shape is applied to develop the standards. However, 
it must be recognized that the relationship between vehicle footprint 
and fuel economy is not necessarily constant over time; newly developed 
technologies, changes in consumer demand, and even the curves 
themselves could influence the observed relationships between the two 
vehicle characteristics. For example, if certain technologies are more 
effective or more marketable for certain types of vehicles, their 
application may not be uniform over the range of vehicle footprints. 
Further, if market demand has shifted between vehicle types, so that 
certain vehicles make up a larger share of the fleet, any underlying 
technological or market restrictions which inform the average shape of 
the curves could change. That is, changes in the technology or market 
restrictions themselves, or a mere re-weighting of different vehicles 
types, could reshape the fit curves.
    For the above reasons, the curve shapes were reconsidered in the 
proposal using the newest available data from MY 2016. With a view 
toward corroboration through different techniques, a range of 
descriptive statistical analyses were conducted that do not require 
underlying engineering models of how fuel economy and footprint might 
be expected to be related, and a separate analysis that uses vehicle 
simulation results as the basis to estimate the relationship from a 
perspective more explicitly informed by engineering theory was 
conducted as well. Despite changes in the new vehicle fleet both in 
terms of technologies applied and in market demand, the underlying 
statistical relationship between footprint and fuel economy has not 
changed significantly since the MY 2008 fleet used for the 2012 final 
rule; therefore, EPA and NHTSA proposed to continue to use the curve 
shapes fit in 2012. The analysis and reasoning supporting this decision 
follows.
b) What statistical analyses did EPA and NHTSA consider?
    In considering how to address the various policy concerns discussed 
above, data from the MY 2016 fleet was considered, and a number of 
descriptive statistical analyses (i.e., involving observed fuel economy 
levels and footprints) using various statistical methods, weighting 
schemes, and adjustments to the data to make the fleets less 
technologically heterogeneous were performed. There were several 
adjustments to the data that were common to all of the statistical 
analyses considered.
    With a view toward isolating the relationship between fuel economy 
and footprint, the few diesels in the fleet were excluded, as well as 
the limited number of vehicles with partial or full electric 
propulsion; when the fleet is normalized so that technology is more 
homogenous, application of these technologies is not allowed. This is 
consistent with the methodology used in the 2012 final rule.
    The above adjustments were applied to all statistical analyses 
considered, regardless of the specifics of each of the methods, 
weights, and technology level of the data, used to view the 
relationship of vehicle footprint and fuel economy. Table V-1, below, 
summarizes the different assumptions considered and the key attributes 
of each. The analysis was performed considering all possible 
combinations of these assumptions, producing a total of eight footprint 
curves.

[[Page 24254]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.058

(1) Current Technology Level Curves
    The ``current technology'' level curves exclude diesels and 
vehicles with electric propulsion, as discussed above, but make no 
other changes to each model year fleet. Comparing the MY 2016 curves to 
ones built under the same methodology from previous model year fleets 
shows whether the observed curve shape has changed significantly over 
time as standards have become more stringent. Importantly, these curves 
will include any market forces which make technology application 
variable over the distribution of footprint. These market forces will 
not be present in the ``maximum technology'' level curves: By making 
technology levels homogenous, this variation is removed. The current 
technology level curves built using both regression types and both 
regression weight methodologies from the MY 2008, MY 2010, and MY 2016 
fleets, shown in more detail in Chapter 4.4.2.1 of the PRIA, support 
the curve slopes finalized in the 2012 final rule. The curves built 
from most methodologies using each fleet generally shift, but remain 
very similar in slope. This suggests that the relationship of footprint 
to fuel economy, including both technology and market limits, has not 
significantly changed.
(2) Maximum Technology Level Curves
    As in prior rulemakings, technology differences between vehicle 
models were considered to be a significant factor producing uncertainty 
regarding the relationship between fuel consumption and footprint. 
Noting that attribute-based standards are intended to encourage the 
application of additional technology to improve fuel efficiency and 
reduce CO2 emissions across the distribution of footprint in 
the fleet, approaches were considered in which technology application 
is simulated for purposes of the curve fitting analysis in order to 
produce fleets that are less varied in technology content. This 
approach helps reduce ``noise'' (i.e., dispersion) in the plot of 
vehicle footprints and fuel consumption levels and identify a more 
technology-neutral relationship between footprint and fuel consumption. 
The results of updated analysis for maximum technology level curves are 
also shown in Chapter 4.4.2.2 of the PRIA. Especially if vehicles 
progress over time toward more similar size-specific efficiency, 
further removing variation in technology application both better 
isolates the relationship between fuel consumption and footprint and 
further supports the curve slopes finalized in the 2012 final rule.
c) What other methodologies were considered?
    The methods discussed above are descriptive in nature, using 
statistical analysis to relate observed fuel economy levels to observed 
footprints for known vehicles. As such, these methods are clearly based 
on actual data, answering the question ``how does fuel economy appear 
to be related to footprint?'' However, being independent of explicit 
engineering theory, they do not answer the question ``how might one 
expect fuel economy to be related to footprint?'' Therefore, as an 
alternative to the above methods, an alternative methodology was also 
developed and applied that, using full-vehicle simulation, comes closer 
to answering the second question, providing a basis either to 
corroborate answers to the first, or suggest that further investigation 
could be important.
    As discussed in the 2012 final rule, several manufacturers have 
confidentially shared with the agencies what they described as 
``physics-based'' curves, with each OEM showing significantly different 
shapes for the footprint-fuel economy relationships. This variation 
suggests that manufacturers face different curves given the other 
attributes of the vehicles in their fleets (i.e., performance

[[Page 24255]]

characteristics) and/or that their curves reflected different levels of 
technology application. In reconsidering the shapes of the proposed MYs 
2021-2026 standards, a similar estimation of physics-based curves 
leveraging third-party simulation work form Argonne National 
Laboratories (Argonne) was developed. Estimating physics-based curves 
better ensures that technology and performance are held constant for 
all footprints; augmenting a largely statistical analysis with an 
analysis that more explicitly incorporates engineering theory helps to 
corroborate that the relationship between fuel economy and footprint is 
in fact being characterized.
    Tractive energy is the amount of energy it will take to move a 
vehicle.\259\ Here, tractive energy effectiveness is defined as the 
share of the energy content of fuel consumed which is converted into 
mechanical energy and used to move a vehicle--for internal combustion 
engine (ICE) vehicles, this will vary with the relative efficiency of 
specific engines. Data from Argonne simulations suggest that the limits 
of tractive energy effectiveness are approximately 25 percent for 
vehicles with internal combustion engines which do not possess 
integrated starter generator, other hybrid, plug-in, pure electric, or 
fuel cell technology.
---------------------------------------------------------------------------

    \259\ Thomas, J. ``Drive Cycle Powertrain Efficiencies and 
Trends Derived from EPA Vehicle Dynamometer Results,'' SAE Int. J. 
Passeng. Cars--Mech. Syst. 7(4):2014, doi:10.4271/2014-01-2562. 
Available at https://www.sae.org/publications/technical-papers/content/2014-01-2562/ (last accessed June 15, 2018).
---------------------------------------------------------------------------

    A tractive energy prediction model was also developed to support 
today's proposal. Given a vehicle's mass, frontal area, aerodynamic 
drag coefficient, and rolling resistance as inputs, the model will 
predict the amount of tractive energy required for the vehicle to 
complete the Federal test cycle. This model was used to predict the 
tractive energy required for the average vehicle of a given footprint 
\260\ and ``body technology package'' to complete the cycle. The body 
technology packages considered are defined in Table V-2, below. Using 
the absolute tractive energy predicted and tractive energy 
effectiveness values spanning possible ICE engines, fuel economy values 
were then estimated for different body technology packages and engine 
tractive energy effectiveness values.
---------------------------------------------------------------------------

    \260\ The mass reduction curves used elsewhere in this analysis 
were used to predict the mass of a vehicle with a given footprint, 
body style box, and mass reduction level. The `Body style Box' is 1 
for hatchbacks and minivans, 2 for pickups, and 3 for sedans, and is 
an important predictor of aerodynamic drag. Mass is an essential 
input in the tractive energy calculation.
[GRAPHIC] [TIFF OMITTED] TR30AP20.059

    Chapter 6 of the PRIA show the resultant CAFE levels estimated for 
the vehicle classes Argonne simulated for this analysis, at different 
footprint values and by vehicle ``box.'' Pickups are considered 1-box, 
hatchbacks and minivans are 2-box, and sedans are 3-box. These 
estimates are compared with the MY 2021 standards finalized in 2012. 
The general trend of the simulated data points follows the pattern of 
the previous MY 2021 standards for all technology packages and tractive 
energy effectiveness values presented in the PRIA. The tractive energy 
curves are intended to validate the curve shapes against a physics-
based alternative, and the analysis suggests that the curve shapes 
track the physical relationship between fuel economy and tractive 
energy for different footprint values.
    Physical limitations are not the only forces manufacturers face; 
their success is dependent upon producing vehicles that consumers 
desire and will purchase. For this reason, in setting future standards, 
the analysis will continue to consider information from statistical 
analyses that do not homogenize technology applications in addition to 
statistical analyses which do, as well as a tractive energy analysis 
similar to the one presented above.
    The relationship between fuel economy and footprint remains 
directionally discernable but quantitatively uncertain. Nevertheless, 
each standard must commit to only one function. Approaching the 
question ``how is fuel economy related to footprint'' from different 
directions and applying different approaches has given EPA and NHTSA 
confidence that the function applied here appropriately and reasonably 
reflects the relationship between fuel economy and footprint.
    The agencies invited comments on this conclusion and the supporting 
analysis. IPI raised concerns that ``. . . several dozen models (mostly 
subcompacts and sports cars) fall in the 30-40 square feet range, which 
are all subject to the same standards'' and that ``manufacturers of 
these models may have an incentive to decrease footprints as a 
compliance strategy, since doing so would not trigger more stringent 
standards.'' \261\ NHTSA and EPA agree that, all else equal, downsizing 
the smallest cars (e.g., Chevrolet Spark, Ford Fiesta, Mini Cooper, 
Mazda MX-5, Porsche 911, Toyota Yaris) would most likely tend to 
degrade overall highway safety. At the same time, as discussed above, 
the agencies recognize that small vehicles do appear attractive to some 
market segments (although obviously the Ford Fiesta and Porsche 911 
compete in different segments).

[[Page 24256]]

Therefore, there is a tension between on one hand, avoiding standards 
that unduly encourage safety-eroding downsizing and, on the other, 
avoiding standards that unduly penalize the market for small vehicles. 
The agencies examined this issue, and note that the market for the 
smallest vehicles has not evolved at all as estimated in the analysis 
supporting the 2012 final rule, and attribute this more to fuel prices 
and consumer demand for larger vehicles than to attribute-based CAFE 
and CO2 standards. For example, the market for vehicles with 
footprints less than 40 square foot was about 45 percent smaller in MY 
2017 than in MY 2010. The agencies also found that among the smallest 
vehicle models produced throughout MYs 2010-2017, most have become 
larger, not smaller. For example, while the Mazda MX-5's footprint 
decreased by 0.1 square foot (0.3 percent) during that time, the MY 
2017 versions of the Mini Cooper, Smart fortwo, Porsche 911, and Toyota 
Yaris had larger footprints than in MY 2010. With the market for very 
small vehicles shrinking, and with manufacturers not evidencing a 
tendency to make the smallest vehicles even smaller, the agencies are 
satisfied that it would be unwise to change the target functions such 
that targets never stop becoming more stringent as vehicle footprint 
becomes ever smaller, because doing so could further impede an already-
shrinking market.
---------------------------------------------------------------------------

    \261\ IPI, NHTSA-2018-0067-12362, p. 14.
---------------------------------------------------------------------------

B. No-Action Alternative

    As in the proposal, the No-Action Alternative applies the augural 
CAFE and final CO2 targets announced in 2012 for MYs 2021-
2025.\262\ For MY 2026, this alternative applies the same targets as 
for MY 2025. The carbon dioxide equivalent of air conditioning 
refrigerant leakage credits, nitrous oxide, and methane emissions are 
included for compliance with the EPA standards for all model years 
under the no-action alternative.\263\
---------------------------------------------------------------------------

    \262\ https://www.govinfo.gov/content/pkg/CFR-2014-title40-vol19/pdf/CFR-2014-title40-vol19-sec86-1818-12.pdf
    \263\ EPA regulations use a different but mathematically 
equivalent approach to specify targets. Rather than using a function 
with nested minima and maxima functions, EPA regulations specify 
requirements separately for different ranges of vehicle footprint. 
Because these ranges reflect the combined application of the listed 
minima, maxima, and linear functions, it is mathematically 
equivalent and more efficient to present the targets as in this 
Section.
[GRAPHIC] [TIFF OMITTED] TR30AP20.060


[[Page 24257]]


    In comments on the DEIS, CBD et al. indicated that it was 
appropriate for NHTSA to use the augural CAFE standards as the baseline 
No Action regulatory alternative.\264\ However, CARB commented that the 
baseline regulatory alternative should include CARB's ZEV mandate, in 
part because EPA must consider ``other regulations promulgated by EPA 
or other government entities,'' and, according to CARB, there will be 
much more vehicle electrification in the future as manufacturers 
respond to market demand and also work to comply with the ZEV 
mandate.\265\ Similarly, EPA's Science Advisory Board recommended--
despite the action taken in the One National Program Action--that the 
baseline include state ZEV mandates ``to be consistent with policies 
that would prevail in the absence of the rule change.'' \266\ EPA's 
Science Advisory Board further recommended including sensitivity 
analyses with different penetration rates of ZEVs.
---------------------------------------------------------------------------

    \264\ CBD et al., NHTSA-2018-0067-12123, Attachment 1, at 13.
    \265\ CARB, NHTSA-2018-0067-11873, at 124-125.
    \266\ SAB at 12 and 29-30.
---------------------------------------------------------------------------

    On the other hand, arguing for consideration of standards less 
stringent than those proposed in the NPRM, Walter Kreucher commented 
that rather than using the augural standards as the baseline, ``a 
better approach would be to assume a clean sheet of paper and start 
from the existing 2016MY fleet and its associated standards as the 
baseline using 0%/year increases for both passenger cars and light 
trucks for MYs 2017-2026.'' \267\ Similarly, AVE argued that because 
previously-promulgated standards for MYs 2018-2021 already present a 
significant challenge that ``will likely require almost every automaker 
to continue using credits for compliance, . . . AVE believes this 
rulemaking should reset . . . the current compliance baseline for cars 
and light trucks at MY 2018 . . .'' \268\ BorgWarner commented 
similarly that ``Beginning in MY 2018, standards should be reset to the 
levels the industry actually achieved. For MY 2018 and beyond, 
succeeding model year targets should be set with an annual rate of 
improvement defined by the slope of improvement the industry has 
achieved over the last six years. . . . Based on these data, our 
analysis suggests the most reasonable and logical rate of improvement 
falls between 2.0% to 2.6% for cars and trucks. Additionally, a single 
rate of improvement for the combined fleet should be considered.'' 
\269\
---------------------------------------------------------------------------

    \267\ Kreucher, W., NHTSA-2018-0067-0444, at 8.
    \268\ AVE, NHTSA-2018-0067-11696, at 8-9.
    \269\ BorgWarner, NHTSA-2018-0067-11895, at 3, 6.
---------------------------------------------------------------------------

    The No-Action Alternative represents expectations regarding the 
world in the absence of a proposal, accounting for applicable laws 
already in place. Although manufacturers are already making significant 
use of compliance credits toward compliance with even MY 2017 
standards, the agencies are obligated to evaluate regulatory 
alternatives against the standards already in place through MY 2025. 
Similarly, even though manufacturers are already producing electric 
vehicles, EPA and NHTSA appropriately excluded California's ZEV mandate 
from the No-Action alternative for the NPRM, for several reasons. 
First, the ZEV mandate is not Federal law; second, as described in the 
proposal and subsequently finalized in regulatory text, the ZEV mandate 
is expressly and impliedly preempted by EPCA; third, EPA proposed to 
withdraw the waiver of CAA preemption in the NPRM and subsequently 
finalized this withdrawal. Accordingly, the agencies have, therefore, 
appropriately excluded the ZEV mandate from the No-Action alternative. 
However, as discussed below, the agencies' analysis does account for 
the potential that under every regulatory alternative, including the 
No-Action Alternative, vehicle electrification could increase in the 
future, especially if batteries become less expensive as gasoline 
becomes more expensive.

C. Action Alternatives

1. Alternatives in Final Rule
    Table V-5 below shows the different alternatives evaluated in 
today's notice.
[GRAPHIC] [TIFF OMITTED] TR30AP20.061


[[Page 24258]]


    With one exception, the alternatives considered in the NPRM 
included the changes in stringency for the above alternatives. 
Alternative 3, the preferred alternative, is newly included for today's 
notice.\270\
---------------------------------------------------------------------------

    \270\ As the agencies indicated in the NPRM, they were 
considering and taking comment ``on a wide range of alternatives and 
have specifically modeled eight alternatives.'' 83 FR at 42990 (Aug. 
24, 2018). The preferred alternative in this final rule was within 
the range of alternatives considered in the proposal, although it 
was not specifically modeled at that time. This issue is discussed 
in further detail below.
---------------------------------------------------------------------------

    Regulations regarding implementation of NEPA requires agencies to 
``rigorously explore and objectively evaluate all reasonable 
alternatives, and for alternatives which were eliminated from detailed 
study, briefly discuss the reasons for their having been eliminated.'' 
\271\ This does not amount to a requirement that agencies evaluate the 
widest conceivable spectrum of alternatives. For example, a State 
considering adding a single travel lane to a preexisting section of 
highway would not be required to consider adding three lanes, or to 
consider dismantling the highway altogether.
---------------------------------------------------------------------------

    \271\ 40 CFR 1502.14.
---------------------------------------------------------------------------

    Among thousands of individual comments that mentioned the proposed 
standards very generally, some comments addressed the range and 
definition of these regulatory alternatives in specific terms, and 
these specific comments include comments on the stringency, structure, 
and particular provisions defining the set of regulatory alternatives 
under consideration.
    As discussed throughout today's notice, the agencies have updated 
and otherwise revised many aspects of the analysis. The agencies have 
also reconsidered whether the set of alternatives studied in detail 
should be expanded to include standards less stringent than the 
proposal's preferred alternative, or to include standards more 
stringent than the proposal's no-action alternative. On one hand, 
comments from Walter Kreucher and AVE cited above indicate the agencies 
should consider relaxing standards below MY 2020 levels, and CEI 
challenged the agencies' failure to include less-stringent alternatives 
in the following comments on this question:

    DOT failed to consider the possibility of freezing CAFE at an 
even more lenient standard than currently exists, nor did it 
consider making its proposed freeze take effect sooner than MY 2020. 
However, as DOT's own analysis strongly indicates, doing so would 
lead to even greater benefits and an even greater reduction in CAFE-
related deaths and injuries. In short, DOT's failure to consider 
this possibility is arbitrary and capricious. It has an opportunity 
to remedy this in its final rule, and it should do so by selecting a 
standard that is even more lenient than the one it proposed. . . . 
It should have gone beyond its original set of alternatives and 
examined less stringent ones as well--until it found one that, for 
some reason or another, failed to produce greater safety benefits or 
failed to meet the statutory factors.\272\
---------------------------------------------------------------------------

    \272\ CEI, NHTSA-2018-0067-12015, at 1.

    On the other hand, a coalition of ten environmental advocacy 
organizations stated that the agencies should consider alternatives 
more stringent than those defining the baseline no action alternative, 
arguing that in light of CEQ guidance and the 2018 IPCC report on 
climate change, ``the increasing danger, increasing urgency, and 
increasing importance of vehicle emissions all rationally counsel for 
strengthening emission standards.'' \273\ CBD et al. observe that 
``none of these alternatives [considered in the NPRM] increases fuel 
economy in comparison with the No Action Alternative, none conserves 
energy . . .'' and go on to assert that ``none represents maximum 
feasible CAFE standards.'' \274\ Similarly, EDF commented that ``. . . 
given its clear statutory directive to maximize fuel savings, NHTSA 
should have considered a range of alternatives that would be more 
protective than the existing standards,'' \275\ and three State 
agencies in Minnesota commented that ``more stringent standards are 
consistent with EPCA's purpose of energy conservation and the CAA's 
purpose of reducing harmful air pollutants.'' \276\ The North Carolina 
Department of Environmental Quality acknowledged the agencies' 
determination in the proposal that alternatives beyond the augural 
standards might be economically impracticable, but nevertheless argued 
that ``alternatives that exceed the stringency of the current standards 
are consistent with EPCA's purpose'' \277\ In oral testimony before the 
agencies, the New York State Attorney General also indicated that the 
agencies should consider alternatives more stringent than the augural 
standards.\278\ A coalition of States and cities commented that ``at a 
minimum, the existing standards should be left in place, but EPA should 
also consider whether to make the standards more stringent, not less, 
just as it has done in prior proposals.'' \279\ More specifically, 
through International Mosaic, some individuals commented that the 
agencies must ``fully and publicly consider a few options that require 
at least a seven annual percent [sic] improvement in vehicle fleet 
mileage.'' \280\ In comments on the DEIS, CBD, et al. went further, 
commenting that ``NHTSA's most stringent alternative must be set at no 
lower than a 9 percent improvement per year.'' \281\ Most manufacturers 
who commented on stringency did not identify specific regulatory 
alternatives that the agencies should consider, although Honda 
suggested that standards be set to increase in stringency at 5 percent 
annually for both passenger cars and light trucks throughout model 
years 2021-2026.282 283
---------------------------------------------------------------------------

    \273\ CBD, et al., NHTSA-2018-0067-12057 p. 10. Also, see 
comments from Senator Tom Carper, NHTSA-2018-0067-11910, at 8-9, and 
from UCS, NHTSA-2018-0067-12039, at 3.
    \274\ CBD, et al., NHTSA-2018-0067-12123, at 12-13.
    \275\ EDF, NHTSA-2018-0067-11996, at 20.
    \276\ Minnesota Pollution Control Agency, Department of 
Transportation, and Department of Health, NHTSA-2018-0067-11706, at 
5.
    \277\ North Carolina Department of Environmental Quality, NHTSA-
2018-0067-12025, at 37-38.
    \278\ New York State Attorney General, Testimony of Austin 
Thompson, NHTSA-2018-0067-12305, at 13.
    \279\ NHTSA-2018-0067-11735, at 49.
    \280\ International Mosaic NHTSA-2018-0067-11154, at 1
    \281\ CBD, et al., NHTSA-2018-0067-12123, at 17.
    \282\ Honda, NHTSA-2018-0067-12019, EPA-HQ-OAR-2018-0283, at 54.
    \283\ In model year 2021, the baseline standards for passenger 
cars and light trucks increase by about 4% and 6.5%, respectively, 
relative to standards for model year 2020. Depending on the 
composition of the future new vehicle fleet (i.e., the footprints 
and relative market shares of passenger cars and light trucks), this 
amounts to an overall average stringency increase of about 5.5% 
relative to model year 2020.
---------------------------------------------------------------------------

    The agencies carefully considered these comments to expand the 
range of stringencies to be evaluated as possible candidates for 
promulgation. To inform this consideration, the agencies used the CAFE 
model to examine a progression of stringencies extending outside the 
range presented in the proposal and draft EIS, and as a point of 
reference, using a case that reverts to MY 2018 standards starting in 
MY 2021. Scenarios included in this initial screening exercise ranged 
as high as increasing annually at 9.5 percent during MYs 2021-2026, 
reaching average CAFE and CO2 requirements of 66 mpg and 120 
g/mi, respectively. Results of this analysis are presented in the 
following tables and charts. Focusing on MY 2029, the tables show 
average required and achieved CAFE (as mpg) and CO2 (as g/
mi) levels for each scenario, along with average per-vehicle costs (in 
2018 dollars, relative to retaining MY 2017 technologies). The proposed 
(0%/0%), final (1.5%/1.5%), and baseline augural standards are shown in 
bold type. The charts present

[[Page 24259]]

the same results on a percentage basis, relative to values shown below 
for the scenario that reverts to MY 2018 standards starting in MY 2021.
    For example, reverting to the MY 2018 CAFE standards starting in MY 
2021 yields an average CAFE requirement of 35 mpg by MY 2029, with the 
industry exceeding that standard by 5 mpg at an average cost of $1,255 
relative to MY 2017 technology. Under the augural standards, the MY 
2029 requirement increases to 47 mpg, the average compliance margin 
falls to 1 mpg, and the average cost increases to $2,770. In other 
words, compared to the scenario that reverts to MY 2018 stringency 
starting in MY 2021, the augural standards increase stringency by 34 
percent (from 35 to 47 mpg), increase average fuel economy by 20 
percent (from 40 to 48 mpg), and increase costs by 121 percent (from 
$1,255 to $2,770).
    As indicated in the following two charts, the reality of 
diminishing returns clearly applies in both directions. On one hand, 
relaxing stringency below the proposed standards by reverting to MY 
2018 or MY 2019 standards reduces average MY 2029 costs by only modest 
amounts ($54-$121). As discussed in Section VIII, the agencies' updated 
analysis indicates that the proposed standards would not be maximum 
feasible considering the EPCA/EISA statutory factors, and would not be 
appropriate under the CAA after considering the appropriate factors. If 
further relaxation of standards appeared likely to yield more 
significant cost reductions, it is conceivable that such savings could 
outweigh further foregoing of energy and climate benefits. However, 
this screening analysis does not show dramatic cost reductions. 
Therefore, the agencies did not include these two less stringent 
alternatives in the detailed analysis presented in Section VII.
    On the other hand, increases in stringency beyond the baseline 
augural standards show relative costs continuing to accrue much more 
rapidly than relative CAFE and CO2 improvements. As 
discussed below in Section VIII, even the no action alternative is 
already well beyond levels that can be supported under the CAA and 
EPCA. If further stringency increases appeared likely to yield more 
significant additional energy and environmental benefits, it is 
conceivable that these could outweigh these significant additional cost 
increases. However, this screening analysis shows no dramatic relative 
acceleration of energy and environmental benefits. Therefore, the 
agencies did not include stringencies beyond the augural standards in 
the detailed analysis presented in Section VII.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.062


[[Page 24260]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.063

[GRAPHIC] [TIFF OMITTED] TR30AP20.064


[[Page 24261]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.065

BILLING CODE 4910-59-C
    Specific to model year 2021, some commenters argued that EPCA's 
lead time requirement prohibits NHTSA from revising CAFE standards for 
model year 2021.\284\ Regarding the revision of standards for model 
year 2021, NHTSA did consider EPCA's lead time requirement, and 
determined that while the agency would need to finalize a stringency 
increase at least 18 months before the beginning of the first affected 
model year, the agency can finalize a stringency decrease closer (or 
even after) the beginning of the first affected model year. The 
agency's reasoning is explained further in Section VIII. Therefore, 
NHTSA did not change regulatory alternatives to avoid any relaxation of 
stringency in model year 2021.
---------------------------------------------------------------------------

    \284\ State of California, et al., NHTSA-2018-0067-11735, at 
78.; CBD, et al., NHTSA-2018-0067-12000, Appendix A, at 66.; 
National Coalition for Advanced Transportation, NHTSA-2018-0067-
11969, at 46.
---------------------------------------------------------------------------

    The Auto Alliance stated that ``the truck increase rate should be 
no greater than the car rate of increase and should be the `equivalent 
task' per fleet.'' \285\ Supporting these Alliance comments, FCA 
elaborated by commenting that ``(1) in MY2017, the latest data we have 
available, most trucks have a larger gap to standards than cars, and 
(2) all of the truck segments are challenged because consumers are 
placing a greater emphasis on capability than fuel economy.'' \286\ 
Similarly, Ford commented that ``. . . the rates of increase in the 
stringency of the standards should remain equivalent between passenger 
cars and light duty trucks.'' \287\ Other commenters expressed general 
support for equalizing the rates at which the stringencies of passenger 
car and light truck standards increase.\288\
---------------------------------------------------------------------------

    \285\ Alliance, NHTSA-2018-0067-12073, at 7-8
    \286\ FCA, NHTSA-2018-0067-11943, at 46-47.
    \287\ Ford, NHTSA-2018-0067-11928, at 3.
    \288\ See, e.g., Global, NHTSA-2018-0067-12032, at 4; NADA, 
NHTSA-2018-0067-12064, at 13; BorgWarner, NHTSA-2018-0067-11895, at 
6.
---------------------------------------------------------------------------

    For the final rule, the agencies have added an alternative in which 
stringency for both cars and trucks increases at 1.5 percent. This is 
consistent with comments received requesting that both fleets' 
standards increase in stringency by the same amount, and 1.5 percent 
represents a rate of increase within the range of rates of increase 
considered in the NPRM.
    Throughout the NPRM, the agencies described their consideration as 
covering a range of alternatives.\289\ The preferred alternative for 
this final rule, an increase in stringency of 1.5 percent for both cars 
and trucks, falls squarely

[[Page 24262]]

within the range of alternatives proposed by the agencies.
---------------------------------------------------------------------------

    \289\ 83 FR at 42986 (Aug. 24, 2018) (explaining, in ``Summary'' 
section of NPRM, that ``comment is sought on a range of alternatives 
discussed throughout this document''); id. at 42988 (stating that 
the agencies are ``taking comment on a wide range of alternatives, 
including different stringencies and retaining existing 
CO2 standards and the augural CAFE standards''); 42990 
(``As explained above, the agencies are taking comment on a wide 
range of alternatives and have specifically modeled eight 
alternatives (including the proposed alternative) and the current 
requirements (i.e., baseline/no action).''); 43197 (``[T]oday's 
notice also presents the results of analysis estimating impacts 
under a range of other regulatory alternatives the agencies are 
considering.''); 43229 (explaining that ``technology availability, 
development and application, if it were considered in isolation, is 
not necessarily a limiting factor in the Administrator's selection 
of which standards are appropriate within the range of the 
Alternatives presented in this proposal.''); 43369 (``As discussed 
above, a range of regulatory alternatives are being considered.'').
---------------------------------------------------------------------------

    The NPRM alternatives were bounded on the upper end by the 
baseline/no action alternative, and the proposed alternative on the 
lower end (0 percent per year increase in stringency for both cars and 
trucks). For passenger cars, the agencies considered a range of 
stringency increases between 0 percent and 2 percent per year for 
passenger cars, in addition to the baseline/no action alternative. For 
light trucks, the agencies considered a range of stringency increases 
between 0 percent and 3 percent per year, in addition to the baseline/
no action alternative.
    The agencies considered the same range of alternatives for this 
final rule. As with the proposal, the alternatives for stringency are 
bounded on the upper end by the baseline/no action alternative and on 
the lower end by 0 percent per year increases for both passenger cars 
and light trucks. Consistent with the proposal, for this final rule, 
the agencies considered stringency increases of between 0 and 2 percent 
per year for passenger cars and between 0 and 3 percent per year for 
light trucks, in addition to the baseline/no action alternative.
    While it was not specifically modeled in the NPRM, the new 
preferred alternative of an increase in stringency of 1.5 percent for 
both cars and trucks was well within the range of alternatives 
considered. The proposal described the alternatives specifically 
modeled as options for the agencies, but also gave notice that they did 
not limit the agencies in selecting from among the range of 
alternatives under consideration.\290\
---------------------------------------------------------------------------

    \290\ See, e.g., 83 FR at 43003 (Aug. 24, 2018) (``These 
alternatives were examined because they will be considered as 
options for the final rule. The agencies seek comment on these 
alternatives, seek any relevant data and information, and will 
review responses. That review could lead to the selection of one of 
the other regulatory alternatives for the final rule or some 
combination of the other regulatory alternatives (e.g., combining 
passenger cars standards from one alternative with light truck 
standards from a different alternative).''); id. at 43229 
(describing a factor relevant to ``the Administrator's selection of 
which standards are appropriate within the range of the Alternatives 
presented in this proposal'').
---------------------------------------------------------------------------

    The agencies explained in the proposal that they were ``taking 
comment on a wide range of alternatives and have specifically modeled 
eight alternatives.'' \291\ As with the proposal, for the final rule, 
the agencies specifically modeled the upper and lower bounds of the 
baseline/no action alternative and 0 percent per year stringency 
increases for both passenger cars and light trucks. In both the 
proposal and the final rule, the agencies also modeled a stringency 
increase of 2 percent per year for passenger cars and 3 percent per 
year for light trucks, as well as a variety of other specific increases 
between 0 and 2 percent for passenger cars and 0 and 3 percent for 
light trucks.
---------------------------------------------------------------------------

    \291\ 83 FR at 42990 (Aug. 24, 2018).
---------------------------------------------------------------------------

    The specific alternatives the agencies modeled for the final rule 
reflect their consideration of public comments. As discussed above, 
multiple commenters expressed support for equalizing the rates at which 
the stringencies of passenger car and light truck standards increase. 
To help the agencies evaluate alternatives that include the same 
stringency increase for passenger cars and light trucks, three of the 
seven alternatives (in addition to the baseline/no action alternative) 
that the agencies specifically modeled for the final rule included the 
same stringency increase for passenger cars and light trucks. This 
includes the new preferred alternative of an increase in stringency of 
1.5 percent for both cars and trucks. This alternative, and all others 
specifically modeled for the final rule, falls within the range of 
alternatives for stringency considered by the agencies in the proposal.
    Beyond these stringency provisions discussed in the NPRM, the 
agencies also sought comment on a number of additional compliance 
flexibilities for the programs, as discussed in Section IX.
2. Additional Alternatives Suggested by Commenters
    Beyond the comments discussed above regarding the shapes of the 
functions defining fuel economy and CO2 targets, regarding 
the inclusion of non-CO2 emissions, and regarding the 
stringencies to be considered, the agencies also received a range of 
other comments regarding regulatory alternatives.
    Some of these additional comments involved how CAFE and 
CO2 standards compare to one another for any given 
regulatory alternative. With a view toward maximizing harmonization of 
the standards, the Alliance, supported by some of its members' 
individual comments, indicated that ``to the degree flexibilities and 
incentives are not completely aligned between the CAFE and 
[CO2] programs, there must be an offset in the associated 
footprint-based targets to account for those differences. Some areas of 
particular concerns are air conditioning refrigerant credits, and 
incentives for advanced technology vehicles. The Alliance urges the 
Agencies to seek harmonization of the standards and flexibilities to 
the greatest extent possible. . . .'' \292\
---------------------------------------------------------------------------

    \292\ Alliance, NHTSA-2018-0067-12073, at 40. See also FCA, 
NHTSA-2018-0067-11943, at 6-7.
---------------------------------------------------------------------------

    On the other hand, discussing consideration of compliance credits 
but making a more general argument, the NYU Institute for Policy 
Integrity commented that ``. . . EPA is not allowed to set lower 
standards just for the sake of harmonization; to the contrary, full 
harmonization may be inconsistent with EPA's statutory 
responsibilities.'' \293\ Similarly, ACEEE argued that ``any 
consideration of an extension or expansion of credit provisions under 
the [carbon dioxide] or CAFE standards program should take as a 
starting point the assumption that the additional credits will allow 
the stringency of the standards to be increased.'' \294\
---------------------------------------------------------------------------

    \293\ IPI, NHTSA-2018-0067-12213, at 21.
    \294\ ACEEE, NHTSA-2018-0067-12122, at 3.
---------------------------------------------------------------------------

    EPCA's requirement that NHTSA set standards at the maximum feasible 
levels is separate and ``wholly independent'' from the CAA's 
requirement, per Massachusetts v. EPA, that EPA issue regulations 
addressing pollutants that EPA has determined endanger public health 
and welfare.\295\ Nonetheless, as recognized by the Supreme Court, 
``there is no reason to think the two agencies cannot both administer 
their obligations and yet avoid inconsistency.'' \296\ This conclusion 
was reached despite the fact that EPCA has a range of very specific 
requirements about how CAFE standards are to be structured, how 
manufacturers are to comply, what happens when manufacturers are unable 
to comply, and how NHTSA is to approach setting standards, and despite 
the fact that the CAA has virtually no such requirements. This means 
that while nothing about either EPCA or the CAA, much less the 
combination of the two, guarantees ``harmonization'' defining ``One 
National Program,'' the agencies are expected to be able to work out 
the differences.
---------------------------------------------------------------------------

    \295\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007).
    \296\ Id.
---------------------------------------------------------------------------

    Since tailpipe CO2 standards are de facto fuel economy 
standards, the more differences there are between CO2 and 
CAFE standards and compliance provisions, the more challenging it is 
for manufacturers to plan year-by-year production that responses to 
both, and the more difficult it is for affected stakeholders and the 
general public to understand regulation in this space. Therefore, even 
if the two statutes, taken together, do not guarantee ``full 
harmonization,'' steps toward greater

[[Page 24263]]

harmonization help with compliance planning and transparency--and meet 
the expectations set forth by the Supreme Court that the agencies avoid 
inconsistencies.
    The agencies have taken important steps toward doing so. For 
example, EPA has adopted separate footprint-based CO2 
standards for passenger cars and light trucks, and has redefined CAFE 
calculation procedures to introduce recognition for the application of 
real-world fuel-saving technology that is not captured with traditional 
EPA two-cycle compliance testing. Detailed aspects of both sets of 
standards and corresponding compliance provisions are discussed at 
length in Section IX. The agencies never set out with the primary goal 
of achieving ``full harmonization,'' such that both sets of standards 
would lead each manufacturer to respond in exactly the same way in 
every model year.\297\ For example, EPA did not adopt the EPCA 
requirement that domestic passenger car fleets each meet a minimum 
standard, or the EPCA cap on compliance credit transfers between 
passenger car fleets. On the other hand, EPA also did not adopt the 
EPCA civil penalty provisions that have allowed some manufacturers to 
pay civil penalties as an alternative method of meeting EPCA 
obligations. These and other differences provide that even if CAFE and 
CO2 standards are ``mathematically'' harmonized, for any 
given manufacturer, the two sets of standards will not be identically 
burdensome in each model year. Inevitably, one standard will be more 
challenging than the other, varying over time, between manufacturers, 
and between fleets. This means manufacturers need to have compliance 
plans for both sets of standards.
---------------------------------------------------------------------------

    \297\ Full harmonization would mean that, for example, if Ford 
would do some set of things over time in response to CAFE standards 
in isolation, it would do exactly the same things on exactly the 
same schedule in response to CO2 standards in isolation.
---------------------------------------------------------------------------

    In 2012, recognizing that EPCA provides no clear basis to address 
HFC, CH4, or N2O emissions directly, the agencies 
``offset'' CO2 targets from fuel economy targets (after 
converting the latter to a CO2 basis) by the amounts of 
credit EPA anticipated manufacturers would, on average, earn in each 
model years by reducing A/C leakage and adopting refrigerants with 
reduced GWPs. In 2012, EPA assumed that by 2021, all manufacturers 
would be earning the maximum available credit, and EPA's analysis 
assumed that all manufacturers would make progress at the same rate. 
However, as discussed above, data highlighted in comments by Chemours, 
Inc., demonstrate that actual manufacturers' adoption of lower-GWP 
refrigerants thus far ranges widely, with some manufacturers (e.g., 
Nissan) having taken no such steps to move toward lower-GWP 
refrigerants, while others (e.g., JLR) have already applied lower-GWP 
refrigerants to all vehicles produced for sale in the U.S. Therefore, 
at least in practice, HFC provisions thus far continue to leave a gap 
(in terms of harmonization) between the two sets of standards. The 
proposal would have taken the additional step of decoupling provisions 
regarding HFC (i.e., A/C leakage credits), CH4, and 
N2O emissions from CO2 standards, addressing 
these in separate regulations to be issued in a new proposal. As 
discussed above, EPA did not finalize this proposal. Accordingly, for 
the regulatory alternatives considered today, EPA has reinstated 
offsets of CO2 targets from fuel economy targets, reflecting 
the assumption that all manufacturers will be earning the maximum 
available A/C leakage credit by MY 2021.
    In addition to general comments on harmonization, the agencies 
received a range of comments on specific provisions--especially 
involving ``flexibilities''--that may or may not impact harmonization. 
With a view toward encouraging further electrification, NCAT proposed 
that EPA extend indefinitely the exclusion of upstream emissions from 
electricity generation, and also extend and potentially restructure 
production multipliers for PHEVs, EVs, and FCVs.\298\ On the other 
hand, connecting its comments back to the stringency of standards, NCAT 
also commented that ``. . . expansion of compliance flexibilities in 
the absence of any requirement to improve [CO2] reduction or 
fuel economy (as under the agencies' preferred option) could result in 
an effective deterioration of existing [CO2] and fuel 
economy performance, as well as little or no effective support for 
advanced vehicle technology development or deployment.'' \299\ Global 
Automakers indicated that the final rule ``should include a package of 
programmatic elements that provide automakers with flexible compliance 
options that promote the full breadth of vehicle technologies,'' such 
options to include the extension of ``advanced technology'' production 
multipliers through MY 2026, the indefinite exclusion of emissions from 
electricity generation, the extension to passenger cars of credits 
currently granted for the application of ``game changing'' technologies 
(e.g., HEVs) only to full-size pickup trucks, an increase (to 15 g/mi) 
of the cap on credits for off-cycle technologies, an updated credit 
``menu'' of off-cycle technologies, and easier process for handling 
applications for off-cycle credits.\300\ The Alliance also called for 
expanded sales multipliers and a permanent exclusion of emissions from 
electricity generation.\301\ Walter Kreucher recommended the agencies 
consider finalizing the proposed standards but also keeping the augural 
standards as ``voluntary targets'' to ``provide compliance with the 
statutes and an aspirational goal for manufacturers.'' \302\
---------------------------------------------------------------------------

    \298\ NCAT, NHTSA-2018-0067-11969, at 3-5.
    \299\ Id.
    \300\ Global Automakers, NHTSA-2018-0067-12032, at 4 et seq.
    \301\ Alliance, NHTSA-2018-0067-12073, at 8.
    \302\ Kreucher, W., NHTSA-2018-0067-0444, at 9.
---------------------------------------------------------------------------

    The agencies have carefully considered these comments, and have 
determined that the current suite of ``flexibilities'' generally 
provide ample incentive more rapidly to develop and apply advanced 
technologies and technologies that produce fuel savings and/or 
CO2 reductions that would otherwise not count toward 
compliance. The agencies also share some stakeholders' concern that 
expanding these flexibilities could increase the risk of ``gaming'' 
that would make compliance less transparent and would unduly compromise 
energy and environmental benefits. Nevertheless, as discussed in 
Section IX, EPA is adopting new multiplier incentives for natural gas 
vehicles. EPA is also finalizing some changes to procedures for 
evaluating applications for off-cycle credits, and expects these 
changes to make this process more accurate and more efficient. Also, 
EPA is revising its regulations to not require manufacturers to account 
for upstream emissions associated with electricity use for electric 
vehicles and plug-in hybrid electric vehicles through model year 2026; 
compliance will instead be based on tailpipe emissions performance only 
and not include emissions from electricity generation until model year 
2027. As discussed below, even with this change, and even accounting 
for continued increases in fuel prices and reductions in battery 
prices, BEVs are projected in this final rule analysis to continue to 
account for less than 5 percent of new light vehicle sales in the U.S. 
through model year 2026. To the extent that this projection turns out 
to reflect reality, this means that the impact of upstream emissions 
from electricity use on the projected CO2

[[Page 24264]]

reductions associated with these standards would likely remain small. 
Regarding comments suggesting that the augural standards should be 
finalized as ``voluntary targets,'' the agencies have determined that 
having such targets exist alongside actual regulatory requirements 
would be, at best, unnecessary and confusing.
    Beyond these additional proposals, some commenters' proposals 
clearly fell outside authority provided under EPCA or the CAA. Ron 
Lindsay recommended the agencies ``consider postponing the rule changes 
until the U.S. can establish a legally binding national and 
international carbon budget and a binding mechanism to adhere to it.'' 
\303\ EPCA requires NHTSA to issue standards for MY 2022 by April 1, 
2020, and previously-issued EPA regulations commit EPA to revisiting MY 
2021-2025 standards on a similar schedule. These statutory and 
regulatory provisions do not include a basis to delay decisions pending 
an international negotiation for which prospects and schedules are both 
unknown.
---------------------------------------------------------------------------

    \303\ Ron Lindsay, EPA-HQ-OAR-2018-0283-1414, at 6.
---------------------------------------------------------------------------

    SCAQMD, supported by Shyam Shukla, indicated that the agencies 
should consider an alternative that keeps the waiver for California's 
CO2 standards in place.\304\ NCAT and the North Carolina DEQ 
offered similar comments and CBD, et al. commented that ``among the set 
of more stringent alternatives that NEPA requires the agency to 
consider, NHTSA must include action alternatives that retain the 
standards California and other states have lawfully adopted.'' \305\ As 
discussed above, the agencies recently issued a final rule addressing 
the issue of California's authority. NEPA does not require NHTSA to 
include action alternatives that cannot be lawfully realized.
---------------------------------------------------------------------------

    \304\ SCAQMD, NHTSA-2018-0067-5666, at 1-2; Shyam Shukla, NHTSA-
2018-0067-5793, at 1-2.
    \305\ NCAT, NHTSA-2018-0067-11969, at 64; NCDEQ, NHTSA-2018-
0067-12025, at 38; CBD et al., NHTSA-2018-0067-12123, Attachment 1, 
at 18.
---------------------------------------------------------------------------

    International Mosiac commented that NHTSA's DEIS ``is fatally 
flawed . . . because it does not consider any market-based alternatives 
(e.g., a `cap and trade' type option).'' \306\ While EPCA/EISA does 
include very specific provisions regarding trading of CAFE compliance 
credits, the statute provides no authority for a broad-based cap-and-
trade program involving other sectors. Similarly, Michalek, et al. 
wrote that ``a more economically efficient approach of, taxing 
emissions and fuel consumption at socially appropriate levels would 
allow households to determine whether to reduce fuel consumption and 
emissions by driving less, by buying a vehicle with more fuel saving 
technologies, or by buying a smaller vehicle--or, alternatively, not to 
reduce fuel consumption and emissions at all but rather pay a cost 
based on the damages they cause. Forcing improvements only through one 
mechanism (fuel-saving technologies) increases the cost of achieving 
these outcomes.'' \307\ While some economists would agree with these 
comments, Congress has provided no clear authority for NHTSA or EPA to 
implement either an emissions tax or a broad-based cap-and-trade 
program in which motor vehicles could participate.
---------------------------------------------------------------------------

    \306\ International Mosaic, NHTSA-2018-0067-11154, at 1-2.
    \307\ Michalek, et al., NHTSA-2018-0067-11903, at 13.
---------------------------------------------------------------------------

3. Details of Alternatives Considered in Final Rule
a) Alternative 1
    Alternative 1 holds the stringency of targets constant and MY 2020 
levels through MY 2026.

[[Page 24265]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.066

b) Alternative 2
    Alternative 2 increases the stringency of targets annually during 
MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by 
0.5 percent for passenger cars and 0.5 percent for light trucks.

[[Page 24266]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.067

c) Alternative 3
    Alternative 3; the final standards promulgated today, increases the 
stringency of targets annually during MYs 2021-2026 (on a gallon per 
mile basis, starting from MY 2020) by 1.5 percent for passenger cars 
and 1.5 percent for light trucks.

[[Page 24267]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.068

d) Alternative 4
    Alternative 4 increases the stringency of targets annually during 
MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by 
1.0 percent for passenger cars and 2.0 percent for light trucks.

[[Page 24268]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.069

e) Alternative 5
    Alternative 5 increases the stringency of targets annually during 
MYs 2022-2026 (on a gallon per mile basis, starting from MY 2021) by 
1.0 percent for passenger cars and 2.0 percent for light trucks.
[GRAPHIC] [TIFF OMITTED] TR30AP20.070


[[Page 24269]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.071

f) Alternative 6
    Alternative 6 increases the stringency of targets annually during 
MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by 
2.0 percent for passenger cars and 3.0 percent for light trucks.
[GRAPHIC] [TIFF OMITTED] TR30AP20.072

[GRAPHIC] [TIFF OMITTED] TR30AP20.073


[[Page 24270]]


g) Alternative 7
    Alternative 7 increases the stringency of targets annually during 
MYs 2022-2026 (on a gallon per mile basis, starting from MY 2021) by 
2.0 percent for passenger cars and 3.0 percent for light trucks.
[GRAPHIC] [TIFF OMITTED] TR30AP20.074

[GRAPHIC] [TIFF OMITTED] TR30AP20.075

    EPCA, as amended by EISA, requires that any manufacturer's 
domestically-manufactured passenger car fleet must meet the greater of 
either 27.5 mpg on average, or 92 percent of the average fuel economy 
projected by the Secretary for the combined domestic and non-domestic 
passenger automobile fleets manufactured for sale in the U.S. by all 
manufacturers in the model year, which projection shall be published in 
the Federal Register when the standard for that model year is 
promulgated in accordance with 49 U.S.C. 32902(b).\308\ Any time NHTSA 
establishes or changes a passenger car standard for a model year, the 
MDPCS for that model year must also be evaluated or re-evaluated and 
established accordingly. Thus, this final rule establishes the 
applicable MDPCS for MYs 2021-2026. Table V-22 lists the minimum 
domestic passenger car standards.
---------------------------------------------------------------------------

    \308\ 49 U.S.C. 32902(b)(4).
    [GRAPHIC] [TIFF OMITTED] TR30AP20.076
    

[[Page 24271]]



VI. Analytical Approach as Applied to Regulatory Alternatives

A. Overview of Methods

    Like analyses accompanying the NPRM and past CAFE and CAFE/
CO2 rulemakings, the analysis supporting today's notice 
spans a range of technical topics, uses a range of different types of 
data and estimates, and applies several different types of computer 
models. The purpose of the analysis is not to determine the standards, 
but rather to provide information for consideration in doing so. The 
analysis aims to answer the question ``what impacts might each of these 
regulatory alternatives have?''
    Over time, NHTSA's and, more recently, NHTSA's and EPA's analyses 
have expanded to address an increasingly wide range of types of 
impacts. Today's analysis involves, among other things, estimating how 
the application of various combinations of technologies could impact 
vehicles' costs and fuel economy levels (and CO2 emission 
rates), estimating how vehicle manufacturers might respond to standards 
by adding fuel-saving technologies to new vehicles, estimating how 
changes in new vehicles might impact vehicle sales and operation, and 
estimating how the combination of these changes might impact national-
scale energy consumption, emissions, highway safety, and public health. 
In addition, the EIS accompanying today's notice addresses impacts on 
air quality and climate. The analysis of these factors informs and 
supports both NHTSA's application of the statutory requirements 
governing the setting of ``maximum feasible'' fuel-economy standards 
under EPCA, including, among others, technological feasibility and 
economic practicability, and EPA's application of the CAA requirements 
for tailpipe emissions.
    Supporting today's analysis, the agencies have brought to bear a 
variety of different types of data, a few examples of which include 
fuel economy compliance reports, historical sales and average 
characteristics of light-duty vehicles, historical economic and 
demographic measures, historical travel demand and energy prices and 
consumption, and historical measures of highway safety. Also supporting 
today's analysis, the agencies have applied several different types of 
estimates, a few examples of which include projections of the future 
cost of different fuel-saving technologies, projections of future GDP 
and the number of households, estimates of the ``gap'' between 
``laboratory'' and on-road fuel economy, and estimates of the social 
cost of CO2 emissions and petroleum ``price shocks.''
    With a view toward transparency, repeatability, and efficiency, the 
agencies have used a variety of computer models to conduct the majority 
of today's analysis. For example, the agencies have applied DOE/EIA's 
National Energy Modeling System (NEMS) to estimate future energy 
prices, EPA's MOVES model to estimate tailpipe emission rates for ozone 
precursors and other criteria pollutants, DOE/Argonne's GREET model to 
estimate emission rates for ``upstream'' processes (e.g., petroleum 
refining), and DOE/Argonne's Autonomie simulation tool to estimate the 
fuel consumption impacts of different potential combinations of fuel-
saving technology. In addition, the EIS accompanying today's notice 
applies photochemical models to estimate air quality impacts, and 
applies climate models to estimate climate impacts of overall emissions 
changes.
    Use of these different types of data, estimates, and models is 
discussed further below in the most closely relevant sections. For 
example, the agencies' use of NEMS is discussed below in the portion of 
Section VI that addresses the macroeconomic context, which includes 
fuel prices, and the agencies use of Autonomie is discussed in the 
portion of Section VI.B.3 that addresses the agencies' approach to 
estimating the effectiveness of various technologies (in reducing fuel 
consumption and CO2 emissions).
    Providing an integrated means to estimate both vehicle 
manufacturers' potential responses to CAFE or CO2 standards 
and, in turn, many of the different potential direct results (e.g., 
changes in new vehicle costs) and indirect impacts (e.g., changes in 
rates of fleet turnover) of those responses, the CAFE Model plays a 
central role in the agencies' analysis supporting today's notice. The 
agencies used the specific models mentioned above to develop inputs to 
the CAFE model, such as fuel prices and emission factors. Outputs from 
the CAFE Model are discussed in Sections VII and VIII of today's 
notice, and in the accompanying RIA. The EIS accompanying today's 
notice makes use of the CAFE Model's estimates of changes in total 
emissions from light-duty vehicles, as well as corresponding changes in 
upstream emissions. These changes in emissions are included in the set 
of inputs to the models used to estimate air quality and climate 
impacts.
    The remainder of this overview focuses on the CAFE Model. The 
purpose of this overview is not to provide a comprehensive technical 
description of the model,\309\ but rather to give an overview of the 
model's functions, to explain some specific aspects not addressed 
elsewhere in today's notice, and to discuss some model aspects that 
were the subject of significant public comment. Some model functions 
and related comments are addressed in other parts of today's notice. 
For example, the model's handling of Autonomie-based fuel consumption 
estimates is addressed in the portion of Section VI.B.3 that discusses 
the agencies' application of Autonomie. The model documentation 
accompanying today's notice provides a comprehensive and detailed 
description of the model's functions, design, inputs, and outputs.
---------------------------------------------------------------------------

    \309\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting 
today's notice.
---------------------------------------------------------------------------

1. Overview of CAFE Model
    The basic design of the CAFE Model is as follows: The system first 
estimates how vehicle manufacturers might respond to a given regulatory 
scenario, and from that potential compliance solution, the system 
estimates what impact that response will have on fuel consumption, 
emissions, and economic externalities. A regulatory scenario involves 
specification of the form, or shape, of the standards (e.g., flat 
standards, or linear or logistic attribute-based standards), scope of 
passenger car and truck regulatory classes, and stringency of the CAFE 
and CO2 standards for each model year to be analyzed.
    Manufacturer compliance simulation and the ensuing effects 
estimation, collectively referred to as compliance modeling, encompass 
numerous subsidiary elements. Compliance simulation begins with a 
detailed user-provided initial forecast of the vehicle models offered 
for sale during the simulation period. The compliance simulation then 
attempts to bring each manufacturer into compliance with the standards 
defined by the regulatory scenario contained within an input file 
developed by the user. For example, a regulatory scenario may define 
CAFE or CO2 standards that increase in stringency by 4 
percent per year for 5 consecutive years.
    The model applies various technologies to different vehicle models 
in each manufacturer's product line to simulate how each manufacturer 
might make progress toward compliance with the specified standard. 
Subject to a variety of user-controlled constraints, the model applies 
technologies based on

[[Page 24272]]

their relative cost-effectiveness, as determined by several input 
assumptions regarding the cost and effectiveness of each technology, 
the cost of compliance (determined by the change in CAFE or 
CO2 credits, CAFE-related civil penalties, or value of 
CO2 credits, depending on the compliance program being 
evaluated and the effective-cost mode in use), and the value of avoided 
fuel expenses. For a given manufacturer, the compliance simulation 
algorithm applies technologies either until the manufacturer runs out 
of cost-effective technologies, until the manufacturer exhausts all 
available technologies, or, if the manufacturer is assumed to be 
willing to pay civil penalties, until paying civil penalties becomes 
more cost-effective than increasing vehicle fuel economy. At this 
stage, the system assigns an incurred technology cost and updated fuel 
economy to each vehicle model, as well as any civil penalties incurred 
by each manufacturer. This compliance simulation process is repeated 
for each model year available during the study period.
    This point marks the system's transition between compliance 
simulation and effects calculations. At the conclusion of the 
compliance simulation for a given regulatory scenario, the system 
contains multiple copies of the updated fleet of vehicles corresponding 
to each model year analyzed. For each model year, the vehicles' 
attributes, such as fuel types (e.g., diesel, electricity), fuel 
economy values, and curb weights have all been updated to reflect the 
application of technologies in response to standards throughout the 
study period. For each vehicle model in each of the model year specific 
fleets, the system then estimates the following: Lifetime travel, fuel 
consumption, carbon dioxide and criteria pollutant emissions, the 
magnitude of various economic externalities related to vehicular travel 
(e.g., noise), and energy consumption (e.g., the economic costs of 
short-term increases in petroleum prices). The system then aggregates 
model-specific results to produce an overall representation of modeling 
effects for the entire industry.
    Different categorization schemes are relevant to different types of 
effects. For example, while a fully disaggregated fleet is retained for 
purposes of compliance simulation, vehicles are grouped by type of fuel 
and regulatory class for the energy, carbon dioxide, criteria 
pollutant, and safety calculations. Therefore, the system uses model-
by-model categorization and accounting when calculating most effects, 
and aggregates results only as required for efficient reporting.
2. Representation of the Market
    As a starting point, the model needs enough information to 
represent each manufacturer covered by the program. As discussed below 
in Section VI.B.1, the MY 2017 analysis fleet contains information 
about each manufacturer's:
     Vehicle models offered for sale--their current (i.e., MY 
2017) production volumes, manufacturer suggested retail prices (MSRPs), 
fuel saving technology content and other attributes (curb weight, drive 
type, assignment to technology class and regulatory class);
     Production considerations--product cadence of vehicle 
models (i.e., schedule of model redesigns and ``freshenings''), vehicle 
platform membership, degree of engine and/or transmission sharing (for 
each model variant) with other vehicles in the fleet; and
     Compliance constraints and flexibilities--preference for 
full compliance or penalty payment/credit application, willingness to 
apply additional cost-effective fuel saving technology in excess of 
regulatory requirements, projected applicable flexible fuel credits, 
and current credit balance (by model year and regulatory class) in 
first model year of simulation.
Representation of Fuel-Saving Technologies
    The modeling system defines technology pathways for grouping and 
establishing a logical progression of technologies that can be applied 
to a vehicle. Technologies that share similar characteristics form 
cohorts that can be represented and interpreted within the CAFE Model 
as discrete entities. The following Table VI-1 shows the technologies 
available within the modeling system used for this final rule. Each 
technology is discussed in detail below. However, an understanding of 
the technologies considered and how they are defined in the model 
(e.g., a 6-speed manual transmission is defined as ``MT6'') is helpful 
for the following explanation of the compliance simulation and the 
inputs required for that simulation.
BILLING CODE 4910-59-P

[[Page 24273]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.077


[[Page 24274]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.078

BILLING CODE 4910-59-C
    These entities are then laid out into pathways (or paths), which 
the system uses to define relations of mutual exclusivity between 
conflicting sets of technologies. For example, as presented in the next 
section, technologies on the Turbo Engine path are incompatible with 
those on the HCR Engine or the Diesel Engine paths. As such, whenever a 
vehicle uses a technology from one pathway (e.g., turbo), the modeling 
system immediately disables the incompatible technologies from one or 
more of the other pathways (e.g., HCR and diesel).
    In addition, each path designates the direction in which vehicles 
are allowed to advance as the modeling system evaluates specific 
technologies for application. Enforcing this directionality within the 
model ensures that a vehicle that uses a more advanced or more 
efficient technology (e.g., AT8) is not allowed to ``downgrade'' to a 
less efficient option (e.g., AT5). Visually, as portrayed in the charts 
in the sections that follow, this is represented by an arrow leading 
from a preceding technology to a succeeding one, where vehicles begin 
at the root of each path, and traverse to each successor technology in 
the direction of the arrows.
    The modeling system incorporates twenty technology pathways for 
evaluation as shown below. Similar to individual technologies, each 
path carries an intrinsic application level that denotes the scope of 
applicability of all technologies present within that path, and whether 
the pathway is evaluated on one vehicle at a time, or on a collection 
of vehicles that share a common platform, engine, or transmission.

[[Page 24275]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.079

    Even though technology pathways outline a logical progression 
between related technologies, all technologies available to the system 
are evaluated concurrently and independently of each other. Once all 
technologies have been examined, the model selects a solution deemed to 
be most cost-effective for application on a vehicle. If the modeling 
system applies a technology that resides later in the pathway, it will 
subsequently disable all preceding technologies from further 
consideration to prevent a vehicle from potentially downgrading to a 
less advanced option. Consequently, the system skips any technology 
that is already present on a vehicle (either those that were available 
on a vehicle from the input fleet or those that were previously applied 
by the model). This ``parallel technology'' approach, unlike the 
``parallel path'' methodology utilized in the preceding versions of the 
model, allows the system always to consider the entire set of available 
technologies instead of foregoing the application of potentially more 
cost-effective options that happen to reside further down the 
pathway.\310\ This revised approach addresses comments summarized 
below, and allows the system to analyze all available technology 
options concurrently and independently of one other without having to 
first apply one or more ``predecessor'' technologies. For example, if 
model inputs are such that a 7-speed transmission is cost-effective, 
but not as cost-effective as an 8-speed transmission, the revised 
approach enables the model to skip over the 7-speed transmission 
entirely, whereas the NPRM version of the model might first apply the 
7-speed transmission and then consider whether to proceed immediately 
to the 8-speed transmission. As such, the model's choices for 
evaluation of new technology solutions becomes slightly less 
restrictive, allowing it immediately to consider and apply more 
advanced options, and increasing the likelihood that the a globally 
optimum solution is selected.
---------------------------------------------------------------------------

    \310\ Previous versions of the CAFE Model followed a ``low-
cost'' first approach where the system would stop evaluating 
technologies residing within a given pathway as soon as the first 
cost-effective option within that path was reached.
---------------------------------------------------------------------------

    Some commenters supported the agencies' use of such pathways in the 
simulation of manufacturers' potential application of technologies. As 
one of a dozen examples of CAFE model design elements that lead to the 
transparent representation of real-world factors, the Alliance 
highlighted ``recognition of the need for manufacturers to follow 
`technology' pathways that retain capital and implementation expertise, 
such as specializing in one type of engine or transmission instead of 
following an unconstrained optimization that would cause manufacturers 
to leap to unrelated technologies and show overly optimistic costs and 
benefits.'' \311\ Similarly, Toyota commented that ``the inertia of 
capital investments and engineering expertise dedicated to one 
compliance technology or set of technologies makes it unreasonable for 
manufacturers to immediately switch to another technology path.'' \312\
---------------------------------------------------------------------------

    \311\ Alliance, NHTSA-2018-0067-12073, at 9.
    \312\ Toyota, NHTSA-2018-0067-12098, at 7.
---------------------------------------------------------------------------

    Other commenters cited the use of technology pathways as inherently 
overly restrictive. For example, as an example of ``arbitrary model 
constraints,'' a coalition of commenters cited the fact the model 
``prohibit[s] manufacturers from switching vehicle technology 
pathways.'' \313\ Also, EDF, UCS, and CARB cited the combination of 
technology pathways, decision making criteria, and model inputs as 
producing unrealistic results.\314\ Regarding the technology pathways, 
specifically, EDF's consultant argued that the technology paths are not

[[Page 24276]]

transparent, and cited the potential that specific paths may not 
necessarily be arranged in progression from least to most cost-
effective--that ``NHTSA ignores the cost of the technology when 
developing this list.'' \315\ Relatedly, as EDF's consultant commented:
---------------------------------------------------------------------------

    \313\ CBD, et al., NHTSA-2018-0067-12057, at 3.
    \314\ EDF, NHTSA-2018-0067-12108, Appendix A, at 57 et seq.; 
UCS, NHTSA-2018-0067-12039, Appendix, at 25 et seq.; Roush 
Industries, NHTSA-2018-0067-11984, at 5.
    \315\ EDF, NHTSA-2018-0067-12108, Appendix B, at 69.

    [T]he Volpe Model is not designed to look backwards along its 
technology paths. Thus, the opportunity to recover the expenditure 
of inefficient technology is missed. NHTSA might argue that a 
manufacturer will not invest in 10-speed transmissions, for example, 
and then return to an older design. Whether or not this is true in 
real life, such a view would put too much stake in the Volpe Model 
projections. The model simply projects what could be done, not what 
will be. Anyone examining the progression of technology and noting 
the reversion of transmission technology could easily modify the 
model inputs to avoid this. Also, if NHTSA evaluated combinations of 
technologies prior to entering them in the model piecemeal, it would 
automatically avoid such apparent problems.\316\
---------------------------------------------------------------------------

    \316\ Ibid., at 70.

    The agencies also received additional public comments on specific 
paths and specific interactions between paths (e.g., involving engines 
and hybridization). These comments are addressed below.
    The agencies have carefully considered these comments and the 
approach summarized below reflects some corresponding revision. As 
mentioned above, the CAFE model now approaches the technology paths in 
a such way that, faced with two cost-effective technologies on the same 
path, the model can proceed directly to the more advanced technology if 
that technology is the more cost effective of the two.
    However, the agencies reject assertions that the model's use of 
technology paths is not transparent. The agencies provided extensive 
explanatory text, figures, model documentation, and model source code 
specifically addressing these paths (and other model features). This 
transparency appears evident in that commenters (sometimes while 
claiming that a specific feature of the model is not transparent) 
presented analytical results involving changes to corresponding inputs 
that required a detailed understanding of that feature's operation.
    Regarding comments that the technology paths should be arranged in 
order of cost-effectiveness, the agencies note that such comments 
presume, without merit, that costs, fuel consumption impacts, and other 
inputs (e.g., fuel prices) that logically impact manufacturers' 
decision-making are not subject to uncertainty. These inputs are all 
subject to uncertainty, and the CAFE Model's arrangement of 
technologies into several paths is responsive to these uncertainties. 
Nevertheless, the agencies maintain that some technologies do reflect a 
higher level of advancement than others (e.g., 10-speed transmissions 
vs. 5-speed transmissions), and while manufacturers may, in practice, 
occasionally revert to less advanced technologies, it is appropriate 
and reasonable to conduct the agencies' analysis in a manner that 
assumes manufacturers will continue to make forward progress. As 
observed by EDF's consultant's remarks, the CAFE Model ``simply 
projects what could be done, not what will be.'' While no model, much 
less any model relying on information that can be made publicly 
available, can hope to represent precisely each manufacturers' actual 
detailed constrains related to product development and planning, such 
constraints are real and important. The agencies agree that the CAFE 
Model's representation of such constraints--including the Model's use 
of technology paths--provides a reasonable means of accounting for 
them.
4. Compliance Simulation
    The CAFE model provides a way of estimating how vehicle 
manufacturers could attempt to comply with a given CAFE standard by 
adding technology to fleets that the agencies anticipate they will 
produce in future model years. This exercise constitutes a simulation 
of manufacturers' decisions regarding compliance with CAFE or 
CO2 standards.
    This compliance simulation begins with the following inputs: (a) 
The analysis fleet of vehicles from model year 2017 discussed below in 
Section VI.B.1, (b) fuel economy improving technology estimates 
discussed below in Section VI.C, (c) economic inputs discussed below in 
Section VI.D, and (d) inputs defining baseline and potential new CAFE 
or CO2 standards discussed above in Section V. For each 
manufacturer, the model applies technologies in both a logical sequence 
and a cost-optimizing strategy in order to identify a set of 
technologies the manufacturer could apply in response to new CAFE or 
CO2 standards. The model applies technologies to each of the 
projected individual vehicles in a manufacturer's fleet, considering 
the combined effect of regulatory and market incentives while 
attempting to account for manufacturers' production constraints. 
Depending on how the model is exercised, it will apply technology until 
one of the following occurs:
    (1) The manufacturer's fleet achieves compliance \317\ with the 
applicable standard and adding additional technology in the current 
model year would be attractive neither in terms of stand-alone (i.e., 
absent regulatory need) cost-effectiveness nor in terms of facilitating 
compliance in future model years;
---------------------------------------------------------------------------

    \317\ When determining whether compliance has been achieved in 
the CAFE program, existing CAFE credits that may be carried over 
from prior model years or transferred between fleets are also used 
to determine compliance status. For purposes of determining the 
effect of maximum feasible CAFE standards, however, EPCA prohibits 
NHTSA from considering these mechanisms for years being considered 
(though it does so for model years that are already final) and the 
agency runs the CAFE model without enabling these options. 49 U.S.C. 
32902(h)(3).
---------------------------------------------------------------------------

    (2) The manufacturer ``exhausts'' available technologies; \318\ or
---------------------------------------------------------------------------

    \318\ In a given model year, it is possible that production 
constraints cause a manufacturer to ``run out'' of available 
technology before achieving compliance with standards. This can 
occur when: (a) An insufficient volume of vehicles are expected to 
be redesigned, (b) vehicles have moved to the ends of each 
(relevant) technology pathway, after which no additional options 
exist, or (c) engineering aspects of available vehicles make 
available technology inapplicable (e.g., secondary axle disconnect 
cannot be applied to two-wheel drive vehicles).
---------------------------------------------------------------------------

    (3) For manufacturers assumed to be willing to pay civil penalties 
(in the CAFE program), the manufacturer reaches the point at which 
doing so would be more cost-effective (from the manufacturer's 
perspective) than adding further technology.
    The model accounts explicitly for each model year, applying 
technologies when vehicles are scheduled to be redesigned or freshened 
and carrying forward technologies between model years once they are 
applied (until, if applicable, they are superseded by other 
technologies). The model then uses these simulated manufacturer fleets 
to generate both a representation of the U.S. auto industry and to 
modify a representation of the entire light-duty registered vehicle 
population. From these fleets, the model estimates changes in physical 
quantities (gallons of fuel, pollutant emissions, traffic fatalities, 
etc.) and calculates the relative costs and benefits of regulatory 
alternatives under consideration.
    The CAFE model accounts explicitly for each model year, in turn, 
because manufacturers actually ``carry forward'' most technologies 
between model years, tending to concentrate the application of new 
technology to vehicle redesigns or mid-cycle ``freshenings,'' and 
design cycles vary widely among manufacturers and specific products.

[[Page 24277]]

Comments by manufacturers and model peer reviewers strongly support 
explicit year-by-year simulation. Year-by-year accounting also enables 
accounting for credit banking (i.e., carry-forward), as discussed 
above, and at least four environmental organizations recently submitted 
comments urging the agencies to consider such credits, citing NHTSA's 
2016 results showing impacts of carried-forward credits.\319\ Moreover, 
EPCA/EISA requires that NHTSA make a year-by-year determination of the 
appropriate level of stringency and then set the standard at that 
level, while ensuring ratable increases in average fuel economy through 
MY 2020. The multi-year planning capability, simulation of ``market-
driven overcompliance,'' and EPCA credit mechanisms (again, for 
purposes of modeling the CAFE program) increase the model's ability to 
simulate manufacturers' real-world behavior, accounting for the fact 
that manufacturers will seek out compliance paths for several model 
years at a time, while accommodating the year-by-year requirement. This 
same multi-year planning structure is used to simulate responses to 
standards defined in grams CO2/mile, and utilizing the set 
of specific credit provisions defined under EPA's program.
---------------------------------------------------------------------------

    \319\ Comment by Environmental Law & Policy Center, Natural 
Resources Defense Council (NRDC), Public Citizen, and Sierra Club, 
Docket ID EPA-HQ-OAR-2015-0827-9826, at 28-29.
---------------------------------------------------------------------------

    After the light-duty rulemaking analysis accompanying the 2012 
final rule that finalized NHTSA's standards through MY 2021, NHTSA 
began work on changes to the CAFE model with the intention of better 
reflecting constraints of product planning and cadence for which 
previous analyses did not account. This involves accounting for 
expected future schedules for redesigning and ``freshening'' vehicle 
models, and accounting for the fact that a given engine or transmission 
is often shared among more than one vehicle model, and a given vehicle 
production platform often includes more than one vehicle model. These 
real product planning considerations are explained below.
    Like earlier versions, the current CAFE model provides the 
capability for integrated analysis spanning different regulatory 
classes, accounting both for standards that apply separately to 
different classes and for interactions between regulatory classes. 
Light vehicle CAFE and CO2 standards are specified 
separately for passenger cars and light trucks. However, there is 
considerable sharing between these two regulatory classes, where a 
single engine, transmission, or platform can appear in both the 
passenger car and light truck regulatory class. For example, some 
sport-utility vehicles are offered in 2WD versions (classified as 
passenger cars for compliance purposes) and 4WD versions (classified as 
light trucks for compliance purposes). Integrated analysis of 
manufacturers' passenger car and light truck fleets provides the 
ability to account for such sharing and reduces the likelihood of 
finding solutions that could involve introducing impractical and 
unrealistic levels of complexity in manufacturers' product lines. In 
addition, integrated fleet analysis provides the ability to simulate 
the potential that manufacturers could earn CAFE and CO2 
credits by over complying with the standard in one fleet and use those 
credits toward compliance with the standard in another fleet (i.e., to 
simulate credit transfers between regulatory classes).\320\
---------------------------------------------------------------------------

    \320\ Note, however, that EPCA prohibits NHTSA from considering 
the availability of such credit trading when setting maximum 
feasible fuel economy standards. 49 U.S.C. 32902(h)(3).
---------------------------------------------------------------------------

    The CAFE model also accounts for EPCA's requirement that compliance 
be determined separately for fleets of domestic passenger cars and 
fleets of imported passenger cars. The model accounts for all three 
CAFE regulatory classes simultaneously (i.e., in an integrated way) yet 
separately: Domestic passenger cars, imported passenger cars, and light 
trucks. The model further accounts for two related specific statutory 
requirements specifically involving this distinction between domestic 
and imported passenger cars. First, EPCA/EISA requires that any given 
fleet of domestic passenger cars meet a minimum standard, irrespective 
of any available compliance credits. Second, EPCA/EISA requires 
compliance with the standards applicable to the domestic passenger car 
fleet without regard to traded or transferred credits.\321\
---------------------------------------------------------------------------

    \321\ 49 U.S.C. 32903(f)(2) and (g)(4).
---------------------------------------------------------------------------

    However, the CAA has no such limitation regarding compliance by 
domestic and imported vehicles; EPA did not adopt provisions similar to 
the aforementioned EPCA/EISA requirements and is not doing so today. 
Therefore, the CAFE model determines compliance for manufacturers' 
overall passenger car and light truck fleets for EPA's program.
    Each manufacturer's regulatory requirement represents the 
production-weighted harmonic mean of their vehicle's targets in each 
regulated fleet. This means that no individual vehicle has a 
``standard,'' merely a target, and each manufacturer is free to 
identify a compliance strategy that makes the most sense given its 
unique combination of vehicle models, consumers, and competitive 
position in the various market segments. As the CAFE model provides 
flexibility when defining a set of regulatory standards, each 
manufacturer's requirement is dynamically defined based on the 
specification of the standards for any simulation and the distribution 
of footprints within each fleet.
    Given this information, the model attempts to apply technology to 
each manufacturer's fleet in a manner that, given product planning and 
engineering-related considerations, optimizes the selected cost-related 
metric. The metric supported by the NPRM version of the model is termed 
``effective cost.'' The effective cost captures more than the 
incremental cost of a given technology; it represents the difference 
between their incremental cost and the value of fuel savings to a 
potential buyer over the first 30 months of ownership.\322\ In addition 
to the technology cost and fuel savings, the effective cost also 
includes the change in CAFE civil penalties from applying a given 
technology and any estimated welfare losses associated with the 
technology (e.g., earlier versions of the CAFE model simulated low-
range electric vehicles that produced a welfare loss to buyers who 
valued standard operating ranges between re-fueling events). Comments 
on this metric are discussed below, as are model changes responding to 
these comments.
---------------------------------------------------------------------------

    \322\ The length of time over which to value fuel savings in the 
effective cost calculation is a model input that can be modified by 
the user. This analysis uses 30 months' worth of fuel savings in the 
effective cost calculation, using the price of fuel at the time of 
vehicle purchase.
---------------------------------------------------------------------------

    This construction allows the model to choose technologies that both 
improve a manufacturer's regulatory compliance position and are most 
likely to be attractive to its consumers. This also means that 
different assumptions about future fuel prices will produce different 
rankings of technologies when the model evaluates available 
technologies for application. For example, in a high fuel price regime, 
an expensive but very efficient technology may look attractive to 
manufacturers because the value of the fuel savings is sufficiently 
high both to counteract the higher cost of the technology and, 
implicitly, to satisfy consumer demand to balance price increases with 
reductions in operating cost.

[[Page 24278]]

    In general, the model adds technology for several reasons but 
checks these sequentially. The model then applies any ``forced'' 
technologies. Currently, only variable valve timing (VVT) is forced to 
be applied to vehicles at redesign since it is the root of the engine 
path and the reference point for all future engine technology 
applications.\323\ The model next applies any inherited technologies 
that were applied to a leader vehicle on the same vehicle platform and 
carried forward into future model years where follower vehicles (on the 
shared system) are freshened or redesigned (and thus eligible to 
receive the updated version of the shared component). In practice, very 
few vehicle models enter without VVT, so inheritance is typically the 
first step in the compliance loop. Next, the model evaluates the 
manufacturer's compliance status, applying all cost-effective 
technologies regardless of compliance status.\324\ Then the model 
applies expiring overcompliance credits (if allowed to do so under the 
perspective of either the ``unconstrained'' or ``standard setting'' 
analysis, for CAFE purposes).\325\ At this point, the model checks the 
manufacturer's compliance status again. If the manufacturer is still 
not compliant (and is unwilling to pay civil penalties, again for CAFE 
modeling), the model will add technologies that are not cost-effective 
until the manufacturer reaches compliance. If the manufacturer exhausts 
opportunities to comply with the standard by improving fuel economy/
reducing emissions (typically due to a limited percentage of its fleet 
being redesigned in that year), the model will apply banked CAFE or 
CO2 credits to offset the remaining deficit. If no credits 
exist to offset the remaining deficit, the model will reach back in 
time to alter technology solutions in earlier model years.
---------------------------------------------------------------------------

    \323\ As a practical matter, this affects very few vehicles. 
More than 95 percent of vehicles in the market file either already 
have VVT present or have surpassed the basic engine path through the 
application of hybrids or electric vehicles.
    \324\ For further explanation of how the CAFE model considers 
the effective cost of applying different technologies see the CAFE 
Model Documentation for the final rule, at S5.3 Compliance 
Simulation Algorithm.
    \325\ As mentioned above, EPCA prohibits consideration of 
available credits when setting maximum feasible fuel economy 
standards. 49 U.S.C. 32902(h)(3).
---------------------------------------------------------------------------

    The CAFE model implements multi-year planning by looking back, 
rather than forward. When a manufacturer is unable to comply through 
cost-effective (i.e., producing effective cost values less than zero) 
technology improvements or credit application in a given year, the 
model will ``reach back'' to earlier years and apply the most cost-
effective technologies that were not applied at that time and then 
carry those technologies forward into the future and re-evaluate the 
manufacturer's compliance position. The model repeats this process 
until compliance in the current year is achieved, dynamically 
rebuilding previous model year fleets and carrying them forward into 
the future, and accumulating CAFE or CO2 credits from over-
compliance with the standard wherever appropriate.
    In a given model year, the model determines applicability of each 
technology to each vehicle platform, model, engine, and transmission. 
The compliance simulation algorithm begins the process of applying 
technologies based on the CAFE or CO2 standards specified 
during the current model year. This involves repeatedly evaluating the 
degree of noncompliance, identifying the next ``best'' technology 
(ranked by the effective cost discussed earlier) available on each of 
the parallel technology paths described above and applying the best of 
these. The algorithm combines some of the pathways, evaluating them 
sequentially instead of in parallel, to ensure appropriate incremental 
progression of technologies.
    The algorithm first finds the best next applicable technology in 
each of the technology pathways and then selects the best among these. 
For CAFE purposes, the model applies the technology to the affected 
vehicles if a manufacturer is either unwilling to pay penalties or if 
applying the technology is more cost-effective than paying penalties. 
Afterwards, the algorithm reevaluates the manufacturer's degree of 
noncompliance and continues application of technology. Once a 
manufacturer reaches compliance (i.e., the manufacturer would no longer 
need to pay penalties), the algorithm proceeds to apply any additional 
technology determined to be cost-effective (as discussed above). 
Conversely, if a manufacturer is assumed to prefer to pay penalties, 
the algorithm only applies technology up to the point where doing so is 
less costly than paying penalties. The algorithm stops applying 
additional technology to this manufacturer's products once no more 
cost-effective solutions are encountered. This process is repeated for 
each manufacturer present in the input fleet. It is then repeated for 
each model year. Once all model years have been processed, the 
compliance simulation algorithm concludes. The process for 
CO2 standard compliance simulation is similar, but without 
the option of penalty payment, such that technologies are applied until 
compliance (accounting for any modeled application of credits) is 
achieved. For both CAFE and CO2 standards, the model also 
applies any additional (i.e., beyond required for compliance) 
technology that ``pays back'' within a specified period (for the NPRM 
and today's analysis, 30 months).
    Some commenters argued that the CAFE model applies constraints that 
excessively limit options manufacturers have to add technology, causing 
the model to overestimate costs to achieve a given level of 
improvement.\326\ Some of these commenters further argued that the 
agencies should assume greater potential to apply technologies that 
contribute to compliance by improving air conditioner efficiency or 
otherwise reducing ``off cycle'' fuel consumption and CO2 
emissions.\327\ Other commenters argued that such constraints, while 
warranting some refinements, help the model to simulate manufacturers' 
decision making realistically and to estimate technology effectiveness 
and costs reasonably.328 329
---------------------------------------------------------------------------

    \326\ NHTSA-2018-0067-12057, CBD, et. al, p. 3.
    \327\ NHTSA-2018-0067-11741, ICCT, Attachment 2, p. 4.
    \328\ NHTSA-2018-0067-12073, Alliance of Automobile 
Manufacturers, pp. 134-36.
    \329\ American Honda Motor Co., ``Honda Comments on the NPRM and 
various proposals contained therein--Prepared for NHTSA, EPA and 
ARB,'' October 17, 2018, pp. 12-16.
---------------------------------------------------------------------------

    Some commenters questioned the ``effective cost'' metric the model 
uses to decide among available options, claiming that the metric also 
causes the model to avoid selection of pathways that are not always 
economically optimal.\330\ One of these commenters recommended the 
agencies modify the effective cost metric for CO2 compliance 
by removing the term placing a monetary value on progress toward 
compliance, and instead dividing the remaining net cost (i.e., the 
increase in technology costs minus a portion of the fuel outlays 
expected to be avoided) by the additional CO2 credits 
earned.\331\ Another of these commenters claimed on one hand, that the 
effective cost metric ``does not include a measurement of the 
technology's reduction in fuel consumption or CO2 
emissions'' and, on the other, that the metric inappropriately places a 
value on avoided fuel consumption.\332\
---------------------------------------------------------------------------

    \330\ NHTSA-2018-0067-11741, ICCT, Attachment 3, p. I-62.
    \331\ NHTSA-2018-0067-12039, UCS, Technical Appendix, pp. 28-32.
    \332\ NHTSA-2018-0067-12108, EDF, Appendix B, p. 67.
---------------------------------------------------------------------------

    One commenter claimed that the model inappropriately allows earned

[[Page 24279]]

credits (including CO2 program credits for which EPA has 
granted a one-time exemption from carry-forward limits) to expire while 
also showing undue degrees overcompliance with standards, and further 
proposed that the model be modified to simulate both credit ``carry 
back'' (aka ``borrowing'') and credit trading between 
manufacturers.\333\
---------------------------------------------------------------------------

    \333\ NHTSA-2018-0067-12039, UCS, Technical Appendix, pp. 36-40.
---------------------------------------------------------------------------

    In addition, some commenters indicated that the agencies' analysis 
(impliedly, its modeling) should account for some States' mandates that 
manufacturers sell minimum quantities of ``Zero Emission Vehicles'' 
(ZEVs).334 335
---------------------------------------------------------------------------

    \334\ NHTSA-2018-0067-12036, Volvo, p. 5.
    \335\ NHTSA-2018-0067-11813, South Coast AQMD, Attachment 1, p. 
4 and EIS comments, p. 9.
---------------------------------------------------------------------------

    Regarding the model's representation of engineering and product 
planning constraints, the agencies maintain that having such 
constraints produces more realistic potential (as mentioned above, not 
``predicted'') pathways forward from manufacturers' current fleets than 
would be the case were these constraints removed. For example, while 
manufacturers' product plans are protected as confidential business 
information (CBI), some manufacturers' public comments demonstrate 
year-by-year balancing such as the CAFE model emulates.\336\ Also, even 
manufacturers that have invested in technologies such as hybrid 
electric powertrains and Atkinson cycle engines have commented that a 
manufacturers' past investments will constrain the pathways it can 
practicably take.\337\ Therefore, the agencies have retained the 
model's basic structural constraints, have updated and expanded the 
model's technology paths (and, as discussed, the model's logic for 
approaching these paths), and have updated inputs defining the range of 
manufacturer-, technology-, and product-specific constraints. These 
updates are discussed below at greater length.
---------------------------------------------------------------------------

    \336\ See, e.g., FCA, pp. 5-6.
    \337\ Toyota, Attachment 1, p. 10.
---------------------------------------------------------------------------

    The agencies have also reconsidered opportunities manufacturers may 
have to expand the application of technologies that contribute to 
compliance by improving air conditioner efficiency or otherwise 
reducing ``off cycle'' fuel consumption and CO2 emissions, 
or to earn credit toward CO2 compliance by using 
refrigerants with lower global warming potential (GWP) or reducing the 
potential for refrigerant leaks. The version of the model used for the 
proposal accommodates inputs that, for each of these adjustments or 
credits, applies the same value to every model year. The agencies have 
revised the model to accommodate inputs that specify the degree of 
adjustment or credit separately for each model year, and have applied 
inputs that assume manufacturers will increase application of these 
improvements to the highest levels reported within the industry.
    Regarding comments on the effective cost metric the model uses to 
compare and select among available options to add technology, the 
agencies have considered changes such as those mentioned above. Given 
the myriad of factors that manufacturers can consider, any weighing to 
be conducted using publicly-available information will constitute a 
simplified representation. Nevertheless, within the model's context, it 
is obvious that any weighing of options should, at a minimum, consider 
some measure of each option's costs and benefits. Since this aspect of 
the model involves simulating manufacturers' decisions, it is also 
clearly appropriate that these costs and benefits be considered from a 
manufacturer perspective rather than a social perspective.
    The effective cost metric used for the NPRM version of the model 
represents the cost of a given option as the cost to apply a given 
technology to a given set of vehicles, and represents the benefit of 
the same option as the extent to which the manufacturer might expect 
buyers would be willing to pay for fuel economy (as represented by a 
portion of the projected fuel savings), combined with any reduction in 
CAFE civil penalties that the manufacturer might ultimately need to 
pass along to buyers. The reduction in CAFE civil penalties places a 
value on progress made toward compliance with CAFE standards. The CAA 
provides no direction regarding CO2 standards, so the model 
accepts inputs specifying an analogous basis for valuing changes in the 
quantity of CO2 credits earned from (or required by) a 
manufacturer's fleet. Because each of these three components 
(technology cost, fuel benefit, and compliance benefit) is expressed in 
dollars, subtracting benefits from costs produces a net cost, and after 
dividing net costs by the number of affected vehicles, it is logical 
to, at each step, select the option that produces the most negative net 
unit cost. This approach can be interpreted as maximizing net benefits 
(to the manufacturer).
    As an alternative, the agencies considered a simpler metric that 
considers only the cost of the option and the extent to which the 
option increases the quantity of earned credits, and does not require 
input assumptions regarding how to value progress toward compliance. 
Such a metric is expressed in dollars per ton or dollars per gallon 
such that seeking options that produce the smallest (positive) values 
can be interpreted as maximizing cost effectiveness (of progress toward 
compliance). However, simply comparing technology costs to 
corresponding compliance improvements would implicitly assume that 
manufacturers do not respond at all to fuel prices. This assumption is 
clearly unrealistic. For example, if diesel fuel costs $5 per gallon 
and gasoline costs $2 per gallon, manufacturers will be reluctant to 
respond to stringent CAFE or CO2 standards by replacing 
gasoline engines with diesel engines. Manufacturers' comments credibly 
assert that fuel prices matter, and in the agencies' judgment, 
simulations of decisions between available options should continue to 
account for avoided fuel outlays.
    On the other hand, while any metric should incorporate some measure 
of progress toward compliance, it is not obvious that this progress 
must be expressed in monetary terms. While the CAFE civil penalty 
provisions provide a logical basis for doing so with respect to CAFE, 
the recently-introduced (through EISA) option to trade credit between 
manufacturers adds an alternative basis that is undefined and 
uncertain, in part because terms of past trades are not known to the 
agencies. Also, as mentioned above, EPCA/EISA's civil penalty 
provisions are not applicable to noncompliance with CO2 
standards.
    Therefore, for the purpose of selecting among available options to 
add technology, the agencies consider it reasonable to use the degree 
of compliance improvement in ``raw'' (i.e., not monetized) form, and to 
divide net costs (i.e., technology costs minus a portion of expected 
avoided fuel outlays) by this improvement. Under a range of side-by-
side tests, this change to the effective cost metric most frequently 
produced lower overall estimates of compliance costs. However, 
differences vary among manufacturers, model years, and regulatory 
alternatives, and also depend on other model inputs. For example, at 
high fuel prices, the new metric tends to select more expensive 
pathways than the NPRM's metric, and with the new metric, a case 
simulating ``perfect trading'' of CO2 compliance credits 
tends to show such trading increasing compliance costs rather than, as 
expected, decreasing such costs.
    The version of the model used for the proposal simulates the 
potential that, for

[[Page 24280]]

a given fleet in a given model year, a manufacturer might be able to 
use credits from an earlier model year or a different fleet. This 
version of the model did not explicitly simulate the potential that, 
for a given fleet in a given model year, a manufacturer might be to use 
credits from a future model year or a different manufacturer. However, 
the agencies did apply model inputs that reflected assumptions 
regarding possible trading of credits actually earned prior to model 
year 2016 (the earliest represented in detail in the agencies' 
analysis), and the agencies did examine a case (included in the 
sensitivity analysis) involving hypothetical ``perfect'' trading of 
CO2 credits among manufacturers by treating the industry as 
a single ``manufacturer.'' Although past versions of the CAFE Model had 
included code under development with a view toward eventually 
simulating one or both of these provisions, this code had never 
proceeded beyond preliminary experimentation, and had never been the 
focus of peer reviews or application in published analyses.
    Nevertheless, the agencies considered expanding the model to 
simulate credit ``carry back'' (or ``borrowing'') and trading 
(explicitly, rather than in an idealized hypothetical way). The 
agencies closely examined the corresponding model revisions proposed by 
UCS and determined that such methods would not produce repeatable 
results. This is because the approach proposed by UCS ``randomly swaps 
items in list to minimize trading bias.'' \338\
---------------------------------------------------------------------------

    \338\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 84-87.
---------------------------------------------------------------------------

    Even if such revisions could be modified to produce non-random 
results, including credit banking and trading would introduce highly 
speculative elements into the agencies' analysis. While manufacturers 
have occasionally indicated plans to carry back credits from future 
model years, those plans have sometimes backfired when projected 
credits have failed to materialize, e.g., by misjudging consumer demand 
for more efficient vehicles. In the agencies' judgment, it would be 
inappropriate to set standards based on an analysis that relies on the 
type of borrowing that has been known to fail. To rely also on credit 
trading during the model years included in the analysis would compound 
this undue speculation. For example, including credit borrowing and 
trading throughout the analysis, as some commenters proposed, would 
lead to an analysis that depends on the potential that, in order to 
comply with the MY 2022 standard for light trucks, FCA could use 
credits it expects to be able to buy from another manufacturer in MY 
2025. Even if the agencies' analysis had knowledge of and made use of 
manufacturers' actual product plans, expectations about the ability to 
borrow others' unearned credits would necessarily be considered risky 
and unreliable. Within an analysis that, to provide for public 
disclosure, extrapolates forward many years from the most recent 
observed fleet, such transactions would add an unreasonable level of 
speculation. Therefore, the agencies have declined to introduce credit 
borrowing and trading into the model's logic.
    The analysis presented in the proposal applied inputs reflecting 
potential application of credits earned earlier than the first year 
modeled explicitly. However, as observed by some commenters, those 
inputs did not fully account for the one-time exemption from the 5-year 
limit on the extent to which manufacturers may carry forward 
CO2 credits. The agencies have updated the analysis fleet to 
MY 2017 and, in doing so, have updated inputs specifying how credits 
earned to MY 2017 might be applied. These updates implement a 
reasonably full accounting of these ``legacy'' credits, including of 
the one-time exemption from the credit life limit.
    As mentioned above, some commenters also indicated that the model 
is unrealistically ``reluctant'' to apply credits carried forward from 
early model years. As explained in the proposal and in the model 
documentation, the model's application of carried-forward credits is 
partially controlled by model inputs, which, for the proposal, were set 
to assume that manufacturers would tend to retain credits as long as 
possible. This assumption is entirely consistent with manufacturers' 
past practice and logical in a context wherein the stringency of 
standards is generally increasing over time. Even though using credits 
in some model years might seem initially advantageous, doing so means 
foregoing actual improvements likely to be needed in later model years.
    Regarding the model's treatment of mandates and credits for the 
sale of ZEVs, as indicated in the model documentation accompanying the 
proposal, these capabilities were experimental in that version of the 
model. The reference case analysis for today's notice, like that for 
the proposal, does not simulate compliance with ZEV mandates.\339\
---------------------------------------------------------------------------

    \339\ The agencies note their finalization of the One National 
Program Final Action, in which EPA partially withdrew a waiver of 
CAA preemption previously granted to the State of California 
relating to its ZEV mandate, and NHTSA finalized regulations 
providing that State ZEV mandates are impliedly and expressly 
preempted by EPCA. This joint action is available at 84 FR 51310.
---------------------------------------------------------------------------

    For the NPRM, the CAFE model was exercised with inputs extending 
this explicit simulation of technology application through MY 2032, as 
the agencies anticipated this was sufficiently beyond MY 2026 that 
nearly all multiyear planning attributable to MY 2026 standards should 
be accounted for, and any compliance credits carried forward from MY 
2026 would have expired. The analysis met this expectation, and the 
agencies presented analysis of the resultant estimated impacts over the 
useful lives of vehicles produced prior to MY 2030. The agencies 
invited comment on all aspects of the analysis, and relevant to this 
aspect of the analysis--i.e., its perspective and temporal span--EDF 
stated that that these led the agencies to overstate the proposal's 
positive impacts on safety, in part because by explicitly representing 
vehicle model years only through 2032, the agencies had failed to 
account for the impact of distant model years prices and fuel economy 
levels on the retention and scrappage of vehicles produced through MY 
2029.\340\ For example, some vehicles produced in MY 2026 will likely 
still be on the road during calendar years (CY) 2033-2050 and the rates 
at which these MY 2026 vehicles will be scrapped during CYs 2033-2050 
will be impacted by the prices and fuel economy levels of vehicles 
produced during MYs 2033-2050.
---------------------------------------------------------------------------

    \340\ EDF, NHTSA-2018-0067-12108, Attachment A at 11 and 
Attachment B at 11-28.
---------------------------------------------------------------------------

    The agencies have addressed this comment by expanding model inputs 
to extend the explicit simulation of technology application through MY 
2050. Most of these expanded model inputs involve the analysis fleet 
and inputs defining the cost and availability of various fuel-saving 
technologies. These inputs are discussed below. The agencies also made 
minor modifications to the model in order to extend model outputs to 
cover this wider span and to carry forward each regulatory 
alternative's standards automatically through the last year to be 
modeled (e.g., extending standards without change from MY 2032 through 
MY 2050). The model documentation discusses these

[[Page 24281]]

minor changes.\341\ In addition, although the agencies published 
detailed model output files documenting all estimated annual impacts 
through calendar year 2089, the notice and PRIA both emphasized the 
above-mentioned ``model year'' perspective, as in past regulatory 
analyses supporting CAFE and CO2 standards. Recognizing that 
an alternative ``calendar year'' perspective is of interest to EDF and, 
perhaps other stakeholders, the agencies have expanded the presentation 
of results in today's notice and FRIA by presenting some physical 
impacts (e.g., fuel consumption and CO2 emissions) as well 
as monetized benefits, costs, and net benefits for each of CYs 2017-
2050. All of these results appear in the model output files published 
with today's notice, as do corresponding results for more specific 
impacts (e.g., year-by-year components of monetized social costs).\342\
---------------------------------------------------------------------------

    \341\ The model and documentation are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
    \342\ Detailed model inputs and outputs are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

5. Calculation of Physical Impacts
    Once it has completed the simulation of manufacturers' potential 
application of technology in response to CAFE/CO2 standards 
and fuel prices, the CAFE Model calculates impacts of the resultant 
changes in new vehicle fuel economy levels and prices. This involves 
several steps.
    The model calculates changes in the total quantity of new vehicles 
sold in each model year as well as the relative shares passenger cars 
and light trucks comprise of the overall new vehicle market. The 
agencies received many comments on the estimation of sales impacts, and 
as discussed below, today's analysis applies methods and corresponding 
estimates that reflect careful consideration of these comments. Related 
to these calculations, the model now operates in an iterated fashion 
with a view toward obtaining sales impacts that are balanced with 
changes in vehicle prices and fuel economy levels. This involves 
solving for compliance, calculating sales impacts, re-solving for 
compliance, and repeating these steps as many times as specified in 
model inputs. For today's analysis, the agencies operated the model 
with four iterations, as early testing suggested three iterations 
should be sufficient for fleetwide results to converge between 
iterations. The model documentation describes the procedures for 
iteration in detail.
    The impacts on outlays for new vehicles occur coincident with the 
sale of these vehicles so the model can simply calculate and record 
these for each model year included in the analysis. However, virtually 
all other impacts result from vehicle operation that extends long after 
a vehicle is produced. Like other models (including, e.g., NEMS), the 
CAFE Model includes procedures (sometimes referred to as ``stock 
models'' or as models of fleet turnover) to estimate annual rates at 
which new vehicles are used and subsequently scrapped. The agencies 
received many comments on procedures for estimating vehicle scrappage 
and on procedures for estimating annual quantities of highway travel, 
accounting for the elasticity of travel demand with respect to per-mile 
costs for fuel. Below, Section VI.D.1 discusses these comments and 
reviews procedures and corresponding estimates that also reflect 
careful consideration of these comments.
    For each vehicle model in each model year, these procedures result 
in estimates of the number of vehicles remaining in service in each 
calendar year, as well as the annual mileage accumulation (i.e., 
vehicle miles traveled, or VMT) in each calendar year. As mentioned 
above, most of the physical impacts of interest derive from this 
vehicle operation. Also discussed above, the simulated application of 
technology results in ``initial'' and ``final'' estimates of the cost, 
fuel type, fuel economy, and fuel share (for, in particular, PHEVs that 
can run on gasoline or electricity) applicable to each vehicle model in 
each model year. Together with quantities of travel, and with estimates 
of the ``gap'' between ``laboratory'' and ``on-road'' fuel economy, 
these enable calculation of quantities of fuel consumed in each year 
during the useful life of each vehicle model produced in each model 
year.\343\ The model documentation provides specific procedures and 
formulas implementing these calculations.
---------------------------------------------------------------------------

    \343\ The agencies have applied the same estimates of the ``on 
road gap'' as applied for the analysis supporting the NPRM. For 
operation on gasoline, diesel, E85, and CNG, this gap is 20 percent; 
for electricity and hydrogen, 30 percent.
---------------------------------------------------------------------------

    As for the NPRM, the model calculates emissions of CO2 
and other air pollutants, reporting emissions both from vehicle 
tailpipes and from upstream processes (e.g., petroleum refining) 
involved in producing and supplying fuels. Section VI.D.3 below reviews 
methods, models, and estimates used in performing these calculations. 
The model also calculates impacts on highway safety, accounting for 
changes in travel demand, changes in vehicle mass, and continued past 
and expected progress in vehicle safety (through, e.g., the application 
of new crash avoidance systems). Section VI.D.2 discusses methods, data 
sources, and estimates involved in estimating safety impacts, comments 
on the same, and changes included in today's analysis. In response to 
the NPRM, some comments urged the agencies also to quantify different 
types of health impacts from changes in air pollution rather than only 
accounting for such impacts in aggregate estimates of the social costs 
of air pollution. Considering these comments, the agencies added such 
calculations to the model, as discussed in Section VI.D.3.
6. Calculation of Benefits and Costs
    Having estimated how technologies might be applied going forward, 
and having estimated the range of resultant physical impacts, the CAFE 
Model calculates a variety of private and social benefits and costs, 
reporting these from the consumer, manufacturer, and social 
perspectives, both in undiscounted and discounted present value form 
(given inputs specifying the corresponding discount rate and present 
year). Estimates of regulatory costs are among the direct outputs of 
the simulation of manufacturers' potential responses to new standards. 
Other benefits and costs are calculated based on the above-mentioned 
estimates of travel demand, fuel consumption, emissions, and safety 
impacts. The agencies received many comments on the NPRM's calculation 
of benefits and costs, and Section VI.D.1 discusses these comments and 
presents the methods, data sources, and estimates used in calculating 
benefits and costs reported here.
7. Structure of Model Inputs and Outputs
    All CAFE Model inputs and outputs described above are specified in 
Microsoft Excel format, and the user can define and edit all inputs to 
the system. Table VI-3 describes (non-exhaustively) which inputs are 
contained within each input file and Table VI-4 describes which outputs 
are contained in each output file. This is important for three reasons: 
(1) Each file is discussed throughout the following sections; (2) 
several commenters conflated aspects of the model with its inputs; and 
(3) several commenters seemed confused about where to find specific 
information in the output files. This information was described in 
detail in the NPRM CAFE Model Documentation, but is reproduced here for 
quick reference. When specifically referencing the input

[[Page 24282]]

or output file used for the NPRM or final rule in the following 
discussion, NPRM or FRM, respectively, will precede the file name.
[GRAPHIC] [TIFF OMITTED] TR30AP20.080

[GRAPHIC] [TIFF OMITTED] TR30AP20.081

    A catalog of the Argonne National Laboratory Autonomie fuel economy 
technology effectiveness value output files are reproduced in the 
following Table VI-5 as well. The left column shows the terminology 
used in this text to refer to the file, while the right column 
describes each file. NPRM or FRM, respectively, may precede the 
terminology in the text as appropriate.

[[Page 24283]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.082

    Finally, Table VI-6 lists the terminologies used to refer to other 
model-related documents which are referred to frequently throughout the 
text. NPRM or FRM, respectively, may precede the terminology in the 
text as appropriate.

[[Page 24284]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.083

B. What inputs does the compliance analysis require?

1. Analysis Fleet
    The starting point for the evaluation of the potential feasibility 
of different stringency levels for future CAFE and CO2 
standards is the analysis fleet, which is a snapshot of the recent 
vehicle market. The analysis fleet provides a baseline from which to 
project what and how additional technologies could feasibly be applied 
to vehicles in a cost-effective manner to raise those vehicles' fuel 
economy and lower their CO2 emission levels.\344\ The fleet 
characterization also provides a reference point with data for other 
factors considered in the analysis, including environmental effects and 
effects estimated by the economic modules (i.e., sales, scrappage, and 
labor utilization). When the scope of the analysis widens, another 
piece of data must be included for each vehicle in the analysis fleet 
to map a given element of the fleet appropriately onto an analysis 
module.
---------------------------------------------------------------------------

    \344\ The CAFE model does not generate compliance paths a 
manufacturer should, must, or will deploy. It is intended as a tool 
to demonstrate a compliance pathway a manufacturer could choose. It 
is almost certain all manufacturers will make compliance choices 
differing from those projected by the CAFE model.
---------------------------------------------------------------------------

    For the analysis presented in this final rule, the analysis fleet 
includes information about vehicles that is essential for each analysis 
module. The first part of projecting how additional technologies could 
be applied to vehicles is knowing which vehicles are produced by which 
manufacturers, the fuel economies of those vehicles, how many of each 
are sold, whether they are passenger cars or light trucks, and their 
footprints. This is important because it improves understanding of the 
overall impacts of different levels of CAFE and CO2 
standards; overall impacts that result from industry's response to 
standards, and industry's response, is made up of individual 
manufacturer responses to the standards in light of the overall market 
and their individual assessment of consumer acceptance. Establishing an 
accurate representation of manufacturers' existing fleets (and the 
vehicle models in them) that will be subject to future standards helps 
in predicting potential individual manufacturer responses to those 
future standards in addition to potential changes in those standards.
    Another part of projecting how additional fuel economy improving 
technologies could be applied to vehicles is knowing which fuel saving 
technologies manufacturers have equipped on which vehicles. In many 
cases, the agencies also collect and reference additional information 
on other vehicle attributes to help with this process.\345\ Accounting 
for technologies already applied to vehicles helps avoid ``double-
counting'' the value of those technologies, by assuming they are still 
available to be applied to improve fuel economy and reduce 
CO2 emissions. It also promotes more realistic 
determinations of what additional technologies can feasibly be applied 
to those vehicles: If a manufacturer has already started down a 
technological path to fuel economy or performance improvements, the 
agencies do not assume it will completely abandon that path because 
doing so would be unrealistic and fails to represent accurately 
manufacturer responses to standards. Each vehicle model (and 
configurations of each model) in the analysis fleet, therefore, has a 
comprehensive list of its technologies, which is important because 
different configurations may have different technologies applied to 
them.\346\ In addition, to properly account for technology costs, the 
agencies assign each vehicle to a technology class and an engine class. 
Technology classes reference each vehicle to a set of full vehicle 
simulations, so that the agencies may project fuel efficiency with 
combinations of additional fuel saving equipment and hybrid and 
electric vehicle battery costs.
---------------------------------------------------------------------------

    \345\ For instance, curb weight, horsepower, drive 
configuration, pickup bed length, oil type, body style, aerodynamic 
drag coefficients, and rolling resistance coefficients, and (if 
applicable) battery sizes are all required to assign technology 
content properly.
    \346\ Considering each vehicle model/configuration also improves 
the ability to consider the differential impacts of different levels 
of potential standards on different manufacturers, since all vehicle 
model/configurations ``start'' at different places, in terms of 
technologies already used and how those technologies are used.
---------------------------------------------------------------------------

    Yet another part of projecting which vehicles might exist in future 
model years is developing reasonable real-world assumptions about when 
and how manufacturers might apply certain technologies to vehicles. The 
analysis fleet accounts for links between vehicles, recognizing vehicle 
platforms will share technologies, and the vehicles that make up that 
platform should receive (or not receive) additional technological 
improvements together. Shared engines, shared transmissions, and shared 
vehicle platforms for mass reduction technology are considered. In 
addition, each vehicle model/configuration in the analysis fleet also 
has information about its redesign

[[Page 24285]]

schedule, i.e., the last year it was redesigned and when the agencies 
expect it to be redesigned again. Redesign schedules are a key part of 
manufacturers' business plans, as each new product can cost more than 
$1B, and involve a significant portion of a manufacturer's scarce 
research, development, and manufacturing and equipment budgets and 
resources.\347\ Manufacturers have repeatedly told the agencies that 
sustainable business plans require careful management of resources and 
capital spending, and that the length of time each product remains in 
production is crucial to recouping the upfront product development and 
plant/equipment costs, as well as the capital needed to fund the 
development and manufacturing equipment needed for future products. 
Because the production volume of any given vehicle model varies within 
a manufacturer's product line, and varies among different 
manufacturers, redesign schedules typically vary for each model and 
manufacturer. Some (relatively few) technological improvements are 
small enough that they can be applied in any model year; a few other 
technological improvements may be applied during a refreshening (when a 
few additional changes are made, but well short of a full redesign), 
but others are major enough that they can only be cost-effectively 
applied at a vehicle redesign, when many other things about the vehicle 
are already changing. Ensuring the CAFE model makes technological 
improvements to vehicles only when it is feasible to do so also helps 
the analysis better represent manufacturer responses to different 
levels of standards.
---------------------------------------------------------------------------

    \347\ Shea, T., Why Does It Cost So Much For Automakers To 
Develop New Models? Autoblog (Jul. 27, 2010), https://www.autoblog.com/2010/07/27/why-does-it-cost-so-much-for-automakers-to-develop-new-models/.
---------------------------------------------------------------------------

    Finally, the agencies restrict the applications of some 
technologies on some vehicles upon determining the technology is not 
compatible with the functional and performance requirements of the 
vehicle, or if the manufacturers are unlikely to apply a specific 
technology to a specific vehicle for reasons articulated with 
confidential business information that the agencies found credible.
    Other data important for the analysis that are referenced to the 
analysis fleet include baseline economic, environmental, and safety 
information. Vehicle fuel tank size is required to estimate range and 
refueling benefit while curb weights and safety class assignments help 
the agencies consider how changes in vehicle mass may affect safety. 
The agencies identify the final assembly location for each vehicle, 
engine, and transmission, as well as the percent of U.S. content to 
support the labor impact analysis. In addition, the aforementioned 
accounting for first-year vehicle production volumes (i.e., the number 
of vehicles of each new model sold in MY 2017, for this analysis) is 
the foundation for estimating how future vehicle sales might change in 
response to different potential standards.
    The input file for the CAFE model characterizing the analysis 
fleet, referred to as the ``market inputs'' file or ``market data'' 
file, accordingly includes a large amount of data about vehicles, their 
technological characteristics, the manufacturers and fleets to which 
they belong, and initial prices and production volumes, which provide 
the starting points for projection (by the sales model) to ensuing 
model years. In the Draft TAR (which utilized a MY 2015 analysis fleet) 
and NPRM (which utilized a MY 2016 analysis fleet), the agencies needed 
to populate about 230,000 cells in the market data file to characterize 
the fleet. For this final rule (which utilized a MY 2017 analysis 
fleet), the agencies populated more than 400,000 cells to characterize 
the fleet. While the fleet is not actually much more heterogeneous in 
reality,\348\ the agencies have provided and collected more data to 
justify the characterization of the analysis fleet, and to support the 
functionality of modules in the CAFE model.
---------------------------------------------------------------------------

    \348\ The expansion of cells is primarily due to (1) considering 
more technologies, and (2) listing trim levels separately, which 
often yields more precise curb weights and more accurate 
manufacturer suggested retail prices.
---------------------------------------------------------------------------

    A solid characterization of a recent model year as an analytical 
starting point helps realistically estimate ways manufacturers could 
potentially respond to different levels of standards, and the modeling 
strives to simulate realistically how manufacturers could progress from 
that starting point. While manufacturers can respond in many ways 
beyond those represented in the analysis (e.g., applying other 
technologies, shifting production volumes, changing vehicle footprint), 
such that it is impossible to predict with any certainty exactly how 
each manufacturer will respond, it is still important to establish a 
solid foundation from which to estimate potential costs and benefits of 
potential future standards. The following sections discuss aspects of 
how the analysis fleet was built for this analysis, and includes 
discussion of the comments on fleet that the agencies received on the 
proposed rule.
a) Principles on Data Sources Used To Populate the Analysis Fleet
    The source data for vehicles in the analysis fleet and their 
technologies is a central input for the analysis. The sections below 
discuss pros and cons of different potential sources and what the 
agencies used for this analysis, and responds to comments the agencies 
received on data sources in the proposal.
(1) Use of Confidential Business Information Versus Publicly-Releasable 
Sources
    Since 2001, CAFE analysis has used either confidential, forward-
estimating product plans from manufacturers, or publicly available data 
on vehicles already sold as a starting point for determining what 
technologies can be applied to what vehicles in response to potential 
different levels of standards. The use of either data source requires 
certain tradeoffs. Confidential product plans comprehensively represent 
what vehicles a manufacturer expects to produce in coming years, 
accounting for plans to introduce new vehicles and fuel-saving 
technologies and, for example, plans to discontinue other vehicles and 
even brands. This information can be very thorough and can improve the 
accuracy of the analysis, but cannot be publicly released. This makes 
it difficult for public commenters to reproduce the analysis for 
themselves as they develop their comments. Some non-industry commenters 
have also expressed concern about manufacturers having an incentive in 
the submitted plans to underestimate (deliberately or not) their future 
fuel economy capabilities and overstate their expectations about, for 
example, the levels of performance of future vehicle models in order to 
affect the analysis. Accordingly, since 2010, EPA and NHTSA have based 
analysis fleets almost exclusively on information from commercial and 
public sources, starting with CAFE compliance data and adding 
information from other sources.
    An analysis fleet based primarily on public sources can be released 
to the public, solving the issue of commenters being unable to 
reproduce the overall analysis. However, industry commenters have 
argued such an analysis fleet cannot accurately reflect manufacturers' 
actual plans to apply fuel-saving technologies (e.g., manufacturers may 
apply turbocharging to improve not just fuel economy, but also to 
improve vehicle performance) or manufacturers' plans to change product 
offerings by introducing some vehicles and brands and discontinuing 
other

[[Page 24286]]

vehicles and brands, precisely because that information is typically 
confidential business information (CBI). A fully-publicly-releasable 
analysis fleet holds vehicle characteristics unchanged over time and 
lacks some level of accuracy when projected into the future. For 
example, over time, manufacturers introduce new products and even 
entire brands. On the other hand, plans announced in press releases do 
not always ultimately bear out, nor do commercially available third-
party forecasts. Assumptions could be made about these issues to 
improve the accuracy of a publicly releasable analysis fleet, but 
concerns include that this information would either be largely 
incorrect, or, if the assumptions were correct, information would be 
released that manufacturers would consider CBI.
    Furthermore, some technologies considered in the rulemaking are 
difficult to observe in the analysis fleet without expensive teardown 
study and time-consuming benchmarking. Not giving credit for these 
technologies puts the analysis at significant risk of double-counting 
the effectiveness of these technologies, as manufacturers cannot equip 
technologies twice to the same vehicle for double the fuel economy 
benefit. As discussed in the Draft TAR, the agencies assigned little 
(if any) technology application in the baseline fleet for some of these 
technologies.\349\ For the NPRM MY 2016 fleet development process, the 
agencies again offered the manufacturers the opportunity to volunteer 
CBI to the agencies to help inform the technology content of the 
analysis fleet, and many manufacturers did. The agencies were able to 
confirm that many manufacturers had already included many hard-to-
observe technologies in the MY 2016 fleet (which they were not properly 
given credit for in the characterization of the MY 2014 and MY 2015 
fleets presented in Draft TAR) so the agencies reflected this new 
information in the NPRM analysis and in the analysis presented today.
---------------------------------------------------------------------------

    \349\ These technologies include low rolling resistance 
technology (incorrectly applied to zero baseline vehicles in Draft 
TAR), low-drag brakes (incorrectly applied to zero baseline vehicles 
in Draft TAR), electric power steering (incorrectly applied to too 
few vehicles in Draft TAR), accessory drive improvements 
(incorrectly applied to zero baseline vehicles in Draft TAR), engine 
friction reduction (previously named LUBEFR1, LUBEFR2, and LUBEFR3), 
secondary axle disconnect and transmission improvements.
---------------------------------------------------------------------------

    In addition, many manufacturers provided confidential comment on 
the potential applicability of fuel-saving technologies to their fleet. 
In particular, many manufacturers confidentially identified specific 
engine technologies that they will not use in the near term, either on 
specific vehicles, or at all. Reasons varied: Some manufacturers cited 
intellectual property concerns, and others stated functional 
performance concerns for some engine types on some vehicles. Other 
manufacturers shared forward-looking product plans, and explained that 
it would be cost prohibitive to scrap significant investments in one 
technology in favor of another. This topic is discussed in more detail 
in Section VI.B.1.b)(6), below.
    The agencies sought comment on how to address this issue going 
forward, recognizing both the competing interests involved and the 
typical timeframes for CAFE and CO2 standards rulemakings.
    Many commenters expressed concern with the agencies using any CBI 
as part of the rulemaking process. Some commenters expressed concern 
that use of CBI would make the CAFE model subject to inaccuracies 
because manufacturers would only provide additional information in 
situations in which a correction to the agencies' baseline assumptions 
would favor the manufacturers.\350\ The agencies recognize this as a 
reasonable concern, but the analysis presented in the Draft TAR 
consistently assumed very little (if any) technology had been applied 
in the baseline. In addition, many manufacturers shared information on 
advanced technologies that were not yet in production in MY 2017, but 
could be used in the future; manufacturer contributions helped the 
agencies better model many advanced engine technologies and to include 
them in today's analysis, and inclusion of these technologies (and 
costs) in the analysis sometimes lowered the projected cost of 
compliance for stringent alternatives. Other commenters expressed 
concern that automakers would supply false or incomplete information 
that would unduly restrict what technologies can be deployed.\351\ When 
possible, the agencies sought independently to verify manufacturer CBI 
(or claims made by other stakeholders) through lab testing and 
benchmarking.\352\ The agencies found no evidence of misrepresentation 
of engineering specifications in the MY 2017 fleet in manufacturer CBI; 
instead, the agencies were able to verify independently many CBI 
submissions, and confirm the credibility of information provided from 
those sources.
---------------------------------------------------------------------------

    \350\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
    \351\ NHTSA-2018-0067-11741, ICCT.
    \352\ For instance, the agencies continue to evaluate tire 
rolling resistance on production vehicles via independent lab 
testing, and the agencies bench-marked the operating behavior and 
calibration of many engines and transmissions.
---------------------------------------------------------------------------

    Some commenters requested that more CBI be used in the analysis. 
For instance, some commenters suggested that the agencies should return 
to the use of product plans and announcements regarding future fleets 
because manufacturers had already committed investments to bring 
announced products to market.\353\ However, if the agencies were to 
assume that these commitments were already in the baseline, the 
agencies would underestimate the cost of compliance for stringent 
alternatives. Moreover, while upfront investments to bring technologies 
to market are significant, the total marginal costs of components are 
typically large in comparison over the entire product life-cycle, and 
these costs have not yet been realized in vehicles not yet produced.
---------------------------------------------------------------------------

    \353\ NHTSA-2018-0067-11956, PA Department of Environmental 
Protection.
---------------------------------------------------------------------------

    The agencies did make use of some forward-looking CBI in the 
analysis. The agencies received many comments from manufacturers on the 
technological feasibility, or functional applicability of some fuel 
saving technologies to certain vehicles, or certain vehicle 
applications, and the agencies took this information into consideration 
when projecting compliance pathways. These cases are discussed 
generally in Section VI.B.1.b)(6), below, and specifically for each 
technology in those technology sections. Some commenters expressed that 
the use of CBI for future product plans would be acceptable, but only 
if the agencies disclosed the CBI affecting all vehicles through MY 
2025 at the time of publication.\354\ Functionally, this is not 
possible. Manufacturer's confidential product plans cannot be made 
public, as prohibited under NHTSA's regulations at 49 CFR part 512, and 
if the information meets the requirements of section 208(c) of the 
Clean Air Act. If the agencies disclosed confidential information, it 
would not only violate the terms on which the agencies obtained the 
CBI, but it is unlikely that manufacturers would continue to offer CBI, 
which in turn would likely degrade the quality of the analysis. The 
agencies believe that the use of CBI in the NPRM and final rule 
analysis--to confirm, reference, or to otherwise modify aspects of the 
analysis that can be made public--threads the needle between a more 
accurate but less transparent analysis (using more CBI) and a less 
accurate but more transparent analysis (using less CBI).
---------------------------------------------------------------------------

    \354\ NHTSA-2018-0067-11741.

---------------------------------------------------------------------------

[[Page 24287]]

(2) Source Data and Vintage Used in the Analysis
    Based on the assumption that a publicly-available analysis fleet 
continued to be desirable, manufacturer compliance submissions to EPA 
and NHTSA were used as a starting point for the NPRM and final rule 
analysis fleets. Generally, manufacturer compliance submissions break 
down vehicle fuel economy and production volume by regulatory class, 
and include some very basic product information (typically including 
vehicle nameplate, engine displacement, basic transmission information, 
and drive configuration). Many different trim levels of a product are 
typically rolled up and reported in an aggregated fashion, and these 
groupings can make decomposition of different fuel-saving, road load 
reducing technologies extremely difficult. For instance, vehicles in 
different test weight classes, with different tires or aerodynamic 
profiles may be aggregated and reported together.\355\ A second portion 
of the compliance submission summarizes production volume by vehicle 
footprints (a key compliance measure for standard setting) by 
nameplate, and includes some basic information about engine 
displacement, transmission, and drive configuration. Often these 
production volumes by footprint do not fit seamlessly together with the 
production volumes for fuel economy, so the agencies must reconcile 
this information.
---------------------------------------------------------------------------

    \355\ Some fuel-economy compliance information for pickup trucks 
span multiple cab and box configurations, but manufacturers reported 
these disparate vehicles together.
---------------------------------------------------------------------------

    Information from the MY 2016 fleet was chosen as the foundation for 
the NPRM analysis fleet because, at the time the rulemaking analysis 
was initiated, the 2016 fleet represented the most up-to-date 
information available in terms of individual vehicle models and 
configurations, production technology levels, and production volumes. 
If MY 2017 data had been used while this analysis was being developed, 
the agencies would have needed to use product planning information that 
could not be made available to the public until a later date.
    The NPRM analysis fleet was initially developed with 2016 mid-model 
year compliance data because final compliance data was not available at 
that time, and the timing provided manufacturers the opportunity to 
review and comment on the characterization of their vehicles in the 
fleet. With a view toward developing an accurate characterization of 
the 2016 fleet to serve as an analytical starting point, corrections 
and updates to mid-year data (e.g., to production estimates) were 
sought, in addition to corroboration or correction of technical 
information obtained from commercial and other sources (to the extent 
that information was not included in compliance data), although future 
product planning information from manufacturers (e.g., future product 
offerings, products to be discontinued) was not requested, as most 
manufacturers view such information as CBI. Manufacturers offered a 
range of corrections to indicate engineering characteristics (e.g., 
footprint, curb weight, transmission type) of specific vehicle model/
configurations, as well as updates to fuel economy and production 
volume estimates in mid-year reporting. After following up on a case-
by-case basis to investigate significant differences, the analysis 
fleet was updated.
    Sales, footprint, and fuel economy values with final compliance 
data were also updated if that data was available. In a few cases, 
final production and fuel economy values were slightly different for 
specific MY 2016 vehicle models and configurations than were indicated 
in the NPRM analysis; however, other vehicle characteristics (e.g., 
footprint, curb weight, technology content) important to the analysis 
were reasonably accurate. While some commenters have, in the past, 
raised concerns that non-final CAFE compliance data is subject to 
change, the potential for change is likely not significant enough to 
merit using final data from an earlier model year reflecting a more 
outdated fleet. Moreover, even ostensibly final CAFE compliance data is 
frequently subject to later revision (e.g., if errors in fuel economy 
tests are discovered), and the purpose of the analysis was not to 
support enforcement actions but rather to provide a realistic 
assessment of manufacturers' potential responses to future standards.
    Manufacturers integrated a significant amount of new technology in 
the MY 2016 fleet, and this was especially true for newly-designed 
vehicles launched in MY 2016. While subsequent fleets will involve even 
further application of technology, using available data for MY 2016 
provided the most realistic detailed foundation for analysis that could 
be made available publicly in full detail, allowing stakeholders to 
reproduce the analysis presented in the proposal independently. Insofar 
as future product offerings are likely to be more similar to vehicles 
produced in 2016 than to vehicles produced in earlier model years, 
using available data regarding the 2016 model year provided the most 
realistic, publicly releasable foundation for constructing a forecast 
of the future vehicle market for this proposal. Many comments 
responding to the Draft TAR, EPA's Proposed Determination, EPA's 2017 
Request for Comment, and the NPRM preceding today's notice stated that 
the most up-to-date analysis fleet possible should be used, because a 
more up-to-date analysis fleet will better capture how manufacturers 
apply technology and will account better for vehicle model/
configuration introductions and deletions.356 357
---------------------------------------------------------------------------

    \356\ 82 FR 39551 (Aug. 21, 2017).
    \357\ For example, in 2016 comments to dockets EPA-HQ-OAR-2015-
0827 and NHTSA-2016-0068, the Alliance of Automobile Manufacturers 
commented that ``the Alliance supports the use of the most recent 
data available in establishing the baseline fleet, and therefore 
believes that NHTSA's selection [of, at the time, model year 2015] 
was more appropriate for the Draft TAR.'' Alliance at 82, Docket ID. 
EPA-HQ-OAR-2015-0827-4089. Global Automakers commented that ``a one-
year difference constitutes a technology change-over for up to 20% 
of a manufacturer's fleet. It was also generally understood by 
industry and the agencies that several new, and potentially 
significant, technologies would be implemented in MY 2015. The use 
of an older, outdated baseline can have significant impacts on the 
modeling of subsequent Reference Case and Control Case 
technologies.'' Global Automakers at A-10, Docket ID EPA-HQ-OAR-
2015-0827-4009.
---------------------------------------------------------------------------

    On the other hand, some commenters suggested that because 
manufacturers continue improving vehicle performance and utility over 
time, an older analysis fleet should be used to estimate how the fleet 
could have evolved had manufacturers applied all technological 
potential to fuel economy rather than continuing to improve vehicle 
performance and utility.\358\ Because manufacturers change and improve 
product offerings over time, conducting analysis with an older analysis 
fleet (or with a fleet using fuel economy levels and CO2 
emissions rates that have been adjusted to reflect an assumed return to 
levels of performance and utility typical of some past model year) 
would miss this real-world trend. While such an analysis could project 
what industry could do if, for example, manufacturers devoted all 
technological improvements toward raising fuel economy and reducing 
CO2 emissions (and if consumers decided to purchase these 
vehicles), the agencies do not believe it would be consistent with a 
transparent examination of what effects different levels of standards 
would have

[[Page 24288]]

on individual manufacturers and the fleet as a whole.
---------------------------------------------------------------------------

    \358\ For example, in 2016 comments to dockets EPA-HQ-OAR-2015-
0827 and NHTSA-2016-0068, UCS stated ``in modeling technology 
effectiveness and use, the agencies should use 2010 levels of 
performance as the baseline.'' UCS at 4, Docket ID. EPA-HQ-OAR-2015-
0827-4016.
---------------------------------------------------------------------------

    All else being equal, using a newer analysis fleet will produce 
more realistic estimates of impacts of potential new standards than 
using an outdated analysis fleet. However, among relatively current 
options, a balance must be struck between input freshness, and input 
completeness and accuracy.\359\ During assembly of the inputs for the 
NPRM analysis, final compliance data was available for the MY 2015 
model year but not, in a few cases, for MY 2016. However, between mid-
year compliance information and manufacturers' specific updates 
discussed above, a robust and detailed characterization of the MY 2016 
fleet was developed. While information continued to develop regarding 
the MY 2017 and, to a lesser extent MY 2018 and even MY 2019 fleets, 
this information was--even in mid-2017--too incomplete and inconsistent 
to be assembled with confidence into an analysis fleet for modeling 
supporting deliberations regarding the NPRM analysis.
---------------------------------------------------------------------------

    \359\ Comments provided through a recent peer review of the CAFE 
model recognize the competing interests behind this balance. For 
example, referring to NHTSA's 2016 Draft TAR analysis, one of the 
peer reviewers commented as follows: ``The NHTSA decision to use MY 
2015 data is wise. In the TAR they point out that a MY 2016 
foundation would require the use of confidential data, which is less 
desirable. Clearly they would also have a qualitative vision of the 
MY 2016 landscape while employing MY 2015 as a foundation. Although 
MY 2015 data may still be subject to minor revision, this is 
unlikely to impact the predictive ability of the model . . . A more 
complex alternative approach might be to employ some 2016 changes in 
technology, and attempt a blend of MY 2015 and MY 2016, while 
relying of estimation gained from only MY 2015 for sales. This 
approach may add some relevancy in terms of technology, but might 
introduce substantial error in terms of sales.''
---------------------------------------------------------------------------

    Manufacturers requested that the baseline fleet supporting the 
final rule incorporate the MY 2018 or most recent information 
available.\360\ Other commenters expressed desire for multiple fleets 
of various vintages to compare the updated model outputs with those of 
previous rule-makings. Specifically, some commenters requested that 
older fleet vintages (MY 2010, for instance) be developed in parallel 
with the MY 2017 fleet so that those too may be used as inputs for the 
model.\361\
---------------------------------------------------------------------------

    \360\ NHTSA-2018-0067-12150, Toyota North America.
    \361\ NHTSA-2018-0067-11741, ICCT.
---------------------------------------------------------------------------

    Between the NPRM and this final rule, manufacturers submitted final 
compliance data for the MY 2017 fleet. When the agencies pulled 
together information for the fleet for the final rule, the agencies 
decided to use the highest-quality, most up-to-date information 
available. Given that pulling this information together takes some 
time, and given that ``final'' compliance submissions often lag 
production by a few years, the agencies decided to use 2017 model year 
as the base year for the analysis fleet, as the agencies stated in the 
NPRM.\362\ While the agencies could have used preliminary 2018 data or 
even very early 2019 data, this information was not available in time 
to support the final rulemaking. Likewise, the agencies chose not to 
revert to a previous model year (for instance 2016 or 2012) because 
many manufacturers have incorporated fuel savings technologies over the 
last few years, realized some benefits for fuel economy, and adjusted 
the performance or sales mix of vehicles to remain competitive in the 
market. Also, using an earlier model year would provide less accurate 
projections because the analysis would be based on what manufacturers 
could have done in past model years and would have estimated the fuel 
economy improvements instead of using known information on the 
technologies that were employed and the actual fuel economy that 
resulted from applying those technologies.
---------------------------------------------------------------------------

    \362\ 83 FR 43006 (``If newer compliance data (i.e., MY 2017) 
becomes available and can be analyzed during the pendency of this 
rulemaking, and if all other necessary steps can be performed, the 
analysis fleet will be updated, as feasible, and made publicly 
available.'').
---------------------------------------------------------------------------

    Some additional information (about off-cycle technologies, for 
instance) was often not reported by manufacturers in MY 2017 formal 
compliance submissions in a way that provided clear information on 
which technologies were included on which products. As part of the 
formal compliance submission, some manufacturers voluntarily submitted 
additional information (about engine technologies, for instance). While 
this data was generally of very high quality, there were some mistakes 
or inconsistencies with publicly available information, causing the 
agencies to contact the manufacturers to understand and correct 
identified issues. In most cases, however, the formal compliance data 
was very limited in nature, and the agencies collected additional 
information necessary to characterize fully the fleet from other 
sources, and scrutinized additional information submitted by 
manufacturers carefully, independently verifying when possible.
    Specifically, the agencies downloaded and reviewed numerous 
marketing brochures and product launch press releases to confirm 
information submitted by manufacturers and to fill in information 
necessary for the analysis fleet that was not provided in the 
compliance data. Product brochures often served as the basis for the 
curb weights used in the analysis. This publicly available manufacturer 
information sometimes also included aerodynamic drag coefficients, 
information about steering architecture, start-stop systems, pickup bed 
lengths, fuel tank capacities, and high-voltage battery capacities. The 
agencies recorded vehicle horsepower, compression ratio, fuel-type, and 
recommended oil weight rating from a combination of manufacturer 
product brochures and owner's manuals. The product brochures, as well 
as online references such as Autobytel, informed which combinations of 
fuel saving technologies were available on which trim levels, and what 
the manufacturer suggested retail price was for many products. Overall 
this information proved helpful for assigning technologies to vehicles, 
and for getting data (such as fuel tank size \363\) necessary for the 
analysis. These reference materials have been included in the 
rulemaking documentation.\364\
---------------------------------------------------------------------------

    \363\ The quality of data for today's analysis fleet is notably 
improved for fuel tank capacity, which factors into the calculation 
of refueling time benefits. In many previous analyses, fuel tank 
sizes were often stated as estimates or proxies, and not sourced so 
carefully.
    \364\ Publicly available data used to supplement analysis fleet 
information is available in the docket.
---------------------------------------------------------------------------

    The agencies elected not to develop fleets of previous model year 
vintages that could be used in parallel as an input to the CAFE model. 
Developing a detailed characterization of the fleet of any vintage 
would be a huge undertaking with few benefits. As the scope has 
increased, and as additional modules are added, going back in time to 
re-characterize a previous fleet in a format that works with CAFE model 
updates can be time- and resource-prohibitive for the agencies, even if 
that work is adapting a fleet that was used in previous rule-making 
analysis. Doing so also offers little value in determining what 
potential fuel saving technology can be added to a more recent fleet 
during the rulemaking timeframe.
    The MY 2017 manufacturer-submitted data, verified and supplemented 
by the agencies with publicly-available information, therefore 
presented the fullest, most up-to-date data set that the agencies could 
have used to support this analysis.

[[Page 24289]]

b) Characterizing Vehicles and Their Technology Content
    The starting point for projecting what additional fuel economy 
improving technologies could feasibly be applied to vehicles is knowing 
what vehicles are produced by which manufacturers and what technologies 
exist on those vehicles. Rows in the market data file are the smallest 
portion of the fleet to which technology may be applied as part of a 
projected compliance pathway. For the analysis presented in this final 
rule, the agencies, when possible, attempted to include vehicle trim 
level information in discrete rows. A manufacturer, for example GM, may 
produce one or more vehicle makes (or brands), for example Chevrolet, 
Buick and others. Each vehicle make may offer one or more vehicle 
models, for example Malibu, Traverse and others. And each vehicle model 
may be available in one or more trim levels (or standard option 
levels), for example ``RS,'' ``Premier'' and others, which have 
different levels of standard options, and in some cases, different 
engines and transmissions.
    Manufacturer compliance submissions, discussed above, were used as 
a starting point to define working rows in the market data file; 
however, often the rows needed to be further disaggregated to correctly 
characterize vehicle information covered in the scope of the analysis, 
and analysis fleet. Manufacturers often grouped vehicles with multiple 
trim levels together because they often included the same fuel-saving 
technologies and may be aggregated to simplify reporting. However, the 
manufacturer suggested retail prices of different trim levels are 
certainly different, and other features relevant to the analysis are 
occasionally different.
    As a result of further disaggregating compliance information, the 
number of rows in the market data file increased from 1,667 rows used 
in the NPRM to 2,952 rows for this final rule analysis. The agencies do 
not have data on sales volumes for each nameplate by trim level, and 
used an approach that evenly distributed volume across offered trim 
levels, within the defined constraints of the compliance data.\365\ 
Evenly distributing the volume across trim levels is a simplification, 
but this action should (1) highlight some difficulties that could be 
encountered when acquiring data for a full-vehicle consumer choice 
model should the agencies pursue developing one in the future 
(discussed further, below), and (2) lower the average sales volume per 
row in the market data file, thereby allowing the application of very 
advanced electrification technologies in smaller lumps. The latter 
effect is responsive to comments (discussed below) that suggested 
electrification technologies could be more cost-effectively deployed in 
lower volumes, and that the CAFE model artificially constrains cost 
effective technologies that may be deployed, resulting in higher costs 
and large over-compliance.
---------------------------------------------------------------------------

    \365\ The sum of volumes by nameplate configuration, for fuel 
economy value, and for footprint value remains the same.
---------------------------------------------------------------------------

(1) Assigning Vehicle Technology Classes
    While each vehicle in the analysis fleet has its list of observed 
technologies and equipment, the ways in which manufacturers apply 
technologies and equipment do not always coincide perfectly with how 
the analysis characterizes the various technologies that improve fuel 
economy and reduce CO2 emissions. To improve how the 
observed vehicle fleet ``fits into'' the analysis, each vehicle model/
configuration is ``mapped'' to the full-vehicle simulation modeling by 
Argonne National Laboratory that is used to estimate the effectiveness 
of the fuel economy-improving/CO2 emissions-reducing 
technologies considered. Argonne produces full-vehicle simulation 
modeling for many combinations of technologies, on many types of 
vehicles, but it did not simulate literally every single manufacturer's 
vehicle model/configuration in the analysis fleet because it would be 
impractical to assemble the requisite detailed information--much of 
which would likely only be provided on a confidential basis--specific 
to each vehicle model/configuration and because the scale of the 
simulation effort would correspondingly increase by at least two orders 
of magnitude. Instead, Argonne simulated 10 different vehicle types 
corresponding to the ``technology classes'' generally used in CAFE 
analysis over the past several rulemakings (e.g., small car, small 
performance car, pickup truck, etc.). Each of those 10 different 
vehicle types was assigned a set of ``baseline characteristics'' to 
which Argonne added combinations of fuel-saving technologies and then 
ran simulations to determine the fuel economy achieved when applying 
each combination of technologies to that vehicle type given its 
baseline characteristics.
BILLING CODE 4910-59-P

[[Page 24290]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.089

BILLING CODE 4910-59-C
    In the analysis fleet, inputs assign each specific vehicle model/
configuration to a technology class, and once there, map to the 
simulation within that technology class most closely matching the 
combination of observed technologies and equipment on that vehicle. 
This mapping to a specific simulation result most closely representing 
a given vehicle model/configuration's initial technology ``state'' 
enables the CAFE model to estimate the same vehicle model/
configuration's fuel economy after application of some other 
combination of technologies, leading to an alternative technology 
state.
(2) Assigning Vehicle Technology Content
    As explained above, the analysis fleet is defined not only by the 
vehicles it contains, but also by the technologies on those vehicles. 
Each vehicle in the analysis fleet has an associated list of observed 
technologies and equipment that can improve fuel economy and reduce 
CO2 emissions.\366\ With a portfolio of descriptive 
technologies arranged by manufacturer and model, the analysis fleet can 
be summarized and project how vehicles in that fleet may increase fuel 
economy over time via the application of additional technology.
---------------------------------------------------------------------------

    \366\ These technologies are generally grouped into the 
following categories: Vehicle technologies include mass reduction, 
aerodynamic drag reduction, low rolling resistance tires, and 
others. Engine technologies include engine attributes describing 
fuel type, engine aspiration, valvetrain configuration, compression 
ratio, number of cylinders, size of displacement, and others. 
Transmission technologies include different transmission 
arrangements like manual, 6-speed automatic, 10-speed automatic, 
continuously variable transmission, and dual-clutch transmissions. 
Hybrid and electric powertrains may complement traditional engine 
and transmission designs or replace them entirely.
---------------------------------------------------------------------------

    In many cases, vehicle technology is clearly observable from the 
2017 compliance data (e.g., compliance data indicates clearly which 
vehicles have turbochargers and which have continuously variable 
transmissions), but in some cases technology levels are less 
observable. For the latter, like levels of mass reduction, the analysis 
categorized levels of technology already used in a given vehicle. 
Similarly, engineering judgment was used to determine if higher mass 
reduction levels may be used practicably and safely for a given 
vehicle.
    Either in mid-year compliance data for MY 2016, final compliance 
data for MY 2017, or separately and at the agencies' invitation prior 
to the NPRM or in comments in responses to the NPRM, most manufacturers 
provided guidance on the technology already present in each of their 
vehicle model/configurations. This information was not as complete for 
all manufacturers' products as needed for the analysis, so, in some 
cases, information was supplemented with publicly available data, 
typically from manufacturer media sites. In limited cases, 
manufacturers did not supply information, and

[[Page 24291]]

information from commercial and publicly available sources was used.
    The agencies continued to evaluate emerging technologies in the 
analysis. In response to comments,\367\ and given recent product 
launches for MY 2020, and some very recently announced future product 
offerings, the agencies elevated some technologies that were discussed 
in the NPRM to the compliance simulation. As a result, several 
additional engine technologies, expanded levels of mass reduction 
technology, and some additional combinations of engines with plug-in 
hybrid, or strong hybrid technology are available in the compliance 
pathways for the final rule analysis.
---------------------------------------------------------------------------

    \367\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    In addition, some redundant technologies, or technologies that were 
inadvertently represented on the technology tree as being available to 
be applied twice, have been consolidated. For instance, previous basic 
versions of engine friction reduction were layered on top of basic 
engine maps, but the efficiency in many modern engine maps already 
include the benefits of that engine friction reduction technology. The 
following Table VI-8 lists the technologies considered in the final 
rule analysis, with the data sources used to map those technologies to 
vehicles in the analysis fleet.
BILLING CODE 4910-59-P

[[Page 24292]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.090


[[Page 24293]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.091


[[Page 24294]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.092

BILLING CODE 4910-59-C
    Industry commenters generally stated the MY 2016 baseline 
technology content presented in the NPRM as an improvement over 
previous analyses because it more accurately accounted for technology 
already used in the fleet.368 369 In contrast, some 
commenters expressed preference for EPA's baseline technology 
assignment assumptions presented in the Draft TAR for mass reduction, 
tire rolling resistance, and aerodynamic drag because those assumptions 
projected very few technology improvements were present in the baseline 
fleet. In assessing the comments, the agencies found that

[[Page 24295]]

using the EPA Draft TAR approach would lead to projected compliance 
pathways with overestimated fuel economy improvements and 
underestimated costs.\370\
---------------------------------------------------------------------------

    \368\ NHTSA-2018-0067-12073, Alliance of Automobile 
Manufacturers.
    \369\ NHTSA-2018-0067-12150, Toyota North America.
    \370\ NHTSA-2018-0067-11741, ICCT.
---------------------------------------------------------------------------

    Many of those assumptions were neither scientifically meritorious, 
nor isolated examples. For instance, for the EPA Draft TAR and Proposed 
Determination analyses, the BMW i3, a vehicle with full carbon fiber 
bodysides and downsized, mass-reduced wheels and tires (some of the 
most advanced mass reducing technologies commercialized in the 
automotive industry), was assumed to have 1.0 percent mass reduction (a 
very minor level of mass reduction). Similarly, previous analyses 
assigned the Chevrolet Corvette, a performance vehicle that has long 
been a platform for commercializing advanced weight saving 
technologies,\371\ with zero mass reduction. For aerodynamic drag, 
previous EPA analysis assumed that pickup trucks could achieve the 
aerodynamic drag profile typical of a sedan, with little regard for 
form drag constraints or frontal area (and headroom, or ground 
clearance) considerations. These assumptions commonly led to 
projections of a 20 percent improvement in mass, aerodynamic drag, and 
tire rolling resistance, even when a large portion of those 
improvements had either already been implemented, or were not 
technologically feasible. On the other hand, in the Draft TAR, NHTSA 
presented methodologies to evaluate content for mass reduction 
technology, aerodynamic drag improvements, and rolling resistance 
technologies that better accounted for the actual level of technologies 
in the analysis fleet. Throughout the rulemaking process, the agencies 
reconciled these differences, jointly presented improved approaches in 
the NPRM similar to what NHTSA presented in the Draft TAR, and again 
used those reconciled approaches in today's analysis.\372\
---------------------------------------------------------------------------

    \371\ See, e.g., Fiberglass to Carbon Fiber: Corvette's 
Lightweight Legacy, GM (August 2012), https://media.gm.com/media/us/en/gm/news.detail.html/content/Pages/news/us/en/2012/Aug/0816_corvette.html.
    \372\ Because these road load technologies are no longer double 
counted, the projected compliance pathway in the NPRM, and in 
today's analysis for stringent alternatives, often requires more 
advanced fuel saving technologies than previously projected, 
including higher projected penetration rates of hybrid and electric 
vehicle technologies.
---------------------------------------------------------------------------

    Many commenters correctly observed that the analysis fleet in the 
NPRM recognized more technology content in the baseline than in the 
Draft TAR (with higher penetration rates of tire rolling resistance and 
aerodynamic drag improvements, for instance), but also that the fuel 
economy values of the fleet had not improved all that much from the 
previous year. Some commenters concluded that the NPRM baseline 
technology assignment process was arbitrary and overstated the 
technology content already present in the baseline 
fleet.373 374 The agencies agree that there was a large 
increase in the amount of road load technology credited in the baseline 
fleet between EPA's Draft TAR and the jointly produced NPRM, and 
clarify that this change was largely due to a recognition of 
technologies that were actually present in the fleet, but not properly 
accounted for in previous analyses. The change in penetration rates of 
road load technologies (after accounting for glider share updates, 
which is discussed in more detail in the mass reduction technology 
section) between the NPRM and today's analysis is relatively small.
---------------------------------------------------------------------------

    \373\ NHTSA-2018-0067-11741, ICCT.
    \374\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
---------------------------------------------------------------------------

    Many commenters noted that the different baseline road load 
assumptions (and other technology modeling) materially affect 
compliance pathways, and projected costs.\375\ ICCT commented that the 
agencies should conduct sensitivity analyses assuming every vehicle in 
the analysis fleet is set to zero percent road load technology 
improvement, to demonstrate how the technology content of the analysis 
fleet affected the compliance scenarios.\376\
---------------------------------------------------------------------------

    \375\ NHTSA-2018-0067-11928, Ford Motor Company.
    \376\ NHTSA-2018-0067-11741, ICCT.
---------------------------------------------------------------------------

    While the agencies have clearly described the methods by which 
initial road load technologies are assigned in Section VI.C.4 Mass 
Reduction, Section VI.C.5 Aerodynamics, and Section VI.C.6 Tire Rolling 
Resistance below, the agencies considered a sensitivity case that 
assumed no mass reduction, rolling resistance, or aerodynamic 
improvements had been made to the MY 2017 fleet (i.e., setting all 
vehicle road levels to zero--MRO, AERO and ROLL0). While this is an 
unrealistic characterization of the initial fleet, the agencies 
conducted a sensitivity analysis to understand any affect it may have 
on technology penetration along other paths (e.g. engine and hybrid 
technology). Under the CAFE program, the sensitivity analysis shows a 
slight decrease in reliance on engine technologies (HCR engines, 
turbocharge engines, and engines utilizing cylinder deactivation) and 
hybridization (strong hybrids and plug-in hybrids) in the baseline 
(relative to the central analysis). The consequence of this shift to 
reliance on lower-level road load technologies is a reduction in 
compliance cost in the baseline of about $300 per vehicle (in MY 2026). 
As a result, cost savings in the preferred alternative are reduced by 
about $200 per vehicle. Under the CO2 program, the general 
trend in technology shift is less dramatic (though the change in BEVs 
is larger) than the CAFE results. The cost change is also comparable, 
but slightly smaller ($200 per vehicle in the baseline) than the CAFE 
program results. Cost savings under the preferred alternative are 
further reduced by about $100. With the lower technology costs in all 
cases, the consumer payback periods decreased as well. These results 
are consistent with the approach taken by manufacturers who have 
already deployed many of the low-level road load reduction 
opportunities to improve fuel economy.
    Some commenters preferred that the agencies develop a different 
methodology based on reported road load coefficients (``A,'' ``B'' and 
``C'' coastdown coefficients) to estimate levels of aerodynamic drag 
improvement and rolling resistance in the baseline fleet that did not 
rely on CBI.\377\ The agencies considered this, but determined that 
using CBI to assign baseline aerodynamic drag levels and rolling 
resistance values was more accurate and appropriate. Estimating 
aerodynamic drag levels and rolling resistance levels from coastdown 
coefficients is not straightforward, and to do it well would require 
information the agencies do not have (much of which is also CBI). For 
instance, rotational inertias of wheel, tire, and brake packages can 
affect coastdown, so mass of the vehicle is not sufficient. The frontal 
area of the vehicles, a key component for calculating aerodynamic drag, 
is rarely known, and often requires manufacturer input to get an 
accurate value. Other important vehicle features like all-wheel-drive 
should also be accounted for, and the agencies would struggle to 
correctly identify improvements in rolling resistance, low-drag brakes, 
and secondary axle disconnect, because all of these technologies would 
present similar signature on a coast down test. All of these 
technologies are represented as technology pathways in today's 
analysis. Manufacturers acknowledged the possibility of using road load 
coefficients to estimate rolling resistance and aerodynamic features, 
but warned that the process ``required

[[Page 24296]]

various assumptions and is not very accurate,'' and stated that the use 
of CBI to assess aerodynamic and rolling resistance technologies is an 
``accurate and practical solution'' to assign these difficult to 
observe technologies.\378\
---------------------------------------------------------------------------

    \377\ NHTSA-2018-0067-11741, ICCT.
    \378\ NHTSA-2018-0067-12073, Alliance of Automobile 
Manufacturers.
---------------------------------------------------------------------------

(3) Assigning Engine Configurations
    Engine technology costs can vary significantly by the configuration 
of the engine. For instance, adding variable valve lift to each 
cylinder on an engine would cost more for an engine with eight 
cylinders than an engine with four cylinders. Similarly, the cost of 
adding a turbocharger to an engine and downsizing the engine would be 
different going from a naturally aspirated V8 to a turbocharged V6 than 
going from a naturally aspirated V6 to a turbocharged I4. As discussed 
in detail in the engine technology section of this document, the cost 
files for the CAFE model account for instances such as these examples.
    Information in the analysis fleet enables the CAFE model to 
reference the intended engine costs. The ``Engine Technology Class 
(Observed)'' lists the architecture of the observed engine. Notably, 
the analysis assumes that nearly all turbo charged engines take 
advantage of downsizing to optimize fuel efficiency, minimize the cost 
of turbo charging, and to maintain performance (to the extent 
practicable) with the naturally aspirated counterpart engine. 
Therefore, engines observed in the fleet that have already been down-
sized must reference costs for a larger basic engine, which assumes 
down-sizing with the application of turbo technology. In these cases, 
the ``Engine Technology Class'' which is used to reference costs will 
be larger than the ``Engine Technology Class (Observed).''
    This is the same process agencies used in the NPRM, and it corrects 
a previous error in the Draft TAR analysis, which incorrectly 
underestimated turbocharged engine costs.\379\ Some commenters 
expressed confusion and disagreement with this correction, with some 
even commenting that the analysis baselessly inflated costs of 
turbocharging technologies between the Draft TAR and the NPRM.\380\ To 
be clear, this was a correction so that the costs used to calculate 
turbocharged engine costs accurately reflected the total costs for a 
turbocharged engine.
---------------------------------------------------------------------------

    \379\ For instance, the Draft TAR engine costs would map an 
observed V6 Turbo engine to I4 Turbo engine costs, by referencing a 
4C1B engine cost.
    \380\ NHTSA-2018-0067-11741, ICCT.
---------------------------------------------------------------------------

(4) Characterizing Shared Vehicle Platforms, Engines, and Transmissions
    Another aspect of characterizing vehicle model/configurations in 
the analysis fleet is based on whether they share a ``platform'' with 
other vehicle model/configurations. A ``platform'' refers to engineered 
underpinnings shared on several differentiated products. Manufacturers 
share and standardize components, systems, tooling, and assembly 
processes within their products (and occasionally with the products of 
another manufacturer) to manage complexity and costs for development, 
manufacturing, and assembly.
    The concept of platform sharing has evolved over time. Years ago, 
manufacturers rebadged vehicles and offered luxury options only on 
premium nameplates (and manufacturers shared some vehicle platforms in 
limited cases). Today, manufacturers share parts across highly 
differentiated vehicles with different body styles, sizes, and 
capabilities that may share the same platform. For instance, the Honda 
Civic and Honda CR-V share many parts and are built on the same 
platform. Engineers design chassis platforms with the ability to vary 
wheelbase, ride height, and even driveline configuration. Assembly 
lines can produce hatchbacks and sedans to cost-effectively utilize 
manufacturing capacity and respond to shifts in market demand. Engines 
made on the same line may power small cars or mid-size sport utility 
vehicles. In addition, although the agencies' analysis, like past CAFE 
analyses, considers vehicles produced for sale in the U.S., the agency 
notes these platforms are not constrained to vehicle models built for 
sale in the U.S.; many manufacturers have developed, and use, global 
platforms, and the total number of platforms is decreasing across the 
industry. Several automakers (for example, General Motors and Ford) 
either plan to, or already have, reduced their number of platforms to 
less than 10 and account for the overwhelming majority of their 
production volumes on that small number of platforms.
    Vehicle model/configurations derived from the same platform are so 
identified in the analysis fleet. Many manufacturers' use of vehicle 
platforms is well documented in the public record and widely recognized 
among the vehicle engineering community. Engineering knowledge, 
information from trade publications, and feedback from manufacturers 
and suppliers was also used to assign vehicle platforms in the analysis 
fleet.
    When the CAFE model is deciding where and how to add technology to 
vehicles, if one vehicle on the platform receives new technology, other 
vehicles on the platform also receive the technology as part of their 
next major redesign or refresh.\381\ Similar to vehicle platforms, 
manufacturers create engines that share parts. For instance, 
manufacturers may use different piston strokes on a common engine 
block, or bore out common engine block castings with different 
diameters to create engines with an array of displacements. Head 
assemblies for different displacement engines may share many components 
and manufacturing processes across the engine family. Manufacturers may 
finish crankshafts with the same tools to similar tolerances. Engines 
on the same architecture may share pistons, connecting rods, and the 
same engine architecture may include both six and eight cylinder 
engines. One engine family may appear on many vehicles on a platform, 
and changes to that engine may or may not carry through to all the 
vehicles. Some engines are shared across a range of different vehicle 
platforms. Vehicle model/configurations in the analysis fleet that 
share engines belonging to the same platform are also identified as 
such.
---------------------------------------------------------------------------

    \381\ The CAFE model assigns mass reduction technology at a 
platform level, but many other technologies may be assigned and 
shared at a vehicle nameplate or vehicle model level.
---------------------------------------------------------------------------

    It is important to note that manufacturers define common engines 
differently. Some manufacturers consider engines as ``common'' if the 
engines shared an architecture, components, or manufacturing processes. 
Other manufacturers take a narrower definition, and only assume 
``common'' engines if the parts in the engine assembly are the same. In 
some cases, manufacturers designate each engine in each application as 
a unique powertrain. For example, a manufacturer may have listed two 
engines separately for a pair that share designs for the engine block, 
the crank shaft, and the head because the accessory drive components, 
oil pans, and engine calibrations differ between the two. In practice, 
many engines share parts, tooling, and assembly resources, and 
manufacturers often coordinate design updates between two similar 
engines. Engine families, designated in the analysis using ``engine 
codes,'' for each manufacturer were tabulated and assigned based on 
data-driven criteria. If engines shared a common cylinder count and 
configuration, displacement, valvetrain, and fuel type, those engines

[[Page 24297]]

may have been considered together. In addition, if the compression 
ratio, horsepower, and displacement of engines were only slightly 
different, those engines were considered the same for the purposes of 
redesign and sharing.
    Vehicles in the analysis fleet with the same engine family will, 
therefore, adopt engine technology in a coordinated fashion. 
Specifically, if such vehicles have different design schedules (i.e., 
refresh and redesign schedules), and a subset of vehicles using a given 
engine add engine technologies during of a redesign or refresh that 
occurs in an early model year (e.g., 2018), other vehicles using the 
same engine ``inherit'' these technologies at the soonest ensuing 
refresh or redesign. This is consistent with a view that, over time, 
most manufacturers are likely to find it more practicable to shift 
production to a new version of an engine than to continue production of 
both the new engine and a ``legacy'' engine indefinitely. By grouping 
engines together, the CAFE model controls future engine families to 
ensure reasonable powertrain complexity. This means, however, that for 
manufacturers that submitted highly atomized engine and transmission 
portfolios, there is a practical cap on powertrain complexity and the 
ability of the manufacturer to optimize the displacement of (i.e., 
``right size'') engines perfectly for each vehicle configuration. This 
concept is discussed further in Section VI.B.4.a), below.
    Like with engines, manufacturers often use transmissions that are 
the same or similar on multiple vehicles. Manufacturers may produce 
transmissions that have nominally different machining to castings, or 
manufacturers may produce transmissions that are internally identical, 
except for the final gear ratio. In some cases, manufacturers sub-
contract with suppliers that deliver whole transmissions. In other 
cases, manufacturers form joint ventures to develop shared 
transmissions, and these transmission platforms may be offered in many 
vehicles across manufacturers. Manufacturers use supplier and joint-
venture transmissions to a greater extent than they do with engines. To 
reflect this reality, shared transmissions were considered for 
manufacturers as appropriate. Transmission configurations are referred 
to in the analysis as ``transmission codes.'' Like the inheritance 
approach outlined for engines, if one vehicle application of a shared 
transmission family upgraded the transmission, other vehicle 
applications also upgraded the transmission at the next refresh or 
redesign year. To define common transmissions, the agencies considered 
transmission type (manual, automatic, dual-clutch, continuously 
variable), number of gears, and vehicle architecture (front-wheel-
drive, rear-wheel-drive, all-wheel-drive based on a front-wheel drive 
platform, or all-wheel-drive based on a rear-wheel-drive platform). If 
vehicles shared these attributes, these transmissions were grouped for 
the analysis. Vehicles in the analysis fleet with the same transmission 
configuration will adopt transmission technology together, as described 
above.
    Having all vehicles that share a platform (or engines that are part 
of a family) adopt fuel economy-improving/CO2 emissions-
reducing technologies together, subject to refresh/redesign 
constraints, reflects the real-world considerations described above, 
but also overlooks some decisions manufacturers might make in the real 
world in response to market pull. Accordingly, even though the analysis 
fleet is incredibly complex, it is also over-simplified in some 
respects compared to the real world. For example, the CAFE model does 
not currently attempt to simulate the potential for a manufacturer to 
shift the application of technologies to improve performance rather 
than fuel economy. Therefore, the model's representation of the 
``inheritance'' of technology can lead to estimates a manufacturer 
might eventually exceed fuel economy standards as technology continues 
to propagate across shared platforms and engines. While the agencies 
have previously seen examples of extended periods during which some 
manufacturers exceeded one or both CAFE and/or CO2 
standards, in plenty of other examples, manufacturers chose to 
introduce (or even reintroduce) technological complexity into their 
vehicle lineups in response to buyer preferences. Going forward, and 
recognizing the recent trend for consolidating platforms, it seems 
likely manufacturers will be more likely to choose efficiency over 
complexity in this regard; therefore, the potential should be lower 
that today's analysis turns out to be oversimplified compared to the 
real world.
    Manufacturers described shared engines, transmissions, and vehicle 
platforms as ``standard business practice'' and they were encouraged 
that the NHTSA analysis in the Draft TAR, and the jointly issued NPRM 
placed realistic limits on the number of unique engines and 
transmissions in a powertrain portfolio.\382\ In previous rulemakings, 
stakeholders pointed out that shared parts and portfolio complexity 
should be considered (but were not), and that the proliferation of 
unique technology combinations resulting from unconstrained compliance 
pathways would jeopardize economies of scale in the real world.\383\
---------------------------------------------------------------------------

    \382\ NHTSA-2018-0067-12150, Toyota North America.
    \383\ Alliance of Automobile Manufacturers, EPA-HQ-OAR-0827 and 
NHTSA-2016-0068.
---------------------------------------------------------------------------

    HD Systems acknowledged that previous rulemakings did not 
appropriately consider part sharing, but contended that in today's 
global marketplace, manufacturers have flexibility to compete in new 
ways that break old part sharing rules.\384\ The agencies acknowledge 
that some transmissions are now sourced through suppliers, and that 
economies of scale could, in the future be achieved at an industry 
level instead of a manufacturer level; however, even when manufacturers 
outsource a transmission, recent history suggests they apply that 
transmission to multiple vehicles to control assembly plant and service 
parts complexity, as they would if they were making the transmission 
themselves. Similarly, even for global platforms, or global 
powertrains, there is little evidence that manufacturers fragment 
powertrain line-ups for a vehicle, or a set of vehicles that have 
typically used the same engine. The agencies will continue to consider 
how to capture more accurately the ways vehicles share engines, 
transmissions, and platforms in future rulemakings, but the part-
sharing and modeling approach presented in the NPRM and this final rule 
represents a marked improvement over previous analysis.
---------------------------------------------------------------------------

    \384\ NHTSA-2018-0067-11985, HD Systems.
---------------------------------------------------------------------------

(5) Characterizing Production Design Cycles
    Another aspect of characterizing vehicles in the analysis fleet is 
based on when they can next be refreshed or redesigned. Redesign 
schedules play an important role in determining when new technologies 
may be applied. Many technologies that improve fuel economy and reduce 
CO2 emissions may be difficult to incorporate without a 
major product redesign. Therefore, each vehicle model in the analysis 
fleet has an associated redesign schedule, and the CAFE model uses that 
schedule to implement significant advances in some technologies (like 
major mass reduction) to redesign years, while allowing manufacturers 
to include minor advances (such as improved tire rolling

[[Page 24298]]

resistance) during a vehicle ``refresh,'' or a smaller update made to a 
vehicle, which can happen between redesigns. In addition to refresh and 
redesign schedules associated with vehicle model/configurations, 
vehicles that share a platform subsequently have platform-wide refresh 
and redesign schedules for mass reduction technologies.
    To develop the refresh/redesign cycles used for the NPRM vehicles 
in the analysis fleet, information from commercially available sources 
was used to project redesign cycles through MY 2022, as was done for 
NHTSA's analysis for the 2016 Draft TAR.\385\ Commercially available 
sources' estimates through MY 2022 are generally supported by detailed 
consideration of public announcements plus related intelligence from 
suppliers and other sources, and recognize that uncertainty increases 
considerably as the forecasting horizon is extended. For MYs 2023-2035, 
in recognition of that uncertainty, redesign schedules were extended 
considering past pacing for each product, estimated schedules through 
MY 2022, and schedules for other products in the same technology 
classes. As mentioned above, potentially confidential forward-looking 
information was not requested from manufacturers; nevertheless, all 
manufacturers had an opportunity to review the estimates of product-
specific redesign schedules. A few manufacturers provided related 
forecasts and, for the most part, that information corroborated the 
estimates.
---------------------------------------------------------------------------

    \385\ In some cases, data from commercially available sources 
was found to be incomplete or inconsistent with other available 
information. For instance, commercially available sources identified 
some newly imported vehicles as new platforms, but the international 
platform was midway through the product lifecycle. While new to the 
U.S. market, treating these vehicles as new entrants would have 
resulted in artificially short redesign cycles if carried forward, 
in some cases. Similarly, commercially available sources labeled 
some product refreshes as redesigns, and vice versa. In these 
limited cases, the data was revised to be consistent with other 
available information or typical redesign and refresh schedules for 
CAFE modeling. In these limited cases, the forecast time between 
redesigns and refreshes was updated to match the observed past 
product timing.
---------------------------------------------------------------------------

    Some commenters suggested supplanting these estimated redesign 
schedules with estimates applying faster cycles (e.g., four to five 
years), and this approach was considered for the analysis. Some 
manufacturers tend to operate with faster redesign cycles and may 
continue to do so, and manufacturers tend to redesign some products 
more frequently than others. However, especially considering that 
information presented by manufacturers largely supports estimates 
discussed above, applying a ``one size fits all'' acceleration of 
redesign cycles would not improve the analysis; instead, assuming a 
fixed, shortened redesign schedule across the industry would likely 
reduce consistency with the real world, especially for light trucks, 
which are redesigned, on average, no less than every six years (see 
Table VI-9, below). Moreover, if some manufacturers accelerate 
redesigns in response to new standards, doing so would likely involve 
costs (greater levels of stranded capital, reduced opportunity to 
benefit from ``learning''-related cost reductions) greater than 
reflected in other inputs to the analysis.
    As discussed in the NPRM, manufacturers use diverse strategies with 
respect to when, and how often they update vehicle designs. While most 
vehicles have been redesigned sometime in the last five years, many 
vehicles have not. In particular, vehicles with lower annual sales 
volumes tend to be redesigned less frequently, perhaps giving 
manufacturers more time to recoup the investment needed to bring the 
product to market. In some cases, manufacturers continue to produce and 
sell vehicles designed more than a decade ago.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.093


[[Page 24299]]


    Each manufacturer may use different strategies throughout their 
product portfolio, and a component of each strategy may include the 
timing of refresh and redesign cycles. Table VI-10 summarizes the 
average time between redesigns, by manufacturer, by vehicle technology 
class. Dashes mean the manufacturer has no volume in that vehicle 
technology class in the MY 2017 analysis fleet. Across the industry, 
manufacturers average 6.6 years between product redesigns.
[GRAPHIC] [TIFF OMITTED] TR30AP20.094

    Trends on redesign schedules identified in the NPRM remain in place 
for today's analysis. Pick-up trucks have much longer redesign 
schedules than small cars. Some manufacturers redesign vehicles often, 
while other manufacturers redesign vehicles less often. Even if two 
manufacturers have similar redesign cadence, the model years in which 
the redesigns occur may still be different and dependent on where each 
of the manufacturer's products are in their life cycle.
    Table VI-11 summarizes the average age of manufacturers' offering 
by vehicle technology class. A value of ``0.0'' means that every 
vehicle for a manufacturer in the vehicle technology class, represented 
by the MY 2017 analysis fleet was new in MY 2017. Across the industry 
manufacturers redesigned MY 2017 vehicles an average of 3.5 years 
earlier, meaning the average MY 2017 vehicle was last redesigned in 
approximately MY 2013, also on average near a midpoint in their product 
lifecycle.

[[Page 24300]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.095

BILLING CODE 4910-59-C
    Some commenters cited examples of vehicles in the NPRM analysis 
fleet where the redesign years were off by a year here or there in the 
2017-2022 timeframe relative to the most recent public announcements, 
or that the extended forecasts were too rigid.\386\ The CAFE model 
structurally requires an input for the redesign years, and the agencies 
worked to make these generally representative without disclosing 
precise CBI product plans. Many of the redesign schedules were carried 
over from the NPRM, with a few minor updates.
---------------------------------------------------------------------------

    \386\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
---------------------------------------------------------------------------

    Some commenters contended that the agencies should not look at the 
historical data to project the timing between redesigns (``business as 
usual''), but should instead adopt a ``policy case'' with an 
accelerated pace of redesigns and refreshes.\387\ Some commenters 
suggested that the agencies use a standard 5 or 6 year redesign 
schedule for all manufacturers and all products as a way to lower 
projected costs.\388\ Other stakeholders commented that the entire 
industry should be modeled with the ability to redesign everything at 
one time in the near term because that would not presuppose precisely 
how manufacturers may adjust their fleet.\389\
---------------------------------------------------------------------------

    \387\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
    \388\ NHTSA-2018-0067-11985, HD Systems.
    \389\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
---------------------------------------------------------------------------

    If the agencies were to implement any such approaches, the agencies 
would need to more precisely account for tooling costs, research and 
development costs, and product lifecycle marketing costs, or risk 
missing ``hidden costs'' of a shortened cadence. To account properly 
for these, the CAFE model would require major changes, and would 
require specific inputs that are currently covered generically under 
the retail price equivalency (RPE) factor.\390\ The agencies considered 
these comments, and decided the process for refresh and redesign 
outlined in the NPRM was a reasonable and realistic approach to 
characterize product changes. The agencies conducted sensitivity 
analysis with compressed redesign and refresh schedules, though these 
ignore the resulting compressed amortization schedules, missing 
important costs that are incorporated in the current RPE assumptions.
---------------------------------------------------------------------------

    \390\ Shorter redesign schedules are likely to put upward 
pressure on RPE, as the manufacturers would have less time to recoup 
investments.
---------------------------------------------------------------------------

    Some commenters claimed that the agency had extraordinarily 
extended redesign schedule of 17.7 years for FCA between 2021-2025, and 
an average redesign time of 25.8 years for Ford between 2022-2025.\391\ 
The agencies found these claims inaccurate and without basis. Table VI-
10, ``Summary of Sales Weighted Average Time

[[Page 24301]]

between Engineering Redesigns, by Manufacturer, by Vehicle Technology 
Class'' summarizes the data used in today's analysis (which is very 
similar to the information used in the NPRM, with some minor 
adjustments and updates to the fleet), and the detailed information 
vehicle-by-vehicle is reported in the ``market data'' file. The 
agencies recognize that the natural sequence of redesigns for some 
manufacturers and some products is not ideal to meet stringent 
alternatives, which is part of the consideration for economic 
practicability and technological feasibility. Manufacturers commented 
supportively on the idea of vehicle specific redesign schedules, and 
the redesign cadence used in the NPRM, as these contribute to realistic 
assessments of new technology penetration within the fleet, and 
acknowledge the heterogeneity in the product development approaches and 
business practices for each manufacturer.\392\ One commenter recognized 
that redesign and refresh schedules represented a vast improvement over 
phase-in caps to model the adoption of mature technologies.\393\
---------------------------------------------------------------------------

    \391\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
    \392\ NHTSA-2018-0067-11928, Ford Motor Company.
    \393\ NHTSA-2018-0067-0444, Walter Kreucher.
---------------------------------------------------------------------------

    Other commenters argued that the structural construct of 
technologies only being available at redesign or at refresh (via 
inheritance) did not reflect real world actions and was not supported 
by any actual data.\394\ Other commenters acknowledged the inheritance 
of engine and transmission technologies at refresh as an important, 
positive feature of the CAFE model.\395\ HD Systems argued that an 
engine or transmission package available in other markets on a global 
platform could be imported to the U.S. market during refresh, and did 
not require a ``leader'' at redesign in the U.S. market to seed 
adoption. HDS cited a few examples where manufacturers have introduced 
strong hybrid powertrains on an existing vehicle a year or two after 
the product launch, not associated with any particular vehicle redesign 
or refresh.
---------------------------------------------------------------------------

    \394\ NHTSA-2018-0067-11985, HD Systems.
    \395\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
---------------------------------------------------------------------------

    The agencies carefully considered these comments, and observed that 
some relatively low volume hybrid options may appear after launch, or 
that some transmissions were quickly replaced shortly after a major 
redesign. In many of these cases, launch delays, warranty claims, or 
other external factors contributed to, at least in part, an atypically 
timed introduction of fuel saving technology to the fleet.\396\ At this 
point, this does not appear to be a mainstream, or preferred industry 
practice. However, the agencies will continue to evaluate this. For 
future rulemaking, the agencies may consider engine refresh and 
redesign cycles for engines and transmissions. These may be separate 
from vehicle redesign and refresh schedules because the powertrain 
product lifecycles may be longer on average than the typical vehicle 
redesign schedules. This approach, if researched and implemented in 
future analysis, could provide some opportunity for manufacturers to 
introduce new powertrain technologies independent of the vehicle 
redesign schedules, in addition to inheriting advanced powertrain 
technology as refresh as already modeled in the NPRM and today's 
analysis.
---------------------------------------------------------------------------

    \396\ Such instances are observable in detailed CAFE and 
CO2 compliance data submitted to EPA and NHTSA.
---------------------------------------------------------------------------

    For today's analysis, the agencies, with a few exceptions based on 
updated publicly available information, carried over redesign cadences 
for each vehicle nameplate as presented in the NPRM. The agencies do 
not claim that the projected redesign years will perfectly match what 
industry does--notably because refresh and redesign information is CBI 
and the agencies have applied more generalized schedules to protect the 
CBI. Also, what any individual manufacturer may choose to do today 
could be completely different than what it chooses to do tomorrow due 
to changing business circumstances and plans--but the agencies have 
worked to ensure the timing of redesigns will be roughly correct 
(especially in the near term), and that the time between redesigns will 
continue forward for each manufacturer as it has based on recent 
history. The agencies have also increased the frequency of refreshes in 
response to comments about the proliferation of some engine and 
transmission families through manufacturers' product portfolios.
    Also for today's analysis, the agencies now explicitly model CAFE 
compliance pathways out through 2050. For the model to work as 
intended, the agencies must project refresh and redesign schedules out 
through 2050. The agencies recognize that the accuracy of predictions 
about the distant future, particularly about refresh and redesign 
cycles through the 2030-2050 timeframe, are likely to be poor. If 
historical evolution of the industry continues, many of the nameplates 
carried forward in the fleet are likely to be out of production, and 
new nameplates not considered in the analysis are sure to emerge. 
Still, carrying forward the MY 2017 fleet with the current refresh and 
redesign cadences is consistent with the current analysis, and imposing 
an alternative schedule on the fleet, or making up new nameplates and 
retiring older nameplates without a clear basis, would lack proper 
foundation.
BILLING CODE 4910-59-P

[[Page 24302]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.096

BILLING CODE 4910-59-C
(6) Defining Technology Adoption Features
    In some circumstances, the agencies may reference full vehicle 
simulation effectiveness data for technology combinations that are not 
able to be, or are not likely to be applied to all vehicles. In some 
cases, a specific technology as modeled only exists on paper, and 
questions remain about the technological feasibility of the efficiency 
characterization.\397\ Or, a technology may perform admirably on the 
test cycle, but fail to meet all functional, or performance 
requirements for certain vehicles.\398\ In other cases, the 
intellectual property landscape may make commercialization of one 
technology risky for a manufacturer without the consent of the 
intellectual property owner.\399\ In such cases, the agencies may not 
allow a technology to be applied to a certain vehicle. The agencies 
designate this in the ``market data'' file with a ``SKIP'' for the 
technology and vehicle. The logic is explained technology by technology 
in this document, as the logic was explained in the PRIA for this rule.
---------------------------------------------------------------------------

    \397\ High levels of aerodynamic drag reduction for some body 
styles, or EPA's previous, speculative characterization of ``HCR2'' 
engines, for example.
    \398\ Examples of applications that are unsuitable for certain 
technologies include low end torque requirements for HCR engines on 
high load vehicles, or towing and trailering applications, 
continuously variable transmissions in high torque applications, and 
low rolling resistance tires on vehicles built for precision 
cornering and high lateral forces, or instant acceleration from a 
stand still.
    \399\ Variable compression ratio engines, for example.
---------------------------------------------------------------------------

    Some commenters argued that the restrictions of technologies on a 
case-by-case basis required case-by-case explanation (and not objective 
specification defined cut-offs), and that the use of CBI for 
performance considerations was unacceptable unless fully 
disclosed.\400\ As discussed above, the agencies are not able to 
disclose CBI. Stakeholders have had plenty of opportunities to comment 
on the applicability of technologies, including the few that have used 
SKIP logic restrictions for a portion of the fleet.
---------------------------------------------------------------------------

    \400\ NHTSA-2018-0067-11741, ICCT.
---------------------------------------------------------------------------

    Other commenters suggested an optimistic and wholly unfounded 
approach to manufacturer innovation, arguing that costs would continue 
to come down (beyond what is currently modeled with cost learning), and 
the list of fuel-saving technologies would continually regenerate 
itself (even if the technological mechanism for fuel saving 
technologies was not yet identified).\401\ Therefore, the argument goes 
that people will figure out new ways to improve fuel saving 
technologies to increase their applicability, and the current 
technology characterization should be enabled for selection with no 
restriction--not because the commenter knows how the technology will be 
adapted, but that the commenter believes the technology could, 
eventually, within the timeline of the rulemaking, be adapted, brought 
to market, and be accepted by consumers. While the agencies recognize 
the improvements that many manufacturers

[[Page 24303]]

have achieved in fuel saving technologies, some of which were difficult 
to foresee, the agencies have an obligation under the law to be 
judicious and specific about technological feasibility, and to avoid 
speculative conclusions about technologies to justify the rulemaking.
---------------------------------------------------------------------------

    \401\ NHTSA-208-0067-12122-33, American Council for an Energy-
Efficient Economy.
---------------------------------------------------------------------------

c) Other Analysis Fleet Data
(1) Safety Classes
    The agencies referenced the mass-size-safety analysis to project 
the effects changes in weight may have on crash fatalities. That 
analysis, discussed in more detail in Section VI.D.2, considers how 
weight changes may affect safety for cars, crossover utility vehicles 
and sport utility vehicles, and pick-up trucks. To consider these 
effects, the agencies mapped each vehicle in the analysis fleet to the 
appropriate ``Safety Class.''
(2) Labor Utilization
    The analysis fleet summarizes components of direct labor for each 
vehicle considered in the analysis. The labor is split into three 
components: (1) Dealership hours worked on sales functions per vehicle, 
(2) direct assembly labor for final assembly, engine, and transmission, 
and (3) percent U.S. content.
    In the MY 2016 fleet for the NPRM, the agencies catalogued 
production locations and plant employment, reviewed annual reports from 
the North American Dealership Association to estimate dealership 
employment (27.8 hours per vehicle sold), and estimated the industry 
average labor hours for final assembly of vehicles (30 hours per 
vehicle produced), engine machining and assembly (4 hours per engine 
produced), and transmission production (5 hours per transmission 
produced).
    Today's analysis fleet carries over the estimated labor 
coefficients for sales and production, but references the most recent 
Part 583 American Automobile Labeling Act Report for percent U.S. 
content and for the location of vehicle assembly, engine assembly, and 
transmission assembly.\402\
---------------------------------------------------------------------------

    \402\ Part 583 American Automobile Labeling Act Report, 
available at https://www.nhtsa.gov/part-583-american-automobile-labeling-act-reports.
---------------------------------------------------------------------------

(3) Production Volumes for Sales Analysis
    A final important aspect of projecting what vehicles will exist in 
future model years and potential manufacturer responses to standards is 
estimating how future sales might change in response to different 
potential standards. If potential future standards appear likely to 
have major effects in terms of shifting production from cars to trucks 
(or vice versa), or in terms of shifting sales between manufacturers or 
groups of manufacturers, that is important for the agencies to 
consider. For previous analyses, the CAFE model used a static forecast 
contained in the analysis fleet input file, which specified changes in 
production volumes over time for each vehicle model/configuration. This 
approach yielded results that, in terms of production volumes, did not 
change between scenarios or with changes in important model inputs. For 
example, very stringent standards with very high technology costs would 
result in the same estimated production volumes as less stringent 
standards with very low technology costs. For this analysis, as in the 
proposal, the CAFE model begins with the first-year production volumes 
(i.e., MY 2017 for today's analysis) and adjusts ensuing sales mix year 
by year (between cars and trucks, and between manufacturers) 
endogenously as part of the analysis, rather than using external 
forecasts of future car/truck split and future manufacturer sales 
volumes. This leads the model to produce different estimates of future 
production volumes under different standards and in response to 
different inputs, reflecting the expectation that regulatory standards 
and other external factors will, in fact, impact the market.
(4) Comments on Other Analysis Fleet Data
    Some commenters suggest that the CAFE model should run as a full 
consumer choice model (and this idea is discussed in more detail in 
Section VI.D.1). While this sounds like a reasonable request on the 
surface, such an approach would place enormous new demands on the data 
characterized in the fleet (and preceding fleets, which may be needed 
to calibrate a model properly). For instance, some model concepts may 
depend on a bevy of product features, such as interior cargo room, 
artistic appeal of the design, and perceived quality of the vehicle. 
But product features alone may not be sufficient. Additional 
information about dealership channels, product awareness and 
advertising effectiveness, and financing terms also may be required. 
Such information could dramatically increase the scope of work needed 
to characterize the analysis fleet for future rulemakings. As described 
in Section VI.D.1.b)(2)(d) Using Vehicle Choice Models in Rulemaking 
Analysis. Accordingly, the agencies decided not to develop such a model 
for this rulemaking.
2. Treatment of Compliance Credit Provisions
    Today's final rule involves a variety of provisions regarding 
``credits'' and other compliance flexibilities. Some recently 
introduced regulatory provisions allow a manufacturer to earn 
``credits'' that will be counted toward a vehicle's rated 
CO2 emissions level, or toward a fleet's rated average 
CO2 or CAFE level, without reference to required levels for 
these average levels of performance. Such flexibilities effectively 
modify emissions and fuel economy test procedures, or methods for 
calculating fleets' CAFE and average CO2 levels. Such 
provisions are discussed below in Section VI.B.2. Other provisions (for 
CAFE, statutory provisions) allow manufacturers to earn credits by 
achieving CAFE or average CO2 levels beyond required levels; 
these provisions may hence more appropriately be termed ``compliance 
credits.''
    EPCA has long provided that, by exceeding the CAFE standard 
applicable to a given fleet in a given model year, a manufacturer may 
earn corresponding ``credits'' that the same manufacturer may, within 
the same regulatory class, apply toward compliance in a different model 
year. EISA amended these provisions by providing that manufacturers 
may, subject to specific statutory limitations, transfer compliance 
credits between regulatory classes, and trade compliance credits with 
other manufacturers. The CAA provides EPA with broad standard-setting 
authority for the CO2 program, with no specific directives 
regarding either CO2 standards or CO2 compliance 
credits.
    EPCA also specifies that NHTSA may not consider the availability of 
CAFE credits (for transfer, trade, or direct application) toward 
compliance with new standards when establishing the standards 
themselves.\403\ Therefore, this analysis, like that presented in the 
NPRM, considers 2020 to be the last model year in which carried-forward 
or transferred credits can be applied for the CAFE program. Beginning 
in model year 2021, today's ``standard setting'' analysis for NHTSA's 
program is conducted assuming each fleet must comply with the CAFE 
standard separately in every model year.
---------------------------------------------------------------------------

    \403\ 49 U.S.C. 32902(h)(3).
---------------------------------------------------------------------------

    The ``unconstrained'' perspective acknowledges that these 
flexibilities exist as part of the program, and, while not considered 
by NHTSA in setting standards, are nevertheless important to consider 
when attempting to estimate the real impact of any alternative. Under

[[Page 24304]]

the ``unconstrained'' perspective, credits may be earned, transferred, 
and applied to deficits in the CAFE program throughout the full range 
of model years in the analysis. The Final Environmental Impact Analysis 
(FEIS) accompanying today's final rule, like the corresponding Draft 
EIS analysis, presents results of ``unconstrained'' modeling. Also, 
because the CAA provides no direction regarding consideration of any 
CO2 credit provisions, today's analysis, like the NPRM 
analysis, includes simulation of carried-forward and transferred 
CO2 credits in all model years.
    Some commenters took issue broadly with this treatment of 
compliance credits. Michalek and Whitefoot wrote that ``we find this 
requirement problematic because the automakers use these flexibilities 
as a common means of complying with the regulation, and ignoring them 
will bias the cost-benefit analysis to overestimate costs.'' \404\
---------------------------------------------------------------------------

    \404\ Michalek, J. and Whitefoot, K., NHTSA-2018-0067-11903, at 
10-11.
---------------------------------------------------------------------------

    Counter to the above general claim, the CAFE model does provide 
means to simulate manufacturers' potential application of some 
compliance credits, and both the analysis of CO2 standards 
and the NEPA analysis of CAFE standards do make use of this aspect of 
the model. As discussed above, NHTSA does not have the discretion to 
consider the credit program--in fact, the agency is prohibited by 
statute from doing so--in establishing maximum feasible standards. 
Further, as discussed below, the agencies also continue to find it 
appropriate for the analysis largely to refrain from simulating two of 
the mechanisms allowing the use of compliance credits.
    The model's approach to simulating compliance decisions accounts 
for the potential to earn and use CAFE credits as provided by EPCA/
EISA. The model similarly accumulates and applies CO2 
credits when simulating compliance with EPA's standards. Like past 
versions, the current CAFE model can be used to simulate credit carry-
forward (a.k.a. banking) between model years and transfers between the 
passenger car and light truck fleets but not credit carry-back (a.k.a. 
borrowing) from future model years or trading between manufacturers.
    Regarding the potential to carry back compliance credits, UCS 
commented that, although past versions of the CAFE model had 
``considered this flexibility in its approach to multiyear modeling,'' 
NHTSA had, without explanation, ``abruptly discontinued support of this 
method of compliance,'' such that ``manufacturers are generally 
incentivized to over comply, regardless of whether carrying forward a 
deficit to be compensated by later overcompliance would be a more cost-
effective method of compliance.'' \405\ Citing the potential that 
manufacturers could make use of carried back credits in the future, UCS 
also stated that ``NHTSA's decision to constrain it in the model is 
unreasonable and arbitrary.'' \406\ UCS effectively implies that the 
agencies should base standards on analysis that presumes manufacturers 
will take full theoretical advantage of provisions allowing credits to 
be borrowed.
---------------------------------------------------------------------------

    \405\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 44.
    \406\ UCS, op. cit., at 77.
---------------------------------------------------------------------------

    The agencies have carefully considered these comments, and while 
EPA's decisions regarding CO2 standards can consider the 
potential to carry back compliance credits from later to earlier model 
years, and NHTSA's ``unconstrained'' evaluation could also do so, past 
examples of failed attempts to carry back CAFE credits (e.g., a MY2014 
carry back default leading to a civil penalty payment) underscore the 
riskiness of such ``borrowing.'' Recent evidence indicates 
manufacturers are disinclined to take such risks,\407\ and both 
agencies find it reasonable and prudent to refrain from attempting to 
simulate such ``borrowing'' in rulemaking analysis.
---------------------------------------------------------------------------

    \407\ Section IX, below, reviews data regarding manufacturers' 
use of CAFE compliance credit mechanism during MYs 2011-2016, and 
shows that the use of ``carry back'' credits is, relative to the use 
of other compliance credit mechanisms, too small to discern.
---------------------------------------------------------------------------

    Unlike past versions, the NPRM and current versions of CAFE model 
provide a basis to specify (in model inputs) CAFE credits available 
from model years earlier than those being explicitly simulated. For 
example, with this analysis representing model years 2017-2050 
explicitly, credits earned in model year 2012 are made available for 
use through model year 2017 (given the current five-year limit on 
carry-forward of credits). The banked credits are specific to both the 
model year and fleet in which they were earned.
    In addition to the above-mentioned comments, UCS also cited as 
``errors'' that ``the model does not accurately reflect the one-time 
exemption from the EPA 5-year credit life for credits earned in the MY 
2010-2015 timeframe'' and ``NHTSA assumes that there will be absolutely 
no credit trading between manufacturers.''
    As discussed below, in the course of updating the analysis fleet 
from MY 2016 to MY 2017, the agencies have updated and expanded the 
manner in which the model accounts for credits earned prior to MY 2017, 
including credits earned as early as MY 2009. In order to increase the 
realism with which the model transitions between the early model year 
(MYs 2017-2020) and the later years that are the subject of this 
action, the agencies have accounted for the potential that some 
manufacturers might trade some of these pre-MY 2017 credits to other 
manufacturers. However, as with the NPRM, the analysis refrains from 
simulating the potential that manufacturers might continue to trade 
credits during and beyond the model years covered by today's action. 
The agencies remain concerned that any realistic simulation of such 
trading would require assumptions regarding which specific pairs of 
manufacturers might actually trade compliance credits, and the evidence 
to date makes it clear that the credit market is far from fully 
``open.'' With respect to the FCA comment cited above, the agencies 
also remain concerned that to set standards based on an analysis that 
presumes the use of program flexibilities risks making the 
corresponding actions mandatory. Some flexibilities--credit carry-
forward (banking) and transfers between fleets in particular--involve 
little risk, because they are internal to a manufacturer and known in 
advance. As discussed above, credit carry-back involves significant 
risk, because it amounts to borrowing against future improvements, 
standards, and production volume and mix--and anticipated market demand 
for fuel efficient vehicles often fail to materialize. Similarly, 
credit trading also involves significant risk, because the ability of 
manufacturer A to acquire credits from manufacturer B depends not just 
on manufacturer B actually earning the expected amount of credit, but 
also on manufacturer B being willing to trade with manufacturer A, and 
on potential interest by other manufacturers. Manufacturers' compliance 
plans have already evidenced cases of compliance credit trades that 
were planned and subsequently aborted, reinforcing the agencies' 
judgment that, like credit banking, credit trading involves too much 
risk to be included in an analysis that informs decisions about the 
stringency of future standards. Nevertheless, recognizing that some 
manufacturers have actually been trading credits, the agencies have, as 
in the NPRM, included in the sensitivity analysis a case that simulates 
``perfect'' trading of compliance credits, focusing

[[Page 24305]]

on CO2 standards to illustrate the hypothetical maximum 
potential impact of trading. The FRIA summarizes results of this and 
other cases included in the sensitivity analysis.
    As discussed in the CAFE model documentation, the model's default 
logic attempts to maximize credit carry-forward--that is, to ``hold 
on'' to credits for as long as possible. If a manufacturer needs to 
cover a shortfall that occurs when insufficient opportunities exist to 
add technology in order to achieve compliance with a standard, the 
model will apply credits. Otherwise the manufacturer carries forward 
credits until they are about to expire, at which point it will use them 
before adding technology that is not considered cost-effective. The 
model attempts to use credits that will expire within the next three 
years as a means to smooth out technology application over time to 
avoid both compliance shortfalls and high levels of over-compliance 
that can result in a surplus of credits. Although it remains impossible 
precisely to predict manufacturer's actual earning and use of 
compliance credits, and this aspect of the model may benefit from 
future refinement as manufacturers and regulators continue to gain 
experience with these provisions, this approach is generally consistent 
with manufacturers' observed practices.
    NHTSA introduced the CAFE Public Information Center to provide 
public access to a range of information regarding the CAFE 
program,\408\ including manufacturers' credit balances. However, there 
is a data lag in the information presented on the CAFE PIC that may not 
capture credit actions across the industry for as much as several 
months. Furthermore, CAFE credits that are traded between manufacturers 
are adjusted to preserve the gallons saved that each credit 
represents.\409\ The adjustment occurs at the time of application 
rather than at the time the credits are traded. This means that a 
manufacturer who has acquired credits through trade, but has not yet 
applied them, may show a credit balance that is either considerably 
higher or lower than the real value of the credits when they are 
applied. For example, a manufacturer that buys 40 million credits from 
Tesla may show a credit balance in excess of 40 million. However, when 
those credits are applied, they may be worth only 1/10 as much--making 
that manufacturer's true credit balance closer to 4 million than 40 
million.
---------------------------------------------------------------------------

    \408\ CAFE Public Information Center, http://www.nhtsa.gov/CAFE_PIC/CAFE_PIC_Home.htm (last visited June 22, 2018).
    \409\ CO2 credits for EPA's program are denominated 
in metric tons of CO2 rather than gram/mile compliance 
credits and require no adjustment when traded between manufacturers 
or fleets.
---------------------------------------------------------------------------

    For the NPRM, the agencies reviewed then-recent credit balances, 
estimated the potential that some manufacturers could trade credits, 
and developed inputs that make carried-forward credits available in 
each of model years 2011-2015, after subtracting credits assumed to be 
traded to other manufacturers, adding credits assumed to be acquired 
from other manufacturers through such trades, and adjusting any traded 
credits (up or down) to reflect their true value for the fleet and 
model year into which they were traded.\410\ For today's analysis, an 
additional model year's data was available in mid-2019, and the 
agencies updated these inputs, as summarized in Table VI-12, Table VI-
13, and Table VI-14. While the CAFE model will transfer expiring 
credits into another fleet (e.g., moving expiring credits from the 
domestic car credit bank into the light truck fleet), some of these 
credits were moved into the initial banks to improve the efficiency of 
application and both to reflect better the projected shortfalls of each 
manufacturer's regulated fleets and to represent observed behavior. For 
context, a manufacturer that produces one million vehicles in a given 
fleet, and experiences a shortfall of 2 mpg, would need 20 million 
credits, adjusted for fuel savings, to offset the shortfall completely.
---------------------------------------------------------------------------

    \410\ The adjustments, which are based upon the CAFE standard 
and model year of both the party originally earning the credits and 
the party applying them, were implemented assuming the credits would 
be applied to the model year in which they were set to expire. For 
example, credits traded into a domestic passenger car fleet for MY 
2014 were adjusted assuming they would be applied in the domestic 
passenger car fleet for MY 2019.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24306]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.097

[GRAPHIC] [TIFF OMITTED] TR30AP20.098


[[Page 24307]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.099

BILLING CODE 4910-59-C
    In addition to the inclusion of these existing credit banks, the 
CAFE model also updated its treatment of credits in the rulemaking 
analysis. EPCA requires that NHTSA set CAFE standards at maximum 
feasible levels for each model year without consideration of the 
program's credit mechanisms. However, as recent NHTSA CAFE/EPA tailpipe 
CO2 emissions rulemakings have evaluated effects of 
standards over longer time periods, the early actions taken by 
manufacturers required more nuanced representation. Accordingly, the 
CAFE model now provides for a setting to establish a ``last year to 
consider credits.'' This adjustment is set at the last year for which 
new standards are not being considered (MY 2020 in this analysis). This 
allows the model to replicate the practical application of existing 
credits toward compliance in early years but also to examine the impact 
of proposed standards based solely on fuel economy improvements in all 
years for which new standards are being considered.
    Regarding the model's simulation of manufacturers' potential 
earning and application of compliance credits, UCS commented that the 
model ``inexplicably lets credits expire'' because ``all technologies 
which pay for themselves within the assumed payback period are applied 
to all manufacturers, regardless of credit status.'' UCS also claimed 
that ``NHTSA did not accurately reflect unique attributes of EPA's 
credit bank,'' that ``credits are not traded between manufacturers,'' 
and that ``NHTSA does not model credit carryback for compliance.'' 
\411\ Relatedly, as discussed above, UCS attributes modeling outcomes 
to the ``effective cost'' metric used to select from among available 
fuel-saving technologies.\412\ As discussed in Section VI.B.1, the 
agencies expect that manufacturers are likely to improve fuel economy 
voluntarily insofar as doing so ``pays back'' economically within a 
short period (30 months), and the agencies note that periods of 
regulatory stability have, in fact, been marked by CAFE levels 
exceeding requirements. As discussed above, the agencies have excluded 
simulation of credit trading (except in MYs prior to those under 
consideration, aside from an idealized case presented in the 
sensitivity analysis) and likewise excluded simulation of potential 
``carryback'' provisions. The agencies have excluded modeling these 
scenarios not just because of the analytical complexities involved (and 
rejecting, for example, the random number generator analysis suggested 
by UCS), but also because the agencies agree that the actual provisions 
regarding trading and borrowing of compliance credits create too much 
risk to be used in the analysis underlying consideration of standards. 
However, as discussed above, the agencies have revised the ``metric'' 
used to prioritize available options to apply fuel-saving technologies. 
As discussed below, the agencies have revised model inputs to include 
the large quantity of ``legacy'' compliance credits EPA has made 
available under its CO2 standards.
---------------------------------------------------------------------------

    \411\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 35-46.
    \412\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 28-30.
---------------------------------------------------------------------------

    The CAFE model has also been modified to include a similar 
representation of existing credit banks in EPA's CO2 
program. While the life of a CO2 credit, denominated in 
metric tons of CO2, has a five-year life, matching the 
lifespan of CAFE credits, such credits earned in the early MY 2009-2011 
years of the EPA program, may be used through MY 2021.\413\ The CAFE 
model was not modified to allow

[[Page 24308]]

exceptions to the life-span of compliance credits, and, to reflect 
statutory requirements, treated them as if they may be carried forward 
for no more than five years, so the initial credit banks were modified 
to anticipate the years in which those credits might be needed. MY 2016 
was simulated explicitly in the NPRM analysis to prohibit the inclusion 
of banked credits in MY 2016 (which could be carried forward from MY 
2016 to MY 2021), and thus underestimated the extent to which 
individual manufacturers, and the industry as a whole, could rely on 
these early credits to comply with EPA standards between MY 2016 and MY 
2021. However, as indicated in the NPRM, the final rule's model inputs 
updated the analysis fleet's basis to MY 2017, such that these 
additional banked credits can be included. The credit banks with which 
the simulations in this analysis were conducted are presented in the 
following Tables:
---------------------------------------------------------------------------

    \413\ In the 2010 rule, EPA placed limits on credits earned in 
MY 2009, which expired prior to this rule. However, credits 
generated in MYs 2010-2011 may be carried forward, or traded, and 
applied to deficits generated through MY 2021.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.100


[[Page 24309]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.101

BILLING CODE 4910-59-C
    While the CAFE model does not simulate the ability to trade credits 
between manufacturers, it does simulate the strategic accumulation and 
application of compliance credits, as well as the ability to transfer 
credits between fleets to improve the compliance position of a less 
efficient fleet by leveraging credits earned by a more efficient fleet. 
The model prefers to hold on to earned compliance credits within a 
given fleet, carrying them forward into the future to offset potential 
future deficits. This assumption is consistent with observed strategic 
manufacturer behavior dating back to 2009.
    From 2009 to present, no manufacturer has transferred CAFE credits 
into a fleet to offset a deficit in the same year in which they were 
earned. This has occurred with credits acquired from other 
manufacturers via trade but not with a manufacturer's own credits. 
Therefore, the current representation of credit transfers between 
fleets--where the model prefers to transfer expiring, or soon-to-be-
expiring credits rather than newly earned credits--is both appropriate 
and consistent with observed industry behavior.
    This may not be the case for CO2 standards, though it is 
difficult to be certain at this point. The CO2 program 
seeded the industry with a large quantity of early compliance credits 
(earned in MYs 2009-2011) \414\ prior to the existence formal 
CO2 standards. Early credits from MYs 2010 and 2011, 
however, do not expire until 2021. Thus, for manufacturers looking to 
offset deficits, it is more sensible to exhaust credits that were 
generated during later model years (which are set to expire within the 
next five years), rather than relying on the initial bank of credits 
from MYs 2010 and 2011. The first model year for which earned credits 
outlive the initial bank is MY 2017, for which final manufacturer 
CO2 performance data (and hence, banked credits) has not yet 
been released. However, considering that under the CO2 
program manufacturers simultaneously comply with passenger car and 
light truck fleets, to more accurately represent the CO2 
credit system the CAFE model allows (and encourages) intra-year 
transfers between regulated fleets for the purpose of simulating 
compliance with the CO2 standards.
---------------------------------------------------------------------------

    \414\ In response to public comment, EPA eliminated the possible 
use of credits earned in MY 2009 for future model years. However, 
credits earned in MY 2010 and MY 2011 remain available for use.
---------------------------------------------------------------------------

a) Off-Cycle and A/C Efficiency Adjustments to CAFE and Average 
CO2 Levels
    In addition to more rigorous accounting of CAFE and CO2 
credits, the model now also accounts for air conditioning efficiency 
and off-cycle adjustments. NHTSA's program considers those adjustments 
in a manufacturer's compliance calculation starting in MY 2017, and the 
NPRM version of the model used the adjustments claimed by each 
manufacturer in MY 2016 as the starting point for all future years. 
Because air conditioning efficiency and off-cycle adjustments are not 
credits in NHTSA's program, but rather adjustments to compliance fuel 
economy (much like the Flexible Fuel Vehicle adjustments due to phase 
out in MY 2019), they may be included under either a ``standard 
setting'' or ``unconstrained'' analysis perspective.
    The manner in which the CAFE model treats the EPA and CAFE A/C 
efficiency and off-cycle credit programs is similar, but the model also 
accounts for A/C leakage (which is not part of NHTSA's program). When 
determining the compliance status of a

[[Page 24310]]

manufacturer's fleet (in the case of EPA's program, PC and LT are the 
only fleet distinctions), the CAFE model weighs future compliance 
actions against the presence of existing (and expiring) CO2 
credits resulting from over-compliance with earlier years' standards, 
A/C efficiency credits, A/C leakage credits, and off-cycle credits.
    Another aspect of credit accounting, implemented in the NPRM 
version of the CAFE model, involved credits related to the application 
of off-cycle and A/C efficiency adjustments, which manufacturers earn 
by taking actions such as special window glazing or using reflective 
paints that provide fuel economy improvements in real-world operation 
but do not produce measurable improvements in fuel consumption on the 
2-cycle test.
    NHTSA's inclusion of off-cycle and A/C efficiency adjustments began 
in MY 2017, while EPA has collected several years' worth of submissions 
from manufacturers about off-cycle and A/C efficiency technology 
deployment. Currently, the level of deployment can vary considerably by 
manufacturer, with several claiming extensive Fuel Consumption 
Improvement Values (FCIV) for off-cycle and A/C efficiency 
technologies, and others almost none. The analysis of alternatives 
presented here (and in the NPRM) does not attempt to project how future 
off-cycle and A/C efficiency technology use will evolve or speculate 
about the potential proliferation of FCIV proposals submitted to the 
agencies. Rather, this analysis uses the off-cycle credits submitted by 
each manufacturer for MY 2017 compliance, and, with a few exceptions, 
carries these forward to future years. Several of the technologies 
described below are associated with A/C efficiency and off-cycle FCIVs. 
In particular, stop-start systems, integrated starter generators, and 
full hybrids are assumed to generate off-cycle adjustments when applied 
to vehicles to improve their fuel economy. Similarly, higher levels of 
aerodynamic improvements are assumed to include active grille shutters 
on the vehicle, which also qualify for off-cycle FCIVs.
    The NPRM analysis assumed that any off-cycle FCIVs that are 
associated with actions outside of the technologies discussed in 
Section VI.C (either chosen from the pre-approved ``pick list,'' or 
granted in response to individual manufacturer petitions) remained at 
the levels claimed by manufacturers in MY 2017. Any additional A/C 
efficiency and off-cycle adjustments that accrued as the result of 
explicit technology application calculated dynamically in each model 
year for each alternative. The NPRM version of the CAFE model also 
represented manufacturers' credits for off-cycle improvements, A/C 
efficiency improvements, and A/C leakage reduction in terms of values 
applicable across all model years.
    Recognizing that application of these improvements thus far varies 
considerably among manufacturers, such that some manufacturers have 
opportunities to earn significantly more of the corresponding 
adjustments over time, the agencies have expanded the CAFE model's 
representation of these credits to provide for year-by-year 
specification of the amounts of each type of adjustment for each 
manufacturer, denominated in grams CO2 per mile,\415\ as 
summarized in the following table:
---------------------------------------------------------------------------

    \415\ For estimating their contribution to CAFE compliance, the 
grams CO2/mile values in Table VI-1711 are converted to 
gallons/mile and applied to a manufacturer's 2-cycle CAFE 
performance. When calculating compliance with EPA's CO2 
program, there is no conversion necessary (as standards are also 
denominated in grams/mile).
    \416\ These values are specified in the ``market_ref.xlsx'' 
input file's ``Credits and Adjustments'' worksheet. The file is 
available with the archive of model inputs and outputs posted at 
https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24311]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.102


[[Page 24312]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.103


[[Page 24313]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.104

BILLING CODE 4910-59-C
    In addition to these refinements to the estimation of the 
quantities of adjustments earned over time by each manufacturer, the 
agencies revised the

[[Page 24314]]

CAFE model to apply estimates of the corresponding costs. For today's 
analysis, the agencies applied estimates developed previously by EPA, 
adjusting these values to 2019 dollars. The following table summarizes 
inputs through model year 2030:
[GRAPHIC] [TIFF OMITTED] TR30AP20.105

    The model currently accounts for any off-cycle adjustments 
associated with technologies that are included in the set of fuel-
saving technologies explicitly simulated as part of this proposal (for 
example, start-stop systems that reduce fuel consumption during idle or 
active grille shutters that improve aerodynamic drag at highway speeds) 
and accumulates these adjustments up to the 10 g/mi cap. As a practical 
matter, most of the adjustments for which manufacturers are claiming 
off-cycle FCIV exist outside of the technology tree, so the cap is 
rarely reached during compliance simulation. The agencies have 
considered the potential to model their application explicitly. 
However, doing so would require data regarding which vehicle models 
already possess these improvements as well as the cost and expected 
value of applying them to other models in the future. Such data is 
currently too limited to support explicit modeling of these 
technologies and adjustments.
b) Alternative Fuel Vehicles
    When establishing maximum feasible fuel economy standards, NHTSA is 
prohibited from considering the availability of alternatively fueled 
vehicles,\417\ and credit provisions related to AFVs that significantly 
increase their fuel economy for CAFE compliance purposes. Under the 
``standard setting'' perspective, these technologies (pure battery 
electric vehicles and fuel cell vehicles) \418\ are not available in 
the compliance simulation to improve fuel economy. Under the 
``unconstrained'' perspective, such as is documented in the DEIS and 
FEIS, the CAFE model considers these technologies in the same manner as 
other available technologies, and may apply them if they represent 
cost-effective compliance pathways. However, under both perspectives, 
the analysis continues to include dedicated AFVs that already exist in 
the MY 2017 fleet (and their projected future volumes). Also, because 
the CAA provides no direction regarding consideration of alternative 
fuels, the final rule's analysis includes simulation of the potential 
that some manufacturers might introduce new AFVs in response to 
CO2 standards. To represent the compliance benefit from such 
a response fully, NHTSA modified the CAFE model to include the specific 
provisions related to AFVs under the CO2 standards. In 
particular, the CAFE model now carries a full representation of the 
production multipliers related to electric vehicles, fuel cell 
vehicles, plug-in hybrids, and CNG vehicles, all of which vary by year 
through MY 2021.
---------------------------------------------------------------------------

    \417\ 49 U.S.C. 32902(h).
    \418\ Dedicated compressed natural gas (CNG) vehicles should 
also be excluded in this perspective but are not considered as a 
compliance strategy under any perspective in this analysis.
---------------------------------------------------------------------------

    EPCA also provides that CAFE levels may, subject to limitations, be 
adjusted upward to reflect the sale of flexible fuel vehicles (FFVs). 
Although these adjustments end after model year 2020, the final rule's 
analysis, like the NPRM's, includes estimated potential use through MY 
2019, as summarized below:

[[Page 24315]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.106

    For its part, EPA has provided that manufacturers selling 
sufficient numbers of PHEVs, BEVs, and FCVs may, when calculating fleet 
average CO2 levels, ``count'' each unit of production as 
more than a single unit. The CAFE model accounts for these 
``multipliers.'' As for the NPRM, the final rule's analysis applies the 
following multipliers:
[GRAPHIC] [TIFF OMITTED] TR30AP20.107

    For example, under EPA's current regulation, when calculating the 
average CO2 level achieved by its MY 2019 passenger car 
fleet, a manufacturer may treat each 1,000 BEVs as 2,000 BEVs. When 
calculating the average level required of this fleet, the manufacturer 
must use the actual production volume (in this example, 1,000 units). 
Similarly, the manufacturer must use the actual production volume when 
calculating compliance credit balances.
    There were no natural gas vehicles in the baseline fleet, and the 
analysis did not apply natural gas technology due to cost 
effectiveness. The application of a 2.0 multiplier for natural gas 
vehicles for MYs 2022-2026 would have no impact on the analysis because 
given the state of natural gas vehicle refueling infrastructure, the 
cost to equip vehicles with natural gas tanks, the outlook for 
petroleum prices, and the outlook for battery prices, we have little 
basis to project more than an inconsequential response to this 
incentive in the foreseeable future.
    For the final rule's analysis, the CAFE model can be exercised in a 
manner that simulates these current EPA requirements, or that simulates 
two alternative approaches. The first includes the above-mentioned 
multipliers in the calculation of average requirements, and the second 
also includes the multipliers in the calculation of credit balances. 
The central analysis reflects current regulations. The sensitivity 
analysis presented in the FRIA includes a case

[[Page 24316]]

applying multipliers to the calculation of achieved and required 
average CO2 levels, and calculation of credit balances.
c) Civil Penalties
    Throughout the history of the CAFE program, some manufacturers have 
consistently achieved fuel economy levels below applicable standards, 
electing instead to pay civil penalties as specified by EPCA. As in 
previous versions of the CAFE model, the current version allows the 
user to specify inputs identifying such manufacturers and to consider 
their compliance decisions as if they are willing to pay civil 
penalties for non-compliance with the CAFE program. As with the NPRM, 
the civil penalty rate in the current analysis is $5.50 per 1/10 of a 
mile per gallon, per vehicle manufactured for sale.
    NHTSA notes that treating a manufacturer as if it is willing to pay 
civil penalties does not necessarily mean that it is expected to pay 
penalties in reality. Doing so merely implies that the manufacturer 
will only apply fuel economy technology up to a point, and then stop, 
regardless of whether or not its corporate average fuel economy is 
above its standard. In practice, the agencies expect that many of these 
manufacturers will continue to be active in the credit market, using 
trades with other manufacturers to transfer credits into specific 
fleets that are challenged in any given year, rather than paying 
penalties to resolve CAFE deficits. The CAFE model calculates the 
amount of penalties paid by each manufacturer, but it does not simulate 
trades between manufacturers. In practice, some (possibly most) of the 
total estimated penalties may be a transfer from one OEM to another.
    Although EPCA, as amended in 2007 by the Energy Independence and 
Security Act (EISA), prescribes these specific civil penalty provisions 
for CAFE standards, the Clean Air Act (CAA) does not contain similar 
provisions. Rather, the CAA's provisions regarding noncompliance 
prohibit sale of a new motor vehicle that is not covered by an EPA 
certificate of conformity, and in order to receive such a certificate 
the new motor vehicle must meet EPA's Section 202 regulations, 
including applicable emissions standards. Therefore, inputs regarding 
civil penalties--including inputs regarding manufacturers' potential 
willingness to treat civil penalty payment as an economic choice--apply 
only to simulation of CAFE standards. On the other hand, some of the 
same manufacturers recently opting to pay civil penalties instead of 
complying with CAFE standards have also recently led adoption of lower-
GWP refrigerants, and the ``A/C leakage'' credits count toward 
compliance only with CO2 standards, not CAFE standards. The 
model accounts for this difference between the programs.
    When considering technology applications to improve fleet fuel 
economy, the model will add technology up to the point at which the 
effective cost of the technology (which includes technology cost, 
consumer fuel savings, consumer welfare changes, and the cost of 
penalties for non-compliance with the standard) is less costly than 
paying civil penalties or purchasing credits. Unlike previous versions 
of the model, the current implementation further acknowledges that some 
manufacturers experience transitions between product lines where they 
rely heavily on credits (either carried forward from earlier model 
years or acquired from other manufacturers) or simply pay penalties in 
one or more fleets for some number of years. The model now allows the 
user to specify, when appropriate for the regulatory program being 
simulated, on a year-by-year basis, whether each manufacturer should be 
considered as willing to pay penalties for non-compliance. This 
provides additional flexibility, particularly in the early years of the 
simulation. As discussed above, this assumption is best considered as a 
method to allow a manufacturer to under-comply with its standard in 
some model years--treating the civil penalty rate and payment option as 
a proxy for other actions it may take that are not represented in the 
CAFE model (e.g., purchasing credits from another manufacturer, carry-
back from future model years, or negotiated settlements with NHTSA to 
resolve deficits).
    For the NPRM, NHTSA relied on past compliance behavior and 
certified transactions in the credit market to designate some 
manufacturers as willing to pay CAFE penalties in some model years. The 
full set of NPRM assumptions regarding manufacturer behavior with 
respect to civil penalties is presented in Table VI-21, which shows all 
manufacturers were assumed to be willing to pay civil penalties prior 
to MY 2020. This was largely a reflection of either existing credit 
balances (which manufacturers will use to offset CAFE deficits until 
the credits reach their expiration dates) or inter-manufacturer trades 
assumed likely to happen in the near future, based on previous 
behavior. The manufacturers in the table whose names appear in bold all 
had at least one regulated fleet (of three) whose CAFE was below its 
standard in MY 2016. Because the NPRM analysis began with the MY 2016 
fleet, and no technology could be added to vehicles that are already 
designed and built, all manufacturers could generate civil penalties in 
MY 2016. However, once a manufacturer is designated as unwilling to pay 
penalties, the CAFE model will attempt to add technology to the 
respective fleets to avoid shortfalls.

[[Page 24317]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.108

    Several of the manufacturers in Table VI-21 that were presumed to 
be willing to pay civil penalties in the early years of the program 
have no history of paying civil penalties. However, several of those 
manufacturers have either bought or sold credits--or transferred 
credits from one fleet to another to offset a shortfall in the 
underperforming fleet. As the CAFE model does not simulate credit 
trades between manufacturers, providing this additional flexibility in 
the modeling avoids the outcome where the CAFE model applies more 
technology than needed in the context of the full set of compliance 
flexibilities at the industry level. By statute, NHTSA cannot consider 
credit flexibilities when setting standards, so most manufacturers 
(those without a history of civil penalty payment) are assumed to 
comply with their standards through fuel economy improvements for the 
model years being considered in this analysis. The notable exception to 
this assumption is Fiat Chrysler Automobiles (FCA), which could still 
satisfy the requirements of the program through a combination of credit 
application and civil penalties through MY 2025 before eventually 
complying exclusively through fuel economy improvements in MY 2026.
    As mentioned above, the CAA does not provide civil penalty 
provisions similar to those provisions specified in EPCA/EISA, and the 
above-mentioned corresponding inputs apply only to simulation of 
compliance with CAFE standards.
    Some stakeholders offering comments related to the analytical 
treatment of civil penalties indicated that NHTSA should tend toward 
assuming manufacturers will take advantage of this EPCA provision as an 
economically attractive alternative to compliance. Other commenters 
implied that NHTSA should tend toward not relying on compliance 
flexibilities in the analysis used to determine the maximum feasible 
stringency of CAFE standards. For example, New York University's 
Institute for Policy Integrity (IPI) offered the following comments:

    NHTSA assumes that most manufacturers will be unwilling to pay 
penalties based in part on the fact that most manufacturers have not 
paid penalties in recent years. The Proposed Rule cites the 
statutory prohibition on NHTSA considering credit trading as a 
reason to assume manufacturers without a history of paying penalties 
will comply through technology alone, whatever the cost. But this is 
an arbitrary assumption and is in no way dictated by the statute. 
NHTSA knows as much, since elsewhere in the proposed rollback, the 
agency explains ``EPCA is very clear as to which flexibilities are 
not to be considered'' and NHTSA is allowed to consider off-cycle 
adjustments because they are not specifically mentioned. But 
considering penalties are not mentioned as off-limits for NHTSA in 
setting the standards either. Instead, the prohibition focuses on 
credit trading and transferring. The penalty safety valve has 
existed in EPCA for decades, and Congress clearly would have known 
how to add penalties to the list of trading and transferring. The 
fact that Congress did not bar NHTSA from considering penalties as a 
safety valve means that NHTSA must consider manufacturer's efficient 
use of penalties as a cost minimizing compliance option. Besides, 
NHTSA does consider penalties for some of the manufacturers making 
its statutory justification even less rational.\419\
---------------------------------------------------------------------------

    \419\ Institute for Policy Integrity, NHTSA-2018-0067-12213, at 
24.

    On the other hand, in more general comments about NHTSA's 
analytical treatment of program flexibilities, FCA stated that ``when 
flexibilities are considered while setting targets, they cease to be 
flexibilities and become simply additional technology mandates.'' \420\
---------------------------------------------------------------------------

    \420\ FCA, Docket #NHTSA-2018-0067-11943, at 6.
---------------------------------------------------------------------------

    NHTSA agrees with IPI that EPCA does not expressly prohibit NHTSA, 
when conducting analysis supporting determinations of the maximum 
feasible stringency of future CAFE standards, from including 
manufacturers' potential tendency to pay civil penalties rather than 
complying with those standards. However, EPCA also does not require 
NHTSA to include this tendency in its analysis. NHTSA also notes, as 
does IPI, that EPCA does prohibit NHTSA from including credit trading, 
transferring, or the availability of credits in such

[[Page 24318]]

analysis (although NHTSA interprets this prohibition to apply only to 
the model years for which standards are being set). This statutory 
difference is logical based on the way credits and penalties function 
differently under EPCA. Because credits help manufacturers achieve 
compliance with CAFE standards, absent the statutory prohibition, 
credits would be relevant to the feasibility of a standard.\421\ 
Penalties, on the other hand, do not enable a manufacturer to comply 
with an applicable standard; penalties are for noncompliance.\422\ When 
Congress added credit trading provisions to EPCA in 2007, NHTSA 
anticipated that competitive considerations would make manufacturers 
reluctant to engage in such trades. Since that time, manufacturers 
actually have demonstrated otherwise, although the reliance on 
trading--especially between specific pairs of OEMs--appears to vary 
widely. At this time, NHTSA considers it most likely that manufacturers 
will shift away from paying civil penalties and toward compliance 
credit trading. Consequently, for NHTSA to include civil penalty 
payment in its analysis would increasingly amount to using civil 
penalty payment as an analytical proxy for credit trading. Having 
further considered the question, NHTSA's current view is, therefore, 
that including civil penalty payment beyond MY 2020 would effectively 
subvert EPCA's prohibition against considering credit trading. 
Therefore, for today's announcement, NHTSA has modified its analysis to 
assume that BMW, Daimler, FCA, JLR, and Volvo would consider paying 
civil penalties through MY 2020, and that all manufacturers would apply 
as much technology as would be needed in order to avoid paying civil 
penalties after MY 2020.
---------------------------------------------------------------------------

    \421\ See 49 U.S.C. 32911(b) (``Compliance is determined after 
considering credits available to the manufacturer . . . . '').
    \422\ See id.
---------------------------------------------------------------------------

3. Technology Effectiveness Values
    The next input required to simulate manufacturers' decision-making 
processes for the year-by-year application of technologies to specific 
vehicles is estimates of how effective each technology would be at 
reducing fuel consumption. In the NPRM, the agencies used full-vehicle 
modeling and simulation to estimate the fuel economy improvements 
manufacturers could make to a fleet of vehicles, considering those 
vehicles' technical specifications and how combinations of technologies 
interact. Full-vehicle modeling and simulation uses computer software 
and physics-based models to predict how combinations of technologies 
perform as a full system under defined conditions.
    A model is a mathematical representation of a system, and 
simulation is the behavior of that mathematical representation over 
time. In this analysis, the model is a mathematical representation of 
an entire vehicle,\423\ including its individual components such as the 
engine and transmission, overall vehicle characteristics such as mass 
and aerodynamic drag, and the environmental conditions, such as ambient 
temperature and barometric pressure. The agencies simulated the model's 
behavior over test cycles, including the 2-cycle laboratory compliance 
tests (or 2-cycle tests),\424\ to determine how the individual 
components interact. 2-cycle tests are test cycles that are used to 
measure fuel economy and emissions for CAFE and CO2 
compliance, and therefore are the relevant test cycles for determining 
technology effectiveness when establishing standards. In the 
laboratory, 2-cycle testing involves sophisticated test and measurement 
equipment, carefully controlled environmental conditions, and precise 
procedures to provide the most repeatable results possible with human 
drivers. Measurements using these structured procedures serve as a 
yardstick for fuel economy and CO2 emissions.
---------------------------------------------------------------------------

    \423\ Our full vehicle model was composed of sub-models, which 
is why the full vehicle model could also be referred to as a full 
system model, composed of sub-system models.
    \424\ EPA's compliance test cycles are used to measure the fuel 
economy of a vehicle. For readers unfamiliar with this process, it 
is like running a car on a treadmill following a program--or more 
specifically, two programs. The ``programs'' are the ``urban 
cycle,'' or Federal Test Procedure (abbreviated as ``FTP''), and the 
``highway cycle,'' or Highway Fuel Economy Test (abbreviated as 
``HFET''), and they have not changed substantively since 1975. Each 
cycle is a designated speed trace (of vehicle speed versus time) 
that all certified vehicles must follow during testing. The FTP is 
meant roughly to simulate stop and go city driving, and the HFET is 
meant roughly to simulate steady flowing highway driving at about 50 
mph. For further details on compliance testing, see the discussion 
in Section VI.B.3.a)(7).
---------------------------------------------------------------------------

    Full-vehicle modeling and simulation was initially developed to 
avoid the costs of designing and testing prototype parts for every new 
type of technology. For example, if a truck manufacturer has a concept 
for a lightweight tailgate and wants to determine the fuel economy 
impact for the weight reduction, the manufacturer can use physics-based 
computer modeling to estimate the impact. The vehicle, modeled with the 
proposed change, can be simulated on a defined test route and under a 
defined test condition, such as city or highway driving in warm ambient 
temperature conditions, and compared against the baseline reference 
vehicle. Full-vehicle modeling and simulation allows the consideration 
and evaluation of different designs and concepts before building a 
single prototype. In addition, full vehicle modeling and simulation is 
beneficial when considering technologies that provide small incremental 
improvements. These improvements are difficult to measure in laboratory 
tests due to variations in how vehicles are driven over the test cycle 
by human drivers, variations in emissions measurement equipment, and 
variations in environmental conditions.\425\
---------------------------------------------------------------------------

    \425\ Difficulty with controlling for such variability is 
reflected, for example, in 40 CFR 1065.210, Work input and output 
sensors, which describes complicated instructions and 
recommendations to help control for variability in real world (non-
simulated) test instrumentation set up.
---------------------------------------------------------------------------

    Full-vehicle modeling and simulation requires detailed data 
describing the individual technologies and performance-related 
characteristics. Those specifications generally come from design 
specifications, laboratory measurements, and other subsystem 
simulations or modeling. One example of data used as an input to the 
full vehicle simulation are engine maps for each engine technology that 
define how much fuel is consumed by the engine technology across its 
operating range.
    Using full-vehicle modeling and simulation to estimate technology 
efficiency improvements has two primary advantages over using single or 
limited point estimates. An analysis using single or limited point 
estimates may assume that, for example, one fuel economy improving 
technology with an effectiveness value of 5 percent by itself and 
another technology with an effectiveness value of 10 percent by itself, 
when applied together achieve an additive improvement of 15 percent. 
Single point estimates generally do not provide accurate effectiveness 
values because they do not capture complex relationships among 
technologies. Technology effectiveness often differs significantly 
depending on the vehicle type (e.g., sedan versus pickup truck) and how 
the technology interacts with other technologies on the vehicle, as 
different technologies may provide different incremental levels of fuel 
economy improvement if implemented alone or in tandem with other 
technologies. Any oversimplification of these complex interactions 
leads to less accurate and often overestimated effectiveness estimates.
    In addition, because manufacturers often implement several fuel-
saving

[[Page 24319]]

technologies simultaneously when redesigning a vehicle, it is difficult 
to isolate the effect of individual technologies using laboratory 
measurement of production vehicles alone. Modeling and simulation 
offers the opportunity to isolate the effects of individual 
technologies by using a single or small number of baseline vehicle 
configurations and incrementally adding technologies to those baseline 
configurations. This provides a consistent reference point for the 
incremental effectiveness estimates for each technology and for 
combinations of technologies for each vehicle type. Vehicle modeling 
also reduces the potential for overcounting or undercounting technology 
effectiveness.
    An important feature of this analysis is that the incremental 
effectiveness of each technology and combinations of technologies be 
accurate and relative to a consistent baseline vehicle. The absolute 
fuel economy values of the full vehicle simulations are used only to 
determine incremental effectiveness and are never used directly to 
assign an absolute fuel economy value to any vehicle model or 
configuration for the rulemaking analysis.
    For this analysis, absolute fuel economy levels are based on the 
individual fuel economy values from CAFE compliance data for each 
vehicle in the baseline fleet. The incremental effectiveness from the 
full vehicle simulations performed in Autonomie, a physics-based full-
vehicle modeling and simulation software developed and maintained by 
the U.S. Department of Energy's Argonne National Laboratory, are 
applied to baseline fuel economy to determine the absolute fuel economy 
of applying the first technology change. For subsequent technology 
changes, incremental effectiveness is applied to the absolute fuel 
economy level of the previous technology configuration.
    For example, if a Ford F150 2-wheel drive crew cab and short bed in 
the baseline fleet has a fuel economy value of 30 mpg for CAFE 
compliance, 30 mpg will be considered the reference absolute fuel 
economy value. A similar full vehicle model in the Autonomie simulation 
may begin with an average fuel economy value of 32 mpg, and with 
incremental addition of a specific technology X its fuel economy 
improves to 35 mpg, a 9.3 percent improvement. In this example, the 
incremental fuel economy improvement (9.3 percent) from technology X 
would be applied to the F150's 30 mpg absolute value.
    For this analysis, the agencies determined the incremental 
effectiveness of technologies as applied to the 2,952 unique vehicle 
models in the analysis fleet. Although, as mentioned above, full-
vehicle modeling and simulation reduces the work and time required to 
assess the impact of moving a vehicle from one technology state to 
another, it would be impractical--if not impossible--to build a unique 
vehicle model for every individual vehicle in the analysis fleet. 
Therefore, as explained further below, vehicle models are built in a 
way that maintains similar attributes to the analysis fleet vehicles, 
which ensures key components are reasonably represented.
    We received a wide array of comments regarding the full-vehicle 
modeling and simulation performed for the NPRM, but there was general 
agreement that full-vehicle modeling and simulation was the appropriate 
method to determine technology effectiveness.\426\ Stakeholders 
commented on other areas, such as full vehicle simulation tools, 
inputs, and assumptions, and these comments will be discussed in the 
following sections. For this final rule, the agencies continued to use 
the same full-vehicle simulation approach to estimate technology 
effectiveness for technology adoption in the rulemaking timeframe. The 
next sections will discuss the details of the explicit input 
specifications and assumptions used for the final rule analysis.
---------------------------------------------------------------------------

    \426\ See NHTSA-2018-0067-12039; NHTSA-2018-0067-12073. UCS and 
AAM both agreed that full vehicle simulation can significantly 
improve the estimates of technology effectiveness.
---------------------------------------------------------------------------

a) Why This Rulemaking Used Autonomie Full-Vehicle Modeling and 
Simulation To Determine Technology Effectiveness
    The NPRM and final rule analysis use effectiveness estimates for 
technologies developed using Autonomie, a physics-based full-vehicle 
modeling and simulation software developed and maintained by the U.S. 
Department of Energy's Argonne National Laboratory.\427\ Autonomie was 
designed to serve as a single tool to meet requirements of automotive 
engineering throughout the vehicle development process, and has been 
under continuous improvement by Argonne for over 20 years. Autonomie is 
commercially available and widely used in the automotive industry by 
suppliers, automakers, and academic researchers (who publish findings 
in peer reviewed academic journals).\428\ DOE and manufacturers have 
used Autonomie and its ability to simulate a large number of powertrain 
configurations, component technologies, and vehicle-level controls over 
numerous drive cycles to support studies on fuel efficiency, cost-
benefit analysis, and carbon dioxide emissions,\429\ and other topics.
---------------------------------------------------------------------------

    \427\ More information about Autonomie is available at https://www.anl.gov/technology/project/autonomie-automotive-system-design 
(last accessed June 21, 2018). As mentioned in the preliminary 
regulatory impact analysis (PRIA) for this rule, the agencies used 
Autonomie version R15SP1, the same version used for the 2016 Draft 
TAR.
    \428\ Rousseau, A. Shidore, N. Karbowski, D. Sharer, ``Autonomie 
Vehicle Validation Summary.'' https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/anl-autonomie-vehicle-model-validation-1509.pdf.
    \429\ Delorme et al. 2008, Rousseau, A, Sharer, P, Pagerit, S., 
& Das, S. ``Trade-off between Fuel Economy and Cost for Advanced 
Vehicle Configurations,'' 20th International Electric Vehicle 
Symposium (EVS20), Monaco (April 2005); Elgowainy, A., Burnham, A., 
Wang, M., Molburg, J., & Rousseau, A. ``Well-To-Wheels Energy Use 
and Greenhouse Gas Emissions of Plug-in Hybrid Electric Vehicles,'' 
SAE 2009-01-1309, SAE World Congress, Detroit, April 2009.
---------------------------------------------------------------------------

    Autonomie has also been used to provide the U.S. government with 
data to make decisions about future research, and is used by DOE for 
analysis supporting budget priorities and plans for programs managed by 
its Vehicle Technologies Office (VTO), and to support decision making 
among competing vehicle technology research and development 
projects.\430\ In addition, Autonomie is the primary vehicle simulation 
tool used by DOE to support its U.S. DRIVE program, a government-
industry partnership focused on advanced automotive and related energy 
infrastructure technology research and development.\431\
---------------------------------------------------------------------------

    \430\ U.S. DOE Benefits & Scenario Analysis publications is 
available at https://www.autonomie.net/publications/fuel_economy_report.html (last accessed September 11, 2019).
    \431\ For more information on U.S. Drive, see https://www.energy.gov/eere/vehicles/us-drive.
---------------------------------------------------------------------------

    Autonomie is a MathWorks-based software environment and framework 
for automotive control-system design, simulation, and analysis.\432\ It 
is designed for rapid and easy integration of models with varying 
levels of detail (low to high fidelity), abstraction (from subsystems 
to systems and entire architectures), and processes (e.g., calibration, 
validation). By building models automatically, Autonomie allows the 
quick simulation of many component technologies and powertrain 
configurations, and, in this case, to assess the energy consumption of 
advanced powertrain technologies. Autonomie simulates subsystems,

[[Page 24320]]

systems, or entire vehicles; evaluates and analyzes fuel efficiency and 
performance; performs analyses and tests for virtual calibration, 
verification, and validation of hardware models and algorithms; 
supports system hardware and software requirements; links to 
optimization algorithms; and supplies libraries of models for 
propulsion architectures of conventional powertrains as well as hybrid 
and electric vehicles.
---------------------------------------------------------------------------

    \432\ Halbach, S. Sharer, P. Pagerit, P., Folkerts, C. & 
Rousseau, A. ``Model Architecture, Methods, and Interfaces for 
Efficient Math-Based design and Simulation of Automotive Control 
Systems,'' SAE 2010-01-0241, SAE World Congress, Detroit, April, 
2010.
---------------------------------------------------------------------------

    With hundreds of pre-defined powertrain configurations along with 
vehicle level control strategies developed from dynamometer test data, 
Autonomie is a highly capable tool for analyzing advantages and 
drawbacks of applying different technology options within each 
technology family, including conventional, parallel hybrid, power-split 
hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles 
(PHEVs), battery electric vehicles (BEV) and fuel cell vehicles (FCVs). 
Autonomie also allows users to evaluate the effect of component sizing 
on fuel consumption for different powertrain technologies as well as to 
define component requirements (e.g., power, energy) to maximize fuel 
displacement for a specific application.\433\ To evaluate properly any 
powertrain-configuration or component-sizing influence, vehicle-level 
control models are critical, especially for electric drive vehicles 
like hybrids and plug-in hybrids. Argonne has extensive expertise in 
developing vehicle-level control models based on different approaches, 
from global optimization to instantaneous optimization, rule-based 
optimization, and heuristic optimization.\434\
---------------------------------------------------------------------------

    \433\ Nelson, P., Amine, K., Rousseau, A., & Yomoto, H. (EnerDel 
Corp.), ``Advanced Lithium-ion Batteries for Plug-in Hybrid-electric 
Vehicles,'' 23rd International Electric Vehicle Symposium (EVS23), 
Anaheim, CA, (Dec. 2007); Karbowski, D., Haliburton, C., & Rousseau, 
A. ``Impact of Component Size on Plug-in Hybrid Vehicles Energy 
Consumption using Global Optimization,'' 23rd International Electric 
Vehicle Symposium (EVS23), Anaheim, CA, (Dec. 2007).
    \434\ Karbowski, D., Kwon, J., Kim, N., & Rousseau, A., 
``Instantaneously Optimized Controller for a Multimode Hybrid 
Electric Vehicle,'' SAE paper 2010-01-0816, SAE World Congress, 
Detroit, April 2010; Sharer, P., Rousseau, A., Karbowski, D., & 
Pagerit, S. ``Plug-in Hybrid Electric Vehicle Control Strategy--
Comparison between EV and Charge-Depleting Options,'' SAE paper 
2008-01-0460, SAE World Congress, Detroit (April 2008); and 
Rousseau, A., Shidore, N., Carlson, R., & Karbowski, D. ``Impact of 
Battery Characteristics on PHEV Fuel Economy,'' AABC08.
---------------------------------------------------------------------------

    Autonomie has been developed to consider real-world vehicle metrics 
like performance, hardware limitations, utility, and drivability 
metrics (e.g., towing capability, shift busyness, frequency of engine 
on/off transitions), which are important to producing realistic 
estimates of fuel economy and CO2 emission rates. This 
increasing realism has, in turn, steadily increased confidence in the 
appropriateness of using Autonomie to make significant investment 
decisions. Autonomie has also been validated for a number of powertrain 
configurations and vehicle classes using Argonne's Advanced Mobility 
Technology Laboratory (AMTL) (formerly Advanced Powertrain Research 
Facility, or APRF) vehicle test data.\435\
---------------------------------------------------------------------------

    \435\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A., 
``Analysis and Model Validation of the Toyota Prius Prime.'' SAE 
2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N, 
Jeong, J. Rousseau, A. & Lohse-Busch, H. ``Control Analysis and 
Thermal Model Development of PHEV,'' SAE 2015-01-1157, SAE World 
Congress, Detroit, April 2015; Kim, N., Rousseau, A. & Lohse-Busch, 
H. ``Advanced Automatic Transmission Model Validation Using 
Dynamometer Test Data,'' SAE 2014-01-1778, SAE World Congress, 
Detroit, Apr. 14; Lee, D. Rousseau, A. & Rask, E. ``Development and 
Validation of the Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE 
World Congress, Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., & 
Duoba, M. ``Validating Volt PHEV Model with Dynamometer Test Data 
using Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit, 
Apr. 13; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model 
Validation with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040, 
SAE World Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A, 
Pagerit, S., & Sharer, P. ``Plug-in Vehicle Control Strategy--From 
Global Optimization to Real Time Application,'' 22th International 
Electric Vehicle Symposium (EVS22), Yokohama, (October 2006).
---------------------------------------------------------------------------

    Argonne has spent several years developing, applying, and expanding 
the means to use distributed computing to exercise its Autonomie full-
vehicle simulation tool over the scale necessary for realistic analysis 
to provide data for CAFE and CO2 standards rulemaking. The 
NPRM and PRIA detailed how Argonne used Autonomie to estimate the fuel 
economy impacts for roughly a million combinations of technologies and 
vehicle types.436 437 Argonne developed input parameters for 
Autonomie to represent every combination of vehicle, powertrain, and 
component technologies considered in this rulemaking. The sequential 
addition of more than 50 fuel economy-improving technologies to ten 
vehicle types generated more than 140,000 unique technology and vehicle 
combinations. Running the Autonomie powertrain sizing algorithms to 
determine the appropriate amount of engine downsizing needed to 
maintain overall vehicle performance when vehicle mass reduction is 
applied and for certain engine technology changes (discussed further, 
below) increased the total number of simulations to more than one 
million. The result of these simulations is a useful dataset 
identifying the impacts of combinations of vehicle technologies on 
energy consumption--a dataset that can be referenced as an input to the 
CAFE model for assessing regulatory compliance alternatives.
---------------------------------------------------------------------------

    \436\ As part of the Argonne simulation effort, individual 
technology combinations simulated in Autonomie were paired with 
Argonne's BatPAC model to estimate the battery cost associated with 
each technology combination based on characteristics of the 
simulated vehicle and its level of electrification. Information 
regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
    \437\ Additionally, the impact of engine technologies on fuel 
consumption, torque, and other metrics was characterized using GT 
POWER simulation modeling in combination with other engine modeling 
that was conducted by IAV Automotive Engineering, Inc. (IAV). The 
engine characterization ``maps'' resulting from this analysis were 
used as inputs for the Autonomie full-vehicle simulation modeling. 
Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
---------------------------------------------------------------------------

    The following sections discuss the full-vehicle modeling and 
simulation inputs and data assumptions, and comments received on the 
NPRM analysis. The discussion is necessarily technical, but also 
important to understand the agencies' decisions to modify (or not) the 
Autonomie analysis for the final rule.
(1) Full-Vehicle Modeling, Simulation Inputs and Data Assumptions
    The agencies provided extensive documentation that quantitatively 
and qualitatively described the over 50 technologies considered as 
inputs to the Autonomie modeling.438 439 These inputs 
consisted of engine technologies, transmission technologies, powertrain 
electrification, light-weighting, aerodynamic improvements, and tire 
rolling resistance improvements.\440\ The PRIA provided an overview of 
the sub-models for each technology, including the internal combustion 
engine model, automatic transmission model, and others.\441\ The 
Argonne NPRM model documentation expanded on these sub-models in detail 
to show the interaction of each sub-model input and output.\442\

[[Page 24321]]

For example, as shown in Figure VI-2, the input for Autonomie's driver 
model (i.e., the model used to approximate the driving behavior of a 
real driver) is vehicle speed, and outputs are accelerator pedal, brake 
pedal, and torque demand.
---------------------------------------------------------------------------

    \438\ NHTSA-2018-0067-12299. Preliminary Regulatory Impact 
Analysis (July 2018).
    \439\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N, 
Rousseau, A. ``A Detailed Vehicle Simulation Process To Support CAFE 
Standards 04262018--Report'' ANL Autonomie Documentation. Aug 21, 
2018. NHTSA-2018-0067-0004. ANL Autonomie Data Dictionary. Aug 21, 
2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main Component 
Assumptions. Aug 21, 2018. NHTSA-2018-0067-0005. ANL Autonomie Model 
Assumptions Summary. Aug 21, 2018. NHTSA-2018-0067-1692. ANL BatPac 
Model 12 55. Aug 21, 2018.
    \440\ SAFE Rule for MY2021-2026 PRIA Chapter 6.2.3 Technology 
groups in Autonomie simulations and CAFE model.
    \441\ PRIA at 189.
    \442\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N, 
Rousseau, A. ``A Detailed Vehicle Simulation Process To Support CAFE 
Standards 04262018--Report'' ANL Autonomie Documentation. Aug 21, 
2018.
[GRAPHIC] [TIFF OMITTED] TR30AP20.109

    Effectiveness inputs for the NPRM and the final rule analysis were 
specifically developed to consider many real world and compliance test 
cycle constraints, to the extent a computer model could capture them. 
Examples include the advanced engine knock model discussed below, in 
addition to other constraints like allowing cylinder deactivation to 
occur in ways that would not negatively impact noise-vibration-
harshness (NVH), and similarly optimizing the number of engine on/off 
events (e.g., from start/stop 12V micro hybrid systems) to balance 
between effectiveness and NVH.
    One major input used in the effectiveness modeling that the 
agencies provided key specifications for in the PRIA are engine fuel 
maps that define how an engine equipped with specific technologies 
operates over a variety of engine load (torque) and engine speed 
conditions. The engine maps used as inputs to the Autonomie modeling 
portion of the analysis were developed by starting with a base map and 
then modifying that base map, incrementally, to model the addition of 
engine technologies. These engine maps, developed using the GT-Power 
modeling tool by IAV, were based off real-world engine designs. 
Simulated operation of these engines included the application of an IAV 
knock model, also developed from real-world engine 
data.443 444 Using this process, which incorporated real-
world data, ensured that real-world constraints were considered for 
each vehicle type. Although the same type of engine map is used for all 
technology classes, the effectiveness varies based on the 
characteristics of each vehicle type. For example, a compact car with a 
turbocharged engine will have different fuel economy and performance 
values than a pickup truck with the same engine technology type. The 
engine map specifications are discussed further in Section VI.C.1 of 
this preamble and Section VI of FRIA.
---------------------------------------------------------------------------

    \443\ Engine knock in spark ignition engines occurs when 
combustion of some of the air/fuel mixture in the cylinder does not 
result from propagation of the flame front ignited by the spark 
plug, but one or more pockets of air/fuel mixture explodes outside 
of the envelope of the normal combustion front.
    \444\ See IAV material submitted to the docket; IAV_20190430_Eng 
22-26 Updated_Docket.pdf, 
IAV_Engine_tech_study_Sept_2016_Docket.pdf, IAV_Study for 4 Cylinder 
Gas Engines_Docket.pdf.
---------------------------------------------------------------------------

    The agencies also provided key details about input assumptions for 
various vehicle specifications like transmission gear ratios, tire 
size, final drive ratios, and individual component weights.\445\ Each 
of these assumptions, to some extent, varied between the ten technology 
classes to capture appropriately real-world vehicle specifications like 
wheel mass or fuel tank mass. These specific input assumptions were 
developed based on the latest test data and current market fleet 
information.\446\ The agencies relied on default assumptions developed 
by the Autonomie team, based on test data and technical publication 
review, for other model inputs required by Autonomie, such as throttle 
time response and shifting strategies for different transmission 
technologies. The Autonomie modeling tool did not simulate vehicle 
attributes determined to have minimal impacts, like whether a vehicle 
had a sun roof or hood scoops, as those attributes would have trivial 
impact in the overall analysis.
---------------------------------------------------------------------------

    \445\ ANL Autonomie Model Assumptions Summary. Aug 21, 2018, 
NHTSA-2018-0067-0005. ANL--Summary of Main Component Performance and 
Assumptions NPRM. Aug 21, 2018, NHTSA-2018-0067-0003.
    \446\ See further details in Section VI.B.1 Analysis Fleet.
---------------------------------------------------------------------------

    Because the agencies model ten different vehicle types to represent 
the 2,952 vehicles in the baseline fleet, improper assumptions about an 
advanced technology could lead to errors in estimating effectiveness. 
Autonomie is a sophisticated full-vehicle modeling tool that requires 
extensive technology characteristics based on both physical and 
intangible data, like proprietary software. With a few technologies, 
the agencies did not have publicly available data, but had received 
confidential business information confirming such technologies 
potential availability in the market during the rulemaking time frame. 
For such technologies, including advanced cylinder deactivation, the 
agencies adopted a method in the CAFE model to represent the 
effectiveness of the technology, and did not explicitly simulate the 
technologies in the Autonomie model. For this limited set of 
technologies, the agencies determined that effectiveness could 
reasonably be represented as a fixed value.\447\ Effectiveness values 
for technologies not explicitly simulated in Autonomie are discussed 
further in the individual technology sections of this preamble.
---------------------------------------------------------------------------

    \447\ For final rule, 9 out of 50 plus technologies use fixed 
offset effectiveness values. The total effectiveness of these 
technologies cannot be captured on the 2-cycle test or, like ADEAC, 
they are a new technology where robust data that could be used as an 
input to the technology effectiveness modeling does not yet exist. 
Specifically, these nine technologies are LDB, SAX, EPS, IACC, EFR, 
ADEAC, DSLI, DSLIAD and TURBOAD.
---------------------------------------------------------------------------

    The agencies sought comments on all effectiveness inputs and input 
assumptions, including the specific data used to characterize the 
technologies,

[[Page 24322]]

such as data to build the technology input, data representing operating 
range of technologies, and data for variation among technology inputs. 
The agencies also sought comment on the effectiveness values used for 
technologies not explicitly defined in Autonomie.
    Meszler Engineering Services, commenting on behalf of the Natural 
Resources Defense Council, and ICCT questioned the accuracy of the 
effectiveness estimates in the Argonne database, and as an example 
Meszler analyzed the fuel economy impacts of a 10-speed automatic 
transmission relative to a baseline 8-speed automatic transmission, 
concluding that the widely ranging effectiveness estimates were 
unexpected. ICCT questioned the accuracy of the IAV engine maps that 
serve as an input to the Autonomie effectiveness modeling, and asked 
whether those could ``reasonably stand as a foundation for automotive 
developments and technology combinations'' discussed elsewhere in their 
comments. ICCT also questioned whether Autonomie realistically and 
validly modeled synergies between technologies, using the effectiveness 
values from CEGR and transmissions as an example. Meszler stated that 
the agencies have an obligation to validate the Autonomie estimates 
before using them to support the NPRM or any other rulemaking. The 
agencies also received comments on the specific effectiveness estimates 
generated by Autonomie; however, those comments will be discussed in 
each individual technology section, below.
    Despite these criticisms, Meszler stated that the critiques of the 
Autonomie technology database were not meant to imply that the 
Autonomie vehicle simulation model used to develop the database was 
fundamentally flawed, or that the model could not be used to derive 
accurate fuel economy impact estimates. Meszler noted that, as with any 
model, estimates derived with Autonomie are only valid for a given set 
of modeling parameters and if those parameters are well defined, the 
estimates should be accurate and reliable. Conversely, if those 
parameters are not well defined, the estimates would be inaccurate and 
unreliable. Meszler stated that the agencies must make the full set of 
modeling assumptions used for the Autonomie database available for 
review and comment.
    We agree with Meszler that, in general, when inputs to a model are 
inaccurate, output effectiveness results may be too high or too low. 
The technology effectiveness estimates from modeling results often vary 
with the type of vehicle and the other technologies that are on that 
vehicle.\448\ The Autonomie output database consists of permutations of 
over 50 technologies for each of the ten technology classes simulated 
by the CAFE model. A wide range of effectiveness is expected when going 
from a baseline technology to an advanced technology across different 
technology classes because there are significant differences in how 
much power is required from the powertrain during 2-cycle testing 
across the ten vehicle types. This impacts powertrain operating 
conditions (e.g., engine speed and load) during 2-cycle testing. Fuel 
economy improving technologies have different effectiveness at each of 
those operating conditions so vehicles that have higher average power 
demands will have different effectiveness than vehicles with lower 
average power demands. Further, the differences in effectiveness at 
higher power and lower power vary by technology so the overall 
relationship is complex. Large-scale full-vehicle modeling and 
simulation account for these interactions and complexities.
---------------------------------------------------------------------------

    \448\ The PRIA Chapter 6.2.2.1, Table 6-2 and Table 6-3 defined 
the characteristics of the reference technology classes that 
representative of the analysis fleet.
---------------------------------------------------------------------------

    Before conducting any full-vehicle modeling and simulation, the 
agencies spent a considerable amount of time and effort developing the 
specific inputs used for the Autonomie analysis. The agencies believe 
that these technology inputs provide reasonable estimates for the 
light-duty vehicle technologies the agencies expect to be available in 
the market in the rulemaking timeframe. As discussed earlier, these 
inputs vary in effectiveness due to how different vehicles, like 
compact cars and pickup trucks, operate on the 2-cycle test and in the 
real world. Some technologies, such as 10-speed automatic transmissions 
(AT10) relative to 8-speed automatic transmissions (AT8), can and 
should have different effectiveness results in the analysis between two 
different technology classes.\449\ These unique synergistic effects can 
only be taken into account through conducting full-vehicle modeling and 
simulation, which the agencies did here.
---------------------------------------------------------------------------

    \449\ Separately, the agencies modified specific transmission 
modeling parameters for the final rule after additional review, 
including a thorough review of public comments, and this review is 
discussed in detail in Section VI.C.2.
---------------------------------------------------------------------------

    With regards to Meszler's comment that the agencies have an 
obligation to validate the Autonomie estimates before using them to 
support the NPRM or any other rulemaking, the agencies would like to 
point Meszler to the description of the Argonne Autonomie team's robust 
process for vehicle model validation that was contained in the 
PRIA.\450\ To summarize, the NPRM and final rule analysis leveraged 
extensive vehicle test data collected by Argonne National 
Laboratory.\451\ Over the past 20 years, the Argonne team has developed 
specific instrumentation lists and test procedures for collecting 
sufficient information to develop and validate full vehicle models. In 
addition, the agencies described the Argonne team's efforts to validate 
specific component models as well, such as the advanced automatic 
transmission and dual clutch transmission models.\452\
---------------------------------------------------------------------------

    \450\ PRIA at 216-7. See also N. Kim, A. Rousseau, E. Rask, 
``Autonomie Model Validation with Test Data for 2010 Toyota Prius,'' 
SAE 2012-01-1040, SAE World Congress, Detroit, Apr12. https://www.autonomie.net/docs/5%20-%20Presentations/Validation/SAE%202012-01-1040.pdf; Vehicle Validation Status, February 2010 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/vehicle_validation_status.pdf; Tahoe HEV Model Development in PSAT, 
SAE paper 2009-01-1307, April 2009 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/tahoe_hev.pdf; PHEV Model 
Validation, U.S.DOE Merit Review 2008 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/phev_model_validation.pdf ; 
PHEV HyMotion Prius model validation and control improvements, 23rd 
International Electric Vehicle Symposium (EVS23), Dec. 2007 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/phev_hymotion_prius.pdf; Integrating Data, Performing Quality 
Assurance, and Validating the Vehicle Model for the 2004 Prius Using 
PSAT, SAE paper 2006-01-0667, April 2006; https://www.autonomie.net/docs/5%20-%20Presentations/Validation/integrating_data.pdf.
    \451\ A list of the vehicles that have been tested at the APRF 
can be found under http://www.anl.gov/energy-systems/group/downloadable-dynamometer-database.
    \452\ Kim, N., Rousseau, N., Lohse-Bush, H. ``Advanced Automatic 
Transmission Model Validation Using Dynamometer Test Data,'' SAE 
2014-01-1778, SAE World Congress, Detroit, April 2014; Kim, N., 
Lohse-Bush, H., Rousseau, A. ``Development of a model of the dual 
clutch transmission in Autonomie and validation with dynamometer 
test data,'' International Journal of Automotive Technologies, March 
2014, Volume 15, Issue 2, pp 263-71.
---------------------------------------------------------------------------

    The agencies also described the process for validating inputs used 
to develop the IAV engine maps,453 454 another input to the 
Autonomie simulations. As discussed in the PRIA, IAV's engine model 
development relied on a collection of sub-models that controlled 
independent combustion characteristics such as heat release, combustion 
knock, friction, heat flow, and other combustion optimization tools. 
These sub-models and other

[[Page 24323]]

computational fluid dynamics models were utilized to convert test data 
for use in the IAV engine map development. Specific combustion 
parameters, like from test data for the coefficient of variation for 
the indicated mean effective pressure (COV of IMEP), which is a common 
variable for combustion stability in a spark ignited engine, was used 
to assure final engine models were reasonable. The assumptions and 
inputs used in the modeling and validation of engine model results 
leveraged IAV's global engine database, which included benchmarking 
data, engine test data, single cylinder test data and prior modeling 
studies, and also technical publications and information presented at 
conferences. The agencies referenced in the PRIA that engine maps were 
validated with engine dynamometer test data to the maximum extent 
possible.\455\ Because the NPRM and the final rule analysis considered 
some technologies not yet in production, the agencies relied on 
technical publications and engine modeling by IAV to develop and 
corroborate inputs and input assumptions where engine dynamometer test 
data was not available.
---------------------------------------------------------------------------

    \453\ See PRIA at 251.
    \454\ See IAV material submitted to the docket; IAV_20190430_Eng 
22-26 Updated_Docket.pdf, 
IAV_Engine_tech_study_Sept_2016_Docket.pdf, IAV_Study for 4 Cylinder 
Gas Engines_Docket.pdf.
    \455\ See PRIA at 288.
---------------------------------------------------------------------------

    In addition, as described earlier in this section, the full set of 
NPRM modeling assumptions used for the Autonomie database were 
available for review and comment in the docket for this 
rulemaking.\456\ The full set of modeling assumptions used for the 
final rule are also available in the docket.\457\
---------------------------------------------------------------------------

    \456\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N, 
Rousseau, A., ``A Detailed Vehicle Simulation Process To Support 
CAFE Standards 04262018--Report'' ANL Autonomie Documentation. Aug 
21, 2018. NHTSA-2018-0067-0004. ANL Autonomie Data Dictionary. Aug 
21, 2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main 
Component Assumptions. Aug 21, 2018. NHTSA-2018-0067-0005. ANL 
Autonomie Model Assumptions Summary. Aug 21, 2018. NHTSA-2018-0067-
1692. ANL BatPac Model 12 55. Aug 21, 2018. Preliminary Regulatory 
Impact Analysis (July 2018). Posted July 2018 and updated August 23 
and October 16, 2018.
    \457\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting 
today's notice.
---------------------------------------------------------------------------

    Both ICCT and Meszler also commented on the availability of 
technologies within the Autonomie database, with Meszler stating that 
with limited exceptions, technologies were not included in the NPRM 
CAFE model if they were not included in the simulation modeling that 
underlay the Argonne database, and accordingly if a combination of 
technologies was not modeled during the development of the Argonne 
database, that package (or combination) of technologies was not 
available for adoption in the CAFE model. Meszler stated that these 
constraints limited the slate of technologies available to respond to 
fuel economy standards, and independently expanding the model to 
include additional technologies or technology combinations is not 
trivial.
    ICCT gave specific examples of key efficiency technologies that it 
stated Autonomie did not include, like advanced DEAC, VCR, Miller 
Cycle, e-boost, and HCCI. ICCT argued that this was especially 
problematic as the agencies appeared to have available engine maps from 
IAV on advanced DEAC, VCR, Miller Cycle, E-boost (and from advanced 
DEAC, VCR, Miller Cycle, E-boost, HCCI from EPA) that Argonne or the 
agencies have been unable to or opted not to include in their modeling. 
ICCT stated that the agencies must disclose how Autonomie had been 
updated to incorporate ``cutting edge'' 2020-2025 automotive 
technologies to ensure they reflect available improvements.\458\
---------------------------------------------------------------------------

    \458\ ICCT also made the same request of EPA's ALPHA model, and 
the agencies' response to that comment is discussed in Section 
VI.C.1 Engine Paths, below.
---------------------------------------------------------------------------

    The agencies have updated the final rule analysis to include 
additional technologies. In the NPRM, the agencies presented the engine 
maps for all of the technologies that ICCT listed, except HCCI, and 
sought comment on the engine maps, technical assumptions and the 
potential use of the technologies for the final rule analysis. Based on 
the available technical information and the ICCT and Meszler comments, 
for the final rule analysis, VCR, Miller Cycle (VTG), and e-boost (VTGe 
with 48V BISG) technologies have been added and included in the 
Autonomie modeling and simulations, and advanced DEAC technology has 
been added using fixed point effectiveness estimates in the CAFE model 
analysis. The agencies disagree with ICCT's assessment of HCCI and do 
not believe it will be available for wide-scale application in the 
rulemaking timeframe, and therefore have not included it as a 
technology. HCCI technology has been in the research phase for several 
decades, and the only production applications to date use a highly-
limited version that restricts HCCI combustion to a very narrow range 
of engine operating conditions.459 460 461 Additional 
discussion of how Autonomie-modeled and non-modeled technologies are 
incorporated into the CAFE Model is located in Section VI.B.3.c), 
below.
---------------------------------------------------------------------------

    \459\ Mazda introduced Skyactiv-X in Europe with a mild hybrid 
technology to assist the engine.
    \460\ Mazda News. ``Revolutionary Mazda Skyactiv-X engine 
details confirmed as sales start,'' May 6, 2019. https://www.mazda-press.com/eu/news/2019/revolutionary-mazda-skyactiv-x-engine-details-confirmed-as-sales-start/. Last accessed Dec. 2, 2019.
    \461\ Confer. K. Kirwan, J. ``Ultra Efficient Light-Duty 
Powertrain with Gasoline Low-Temperature Combustion.'' DOE Merit 
Review. June 9, 2017. https://www.energy.gov/sites/prod/files/2017/06/f34/acs094_confer_2017_o.pdf. Last accessed Dec. 2, 2019.
---------------------------------------------------------------------------

    ICCT and Meszler also commented that the agencies overly limited 
the availability of several technologies in the NPRM analysis. In 
response, the agencies reconsidered the restrictions that were applied 
in the NPRM analysis, and agree with the commenters for several 
technologies and technology classes. Many technologies identified by 
the commenters are now in production for the MY2017 as well as MY2018 
and MY2019. The agencies also think that the baseline fleet compliance 
data reflects adoption of many of these technologies. For the final 
rule analysis, the agencies have expanded the availability of several 
technologies. In the CAFE model, the agencies are now allowing parallel 
hybrids (SHEVP2) to be adopted with high compression Atkinson mode 
engines (HCR0 and HCR1). In addition, as mentioned above, the Autonomie 
full-vehicle modeling included Variable Compression Ratio engine (VCR), 
Miller Cycle Engine (VTG), E-boost (VTGe) technologies, and cylinder 
deactivation technologies (DEAC) to be applied to turbocharged engines 
(TURBO1). As these changes relate to the technology effectiveness 
modeling, the CAFE model analysis now includes effectiveness estimates 
based on full vehicle simulations for all of these technology 
combinations.
    We disagree with comments stating the agencies should allow every 
technology to be available to every vehicle class.\462\ Discussed 
earlier in this section, Autonomie models key aspects of vehicle 
operation that are most relevant to assessing fuel economy, vehicle 
performance and certain aspects of drivability (like EPA 2-cycle tests, 
EPA US06 cycle tests, gradability, low speed acceleration time from 0-
to-60 mph, passing acceleration time from 50 to 80 mph, and number of 
transmission shifts). However, there are other critical aspects of 
vehicle functionality and operation that the agencies considered beyond 
those criteria, that cannot necessarily be reflected in the Autonomie 
modeling. For example, a pickup truck can be modeled with a

[[Page 24324]]

continuously variable transmission (CVT) and show improvements on the 
2-cycle tests. However, pickup trucks are designed to provide high load 
towing utility.\463\ CVTs lack the torque levels needed to provide that 
towing utility, and would fail mechanically if subject to high load 
towing.\464\ The agencies provided discussions of some of these 
technical considerations in the PRIA, and explained why the agencies 
had limited technologies for certain vehicle classes, such as limiting 
CVTs on pickups as in the example above. These and other limitations 
are discussed further in the individual technology sections.
---------------------------------------------------------------------------

    \462\ NHTSA-2018-0067-11723. NRDC Attachment2 at p. 4.
    \463\ SAE J2807. ``Performance Requirements for Determining Tow-
Vehicle Gross Combination Weight Rating and Trailer Weight Rating.'' 
Feb. 4, 2016.
    \464\ PRIA at p. 223 and 340.
---------------------------------------------------------------------------

    The agencies also received a variety of comments that conflated 
aspects of the Autonomie models with technology inputs and input 
assumptions. For example, commenters expressed concern about the 
transmission gear set and final drive values used for the NPRM 
analysis, or more specifically, that the gear ratios were held constant 
across applications.\465\ In this case, both the inputs (gear set and 
final drive ratio) and input assumption (ratios held constant) were 
discussed by the commenters. Because these comments are actually about 
technology inputs to the Autonomie model, for these and similar cases, 
the agencies are addressing the comments in the individual technology 
sections which discuss the technology inputs and input assumptions that 
impact the effectiveness values for those technologies.
---------------------------------------------------------------------------

    \465\ NHTSA-2018-0067-11873. Comments from Roush Industries, 
Attachment 1, at p. 14-15. NHTSA-2018-0067-11873. Comments from 
CARB, at p.110.
---------------------------------------------------------------------------

    For the NPRM analysis, the agencies prioritized using inputs that 
were based on data for identifiable technology configurations and that 
reflected practical real world constraints. The agencies provided 
detailed information on the NPRM analysis inputs and input assumptions 
in the NPRM Preamble, PRIA and Argonne model documentation for engine 
technologies, transmission technologies, powertrain electrification, 
light-weighting, aerodynamic improvements, tire rolling resistance 
improvements, and other vehicle technologies. Comments and the 
agencies' assessment of comments for each technology are discussed in 
the individual technology sections below. Through careful consideration 
of the comments, the agencies have updated analytical inputs associated 
with several technologies, and as discussed above, have included 
several advanced technologies for which technical information was 
included in the NPRM. However, for most technologies, the agencies have 
determined that the technology inputs and input assumptions that were 
used in the NPRM analysis remain reasonable and the best available for 
the final rule analysis.
(2) How The Agencies Defined Different Vehicle Types in Autonomie
    As described in the NPRM, Argonne produced full-vehicle models and 
ran simulations for many combinations of technologies, on many types of 
vehicles, but it did not simulate literally every single vehicle model/
configuration in the analysis fleet because it would be impractical to 
assemble the requisite detailed information--much of which would likely 
only be provided on a confidential basis--specific to each vehicle 
model/configuration and because the scale of the simulation effort 
would correspondingly increase by orders of magnitude. Instead, Argonne 
simulated 10 different vehicle types, corresponding to the five 
``technology classes'' generally used in CAFE analysis over the past 
several rulemakings, each with two performance levels and corresponding 
vehicle technical specifications (e.g., small car, small performance 
car, pickup truck, performance pickup truck, etc.).
    Technology classes are a means of specifying common technology 
input assumptions for vehicles that share similar characteristics. 
Because each vehicle technology class has unique characteristics, the 
effectiveness of technologies and combinations of technologies is 
different for each technology class. Conducting Autonomie simulations 
uniquely for each technology class provides a specific set of 
simulations and effectiveness data for each technology class. Like the 
Draft TAR analysis, there are separate technology classes for compact 
cars, midsize cars, small SUVs, large SUVs, and pickup trucks. However, 
new for the NPRM analysis and carried into this final rule analysis, 
each of those vehicle types has been split into ``low'' (or 
``standard'') performance and a ``high'' performance versions, which 
represent two classes with similar body styles but different levels of 
performance attributes (for a total of 10 technology classes). The 
separate technology classes for high performance and low performance 
vehicles better account for performance diversity across the fleet.
    NHTSA directed Argonne to develop a vehicle assumptions database to 
capture vehicle attributes that would comprise the full vehicle models. 
For each vehicle technology class, representative vehicle attributes 
and characteristics were identified from publicly available information 
and automotive benchmarking databases like A2Mac1,\466\ Argonne's 
Downloadable Dynamometer Database (D\3\),\467\ and EPA compliance and 
fuel economy data,\468\ EPA's guidance on the cold start penalty on 2-
cycle tests.\469\ The resulting vehicle assumptions database consists 
of over 100 different attributes like vehicle frontal area, drag 
coefficient, fuel tank weight, transmission housing weight, 
transmission clutch weight, hybrid vehicle component weights, and 
weights for components that comprise engines and electric machines, 
tire rolling resistance, transmission gear ratios and final drive 
ratio. Each of the 10 different vehicle types was assigned a set of 
these baseline attributes and characteristics, to which combinations of 
fuel-saving technologies were added as inputs for the Autonomie 
simulations. For example, the characteristics of the MY 2016 Honda Fit 
were considered along with a wide range of other compact cars to 
identify representative characteristics for the Autonomie simulations 
for the base compact car technology class. The simulations determined 
the fuel economy achieved when applying each combination of 
technologies to that vehicle type, given its baseline characteristics.
---------------------------------------------------------------------------

    \466\ A2Mac1: Automotive Benchmarking. (Proprietary data). 
Retrieved from https://a2mac1.com.
    \467\ Downloadable Dynamometer Database (D\3\). ANL Energy 
Systems Division. https://www.anl.gov/es/downloadable-dynamometer-database. Last accessed Oct. 31, 2019.
    \468\ Data on Cars used for Testing Fuel Economy. EPA Compliance 
and Fuel Economy Data. https://www.epa.gov/compliance-and-fuel-economy-data/data-cars-used-testing-fuel-economy. Last accessed Oct. 
31, 2019.
    \469\ EPA PD TSD at p.2-265--2-266.
---------------------------------------------------------------------------

    For each vehicle technology class and for each vehicle attribute, 
Argonne estimated the attribute value using statistical distribution 
analysis of publicly available data and data obtained from the A2Mac1 
benchmarking database.\470\ Some

[[Page 24325]]

vehicle attributes were also based on test data and vehicle 
benchmarking, like the cold-start penalty for the FTP test cycle and 
vehicle electrical accessories load. The analysis of vehicle attributes 
used in the NPRM was discussed in the Argonne model documentation,\471\ 
and values for each vehicle technology class were provided with the 
NPRM for public review.\472\
---------------------------------------------------------------------------

    \470\ A2Mac1 is subscription-based benchmarking service that 
conducts vehicle and component teardown analyses. Annually, A2Mac1 
removes individual components from production vehicles such as oil 
pans, electric machines, engines, transmissions, among the many 
other components. These components are weighed and documented for 
key specifications which is then available to their subscribers.
    \471\ NHTSA-2018-0067-0007, at 131. Islam, E., S, Moawad, A., 
Kim, N, Rousseau, A., ``A Detailed Vehicle Simulation Process To 
Support CAFE Standards 04262018--Report'' ANL Autonomie 
Documentation. Aug 21, 2018.
    \472\ NHTSA-2018-0067-0003. ANL Autonomie Summary of Main 
Component Assumptions. Aug 21, 2018.
---------------------------------------------------------------------------

    The agencies did not believe it was appropriate to assign one 
single engine mass for each vehicle technology class in the NPRM 
analysis. To account for the difference in weight for different engine 
types, Argonne performed a regression analysis of engine peak power 
versus weight, based on attribute data taken from the A2Mac1 
benchmarking database. For example, to account for weight of different 
engine sizes like 4-cylinder versus 8-cylinder, Argonne developed a 
relationship curve between peak power and engine weight based on the 
A2Mac1 benchmarking data. For the NPRM analysis, this relationship was 
used to estimate mass for all engine types regardless of technology 
type (e.g., variable valve lift and direct injection). Secondary weight 
reduction associated with changes in engine technology was applied by 
using this linear relationship between engine power and engine weight 
from the A2Mac1 benchmarking database. When a vehicle in the analysis 
fleet with an 8-cylinder engine adopted a more fuel efficient 6-
cylinder engine, the total vehicle weight would reflect the updated 
engine weight with two less cylinders based on the peak power versus 
engine weight relationship. The impact of engine mass reduction on 
effectiveness is accounted for directly in the Autonomie simulation 
data through the application of the above relationship. Engine mass 
reduction through downsizing is, therefore, appropriately not included 
as part of vehicle mass reduction technology that is discussed in 
Section VI.C.4 because doing so would result in double counting the 
impacts. As discussed further below, for the final rule the agencies 
improved upon the precision of engine weights by creating two curves to 
separately represent naturally aspirated engine designs and 
turbocharged engine designs.
    In addition, certain attributes were held at constant levels within 
each technology class to maintain vehicle functionality, performance 
and utility including noise, vibration, and harshness (NVH), safety, 
performance and other utilities important for customer satisfaction. 
For example, in addition to the vehicle performance constraints 
discussed in Section VI.B.3.a)(6), the analysis does not allow the 
frontal area of the vehicle to change, in order to maintain utility 
like ground clearance, head-room space, and cargo space, and a cold-
start penalty is used to account for fuel economy degradation for 
heater performance and emissions system catalyst light-off.\473\ This 
allows us to capture the discrete improvement in technology 
effectiveness while maintaining vehicle attributes that are important 
vehicle utility, consumer acceptance and compliance with criteria 
emission standards, and considering these constraints similar to how 
manufacturers do in the real world.
---------------------------------------------------------------------------

    \473\ The catalyst light-off is the temperature necessary to 
initiate the catalytic reaction and this energy is generated from 
engine.
---------------------------------------------------------------------------

    The agencies sought comment on the analytical approach used to 
determine vehicle attributes and characteristics for the Autonomie 
modeling. In response, the agencies received a wide variety of comments 
on vehicle attributes ranging from discussions of performance increase 
from technology adoption (e.g., if a vehicle adopting an electrified 
powertrain improved its time to accelerate from 0-60 mph), to comments 
on vehicle attributes not modeled in Autonomie, like heated seats and 
cargo space.
    Toyota and the Alliance commented that the inclusion of performance 
vehicle classes addressed the market reality that some consumers will 
purchase vehicles for their performance attributes and will accept the 
corresponding reduction in fuel economy. Furthermore, Toyota commented 
that some gain in performance is more realistic, and that ``dedicating 
all powertrain improvements to fuel efficiency is inconsistent with 
market reality.'' Toyota ``supports the agencies' inclusion of 
performance classes in compliance modeling where a subset of certain 
models is defined to have higher performance and a commensurate 
reduction in fuel efficiency.'' \474\ Also, in support of the addition 
of performance vehicle classes, the Alliance commented that ``vehicle 
categories have been increased to 10 to better recognize the range of 
0-60 performance characteristics within each of the 5 previous 
categories, in recognition of the fact that many vehicles in the 
baseline fleet significantly exceeded the previously assumed 0-60 
performance metrics. This provides better resolution of the baseline 
fleet and more accurate estimates of the benefits of technology.'' 
\475\
---------------------------------------------------------------------------

    \474\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at 
p. 6.
    \475\ Alliance of Automobile Manufacturers, Attachment ``Full 
Comment Set,'' Docket No. NHTSA-2018-0067-12073, at p.135.
---------------------------------------------------------------------------

    UCS commented that the CAFE model incorporates technology 
improvements to each vehicle by applying the effectiveness improvement 
of the average vehicle in the technology class, leading to discrete 
``stepped'' effectiveness levels for technologies across the different 
vehicle types. UCS stated that in contrast, the OMEGA model takes into 
account a vehicle's performance characteristics through response-
surface modeling based on relative deviation from the class average 
modeled in ALPHA.\476\
---------------------------------------------------------------------------

    \476\ NHTSA-2018-0067-12039, at p.24.
---------------------------------------------------------------------------

    Although differences between the ALPHA and Autonomie models are 
discussed in more detail below, for the NPRM vehicle simulation 
analysis the agencies expanded the number of vehicle classes from the 
five classes used in the Draft TAR to ten classes, to represent better 
the diversity of vehicle characteristics across the fleet. Each of 
these ten vehicle technology classes are empirically built from 
benchmarking data and other information from various sources, amounting 
to hundreds of vehicle characteristics data points to develop each 
vehicle class. The agencies expand on these vehicle classes and 
characteristics in Section VI.B.3.(a)(2) Vehicle Types in Autonomie and 
Section VI.B.3.(a)(3) How Vehicle Models are Built in Autonomie and 
Optimized for Simulation. The agencies believe that the real-world data 
used to define vehicle characteristics for each of the ten vehicle 
classes, in addition to the ten vehicle technology classes themselves, 
ensures the analysis reasonably accounts for the diversity in vehicle 
characteristics across the fleet.
    The agencies believe that UCS's characterization of how technology 
improvements are applied in the analysis is a misleading 
oversimplification. While the analysis approach in the final rule uses 
a representative effectiveness value, the value is not linked solely to 
the vehicle technology class, as the UCS implies. The entire technology 
combination, or technology key, which includes the vehicle technology 
class, is used to

[[Page 24326]]

determine the value for the platform being considered. Within each 
vehicle class, the interactions between the added technology and the 
full vehicle system (including other technologies and substantial road 
load characteristics) are considered in the effectiveness values 
calculated for each technology during compliance modeling. As discussed 
under each of the technology pathways sections, the effectiveness for 
most technologies is reported as a range rather than a single value. 
The range exists because the effectiveness for each technology is 
adjusted based on the technologies it is coupled with and the major 
road load characteristics of the full vehicle system. This approach, in 
combination with using the baseline vehicle's initial performance 
values as a starting point for performance improvement, results in a 
widely variable level of improvement for the system, dependent on 
individual vehicle platform characteristics. As a result, the 
application of a response-surface approach would likely result in 
minimal improvement in accuracy for the Autonomie and CAFE model 
analysis approach.
    For the final rule analysis, the agencies used the same process to 
obtain the vehicle attributes and characteristics for the vehicle 
technology classes. Data was acquired from publicly available sources, 
Argonne D\3\, EPA compliance and fuel economy data, and A2mac1 
benchmarking data. Accordingly, the attributes and characteristics of 
the modeled vehicles reflect actual vehicles that meet customer 
expectations and automakers' capabilities to manufacture the vehicles. 
In addition, for the final rule, the agencies improved the NPRM 
analysis by updating some of the attribute values to account for 
changes in the fleet. For example, the agencies have updated vehicle 
electrical accessory load on the test cycle to reflect higher 
electrical loads associated with contemporary vehicle features.
(3) How This Rulemaking Builds Vehicle Models for Autonomie and 
Optimize Them for Simulation
    Before any simulation is initiated in Autonomie, Argonne must 
``build'' a vehicle by assigning reference technologies and initial 
attributes to the components of the vehicle model representing each 
technology class.\477\ The reference technologies are baseline 
technologies that represent the first step on each technology pathway 
used in the analysis. For example, a compact car is built by assigning 
it a baseline engine, a baseline 6-speed automatic transmission (AT6), 
a baseline level of aerodynamic improvement (AERO0), a baseline level 
of rolling resistance improvement (ROLL0), a baseline level of mass 
reduction technology (MR0), and corresponding attributes from the 
Argonne vehicle assumptions database like individual component 
weights.\478\ A baseline vehicle will have a unique starting point for 
the simulation and a unique set of assigned inputs and attributes, 
based on its technology class.
---------------------------------------------------------------------------

    \477\ For the NPRM analysis, Chapter 8 Vehicle-Sizing Process in 
the ANL Model Documentation had discussed this process in detail. 
Further discussion of this process is located in Chapter 8 of the 
ANL Model Documentation for this final rule.
    \478\ See Section VI.A.7.
---------------------------------------------------------------------------

    The next step in the process is to run a powertrain sizing 
algorithm that ensures the built vehicle meets or exceeds defined 
performance metrics, including low-speed acceleration (i.e., time 
required to accelerate from 0-60 mph), high-speed passing acceleration 
(time required to accelerate from 50-80 mph), gradeability (e.g. the 
ability of the vehicle to maintain constant 65 miles per hour speed on 
a six percent upgrade), and towing capacity. Together, these 
performance criteria are widely used by industry as metrics to quantify 
vehicle performance attributes that consumers observe and that are 
important for vehicle utility and customer satisfaction.
    In the compact car example used above, the agencies assigned an 
initial specific engine design and engine power, transmission, AERO, 
ROLL, and MR technologies, and other attributes like vehicle weight. If 
the built vehicle does not meet all the performance criteria in the 
first iteration, then the engine power is increased to meet the 
performance requirement. This increase in power is from higher engine 
displacement, which could involve an increase in number of cylinders, 
leading to an increase in the engine weight. The iterative process 
continues to check whether the compact car with updated engine power, 
and corresponding updated engine weight, meets its defined performance 
metrics. The loop stops once all the metrics are met, and at this 
point, a compact car technology class vehicle model becomes ready for 
simulation. For further discussion of the vehicle performance metrics, 
see Section VI.B.3.(a).
    Autonomie then adopts a single fuel saving technology to the 
baseline vehicle model, keeping everything else the same except for 
that one technology and the attributes associated with it. For example, 
the model would apply an 8-speed automatic transmission in place of the 
baseline 6-speed automatic transmission, which would lead to either an 
increase or decrease in the total weight of the vehicle based on the 
technology class assumptions. At this point, Autonomie confirms whether 
performance metrics are met for this new vehicle model through the 
previously discussed sizing algorithm. Once a technology has been 
assigned to the vehicle model and the resulting vehicle meets its 
performance metrics, those vehicle models will be used as inputs to the 
full vehicle simulations. So, in the example of the 6-speed to 8-speed 
automatic transmission technology update, the agencies now have the 
initial ten vehicle models (one for each technology class), plus the 
ten new vehicle models with the updated 8-speed automatic transmission, 
which adds up to 20 different vehicle models for simulation. This 
permutation process is conducted for each of the over 50 technologies 
considered, and for all ten technology classes, which results in more 
than one million optimized vehicle models.
    Figure VI-3 shows the process for building vehicles in Autonomie 
for simulation.

[[Page 24327]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.110

    Some of the technologies require extra steps for optimization 
before the vehicle models are built for simulation; for example, the 
sizing and optimization process is more complex for the electrified 
vehicles (i.e., HEVs, PHEVs) compared to vehicles with internal 
combustion engines, as discussed further, below. Throughout the vehicle 
building process, the following items are considered for optimization:
     Vehicle weight is decreased or increased in response to 
switching from one type of technology to another for the technologies 
for which the agencies consider weight, such as different engine and 
transmission types;
     Vehicle performance is decreased or increased in response 
to the addition of mass reduction technologies when switching from one 
vehicle model to another vehicle model for the same engine;
     Vehicle performance is decreased or increased in response 
to the addition of a new technology when switching from one vehicle 
model to another vehicle model for the same hybrid electric machine; 
and
     Electric vehicle battery size is decreased or increased in 
response to the addition of mass, aero and/or tire rolling resistance 
technologies when switching from one vehicle model to another vehicle 
model.
    Every time a vehicle adopts a new technology, the vehicle weight is 
updated to reflect the new component weight. For some technologies, the 
direct weight change is easy to assess. For example, in the NPRM the 
agencies designated weights for transmissions so, when a vehicle is 
updated to a higher geared transmission, the weight of the original 
transmission is replaced with the corresponding transmission weight 
(e.g., the weight of a vehicle moving from a 5-speed automatic 
transmission to an 8-speed automatic transmission will be updated based 
on the 8-speed transmission weight).
    For other technologies, like engine technologies, assessing the 
updated vehicle weight is much more complex. Discussed earlier, 
modeling a change in engine technology involves both the new technology 
adoption and a change in power (because the reduction in vehicle weight 
leads to lower engine loads, and a resized engine). When a new engine 
technology is adopted on a vehicle the agencies account for the 
associated weight change to the vehicle based on the earlier discussed 
regression analysis of weight versus power. For the NPRM engine weight 
regression analysis, the agencies considered 19 different engine 
technologies that consisted of unique components to achieve fuel 
economy improvements. This regression analysis is technology agnostic 
by taking the approach of using engine peak power versus engine weight 
because it removed biases to any specific engine technology in the 
analysis. Although the agencies do not estimate the specific weight for 
each individual engine technology, such as VVT and SGDI, this process 
provides a reasonable estimate of the weight differences among engine 
technologies.

[[Page 24328]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.111

    For the final rule analysis, the agencies used the same process to 
assign initial weights to the original 19 engines, plus the added 
engines. However, the agencies improved upon precision of the weights 
by creating two separate curves separately to represent naturally 
aspirated engine designs and turbocharged engine designs.\479\ This 
update resulted in two benefits. First, small naturally aspirated 4-
cylinder engines that adopted turbocharging technology reflected the 
increased weight of associated components like ducting, clamps, the 
turbocharger itself, a charged air cooler, wiring, fasteners, and a 
modified exhaust manifold. Second, larger cylinder count engines like 
naturally aspirated 8-cylinder and 6-cylinder engines that adopted 
turbocharging and downsized technologies would have lower weight due to 
having fewer engine cylinders. For example, a naturally aspirated 8-
cylinder engine that adopts turbocharging technology when downsized to 
a 6-cylinder turbocharged engine appropriately reflects the added 
weight of turbocharging components, and the lower weight of fewer 
cylinders.
---------------------------------------------------------------------------

    \479\ ANL Model Documentation for the final rule analysis, 
Chapter 5.2.9 Engine Weight Determination.
---------------------------------------------------------------------------

    As with conventional vehicle models, electrified vehicle models 
were built from the ground up. For the NPRM analysis, Argonne used data 
from the A2mac1 database and vehicle test data to define different 
attributes like weights and power. Argonne used one electric motor 
specific power for each type of hybrid and electric vehicle.\480\ For 
MY2017, the U.S. market has an expanded number of available hybrid and 
electric vehicle models. To capture appropriately the improvements for 
electrified vehicles for the final rule analysis, the agencies applied 
the same regression analysis process that considers electric motor 
weight versus electric motor power for vehicle models that have adopted 
electric motors. Benchmarking data for hybrid and electric vehicles 
from the A2Mac1 database was analyzed to develop a regression curve of 
electric motor peak power versus electric motor weight.\481\
---------------------------------------------------------------------------

    \480\ NHTSA-2018-0067-0005. ANL Autonomie Model Assumptions 
Summary. Aug 21, 2018. Non_Vehicle_Attributes tab. Specific power 
for PS and P2 HEVs was set to 2750 watts/kg, plug-in HEVs were set 
to 375 watts/kg, and electric vehicles were set to 1400 watts/kg.
    \481\ ANL Model Documentation for the final rule analysis, 
Chapter 5.2.10 Electric Machines System Weight.
---------------------------------------------------------------------------

(4) How Autonomie Sizes Powertrains for Full Vehicle Simulation
    The agencies maintain performance neutrality of the full vehicle 
simulation analysis by resizing engines, electric machines, and hybrid 
electric vehicle battery packs at specific incremental technology 
steps. To address product complexity and economies of scale, engine 
resizing is limited to specific incremental technology changes that 
would typically be associated with a major vehicle or engine 
redesign.\482\ Manufacturers have repeatedly told the agencies that the 
high costs for redesign and the increased manufacturing complexity that 
would result from resizing engines for small technology changes 
preclude them from doing so. It would be unreasonable and unaffordable 
to resize powertrains for every unique combination of technologies, and 
exceedingly so for every unique combination of technologies across 
every vehicle model due to the extreme manufacturing complexity that 
would be required to do so. The agencies reiterated in the NPRM that 
the analysis should not include engine resizing with the application of 
every technology or for combinations of technologies that drive small 
performance changes so that the analysis better reflects what is 
feasible for manufacturers.\483\
---------------------------------------------------------------------------

    \482\ See 83 FR 43027 (Aug. 24, 2018).
    \483\ For instance, a vehicle would not get a modestly bigger 
engine if the vehicle comes with floor mats, nor would the vehicle 
get a modestly smaller engine without floor mats. This example 
demonstrates small levels of mass reduction. If manufacturers 
resized engines for small changes, manufacturers would have 
dramatically more part complexity, potentially losing economies of 
scale.
---------------------------------------------------------------------------

    When a powertrain does need to be resized, Autonomie attempts to 
mimic manufacturers' development approaches to the extent possible. 
Discussed earlier, the Autonomie vehicle building process is initiated 
by building a baseline vehicle model with a baseline engine, 
transmission, and other baseline vehicle technologies. This baseline 
vehicle model (for each technology class) is sized to meet a specific 
set of

[[Page 24329]]

performance criteria, including acceleration and gradeability.
    The modeling also accounts for the industry practice of platform, 
engine, and transmission sharing to manage component complexity and the 
associated costs.\484\ At a vehicle refresh cycle, a vehicle may 
inherit an already resized powertrain from another vehicle within the 
same engine-sharing platform that adopted the powertrain in an earlier 
model year. In the Autonomie modeling, when a new vehicle adopts fuel 
saving technologies that are inherited, the engine is not resized (the 
properties from the baseline reference vehicle are used directly and 
unchanged) and there may be a small change in vehicle performance. For 
example, in Figure VI-3, Vehicle 2 inherits Eng01 from Vehicle 1 while 
updating the transmission. Inheritance of the engine with new 
transmission may change performance. This example illustrates how 
manufacturers generally manage manufacturing complexity for engines, 
transmissions, and electrification technologies.
---------------------------------------------------------------------------

    \484\ Ford EcoBoost Engines are shared across ten different 
models in MY2019. https://www.ford.com/powertrains/ecoboost/. Last 
accessed Nov. 05, 2019.
---------------------------------------------------------------------------

    Autonomie implements different powertrain sizing algorithms 
depending on the type of powertrain being considered because different 
types of powertrains contain different components that must be 
optimized.\485\ For example, the conventional powertrain resizing 
considers the reference power of the conventional engine (e.g., Eng01, 
a basic VVT engine, is rated at 108 kilowatts and this is the starting 
reference power for all technology classes) against the power-split 
hybrid (SHEVPS) resizing algorithm that must separately optimize engine 
power, battery size (energy and power), and electric motor power. An 
engine's reference power rating can either increase or decrease 
depending on the architecture, vehicle technology class, and whether it 
includes other advanced technologies.
---------------------------------------------------------------------------

    \485\ ANL Model Documentation for the final rule Analysis, 
Chapter 8.3.1 Conventional-Vehicle Sizing Algorithm; Chapter 8.3.2 
Split-HEV Sizing Algorithm; 8.3.4 Blended PHEV sizing Algorithm; 
8.3.5 Voltec PHEV (Extended Range) Vehicle Sizing Algorithm; Chapter 
8.3.6 BEV Sizing Algorithm.
---------------------------------------------------------------------------

    Performance requirements also differ depending on the type of 
powertrain because vehicles with different powertrain types may need to 
meet different criteria. For example, a plug-in hybrid electric vehicle 
(PHEV) powertrain that is capable of traveling a certain number of 
miles on its battery energy alone (referred to as all-electric range, 
or AER, or as performing in electric-only mode) is also sized to ensure 
that it can meet the performance requirements of a US06 cycle in 
electric-only mode.
    The powertrain sizing algorithm is an iterative process that 
attempts to optimize individual powertrain components at each step. For 
example, the sizing algorithm for conventional powertrains estimates 
required power to meet gradeability and acceleration performance and 
compares it to the reference engine power for the technology class. If 
the power required to meet gradeability and acceleration performance 
exceeds the reference engine power, the engine power is updated to the 
new value. Similarly, if the reference engine power exceeds the 
gradeability and acceleration performance power, it will be decreased 
to the lower power rating. As the change in power requires a change 
design of the engine, like increasing displacement (e.g., going from a 
5.2-liter to 5.6-liter engine, or vice versa) or increasing cylinder 
count (e.g., going from an I4 to a V6 or vice versa), the engine weight 
will also change. The new engine power is used to update the weight of 
the engine.
    Next, the conventional powertrain sizing algorithm enters an 
acceleration algorithm loop to verify low-speed acceleration 
performance (time it takes to go from 0 mph to 60 mph). In this step, 
Autonomie adjusts engine power to maintain a performance attribute for 
the given technology class and updates engine weight accordingly. Once 
the performance criteria are met, Autonomie ends the low-speed 
acceleration performance algorithm loop and enters a high-speed 
acceleration (time it takes to go from 50 mph to 80 mph) algorithm 
loop. Again, Autonomie might need to adjust engine power to maintain a 
performance attribute for the given technology, and it exits this loop 
once the performance criteria have been met. At this point, the sizing 
algorithm is complete for the conventional powertrain based on the 
designation for engine type, transmissions type, aero type, mass 
reduction technology and low rolling resistance technology.
    Figure VI-5 below shows the sizing algorithm for conventional 
powertrains.

[[Page 24330]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.112

    Depending on the type of powertrain considered, the sizing 
algorithms may also size to meet different performance criteria in 
different order. The powertrain sizing algorithms for electrified 
vehicles are considerably more complex, and are discussed in further 
detail in Section VI.C.3, below.
(5) How the Agencies Considered Maintaining Vehicle Attributes
    For this rulemaking analysis, consistent with past CAFE and 
CO2 rulemakings, the agencies have analyzed technology 
pathways manufacturers could use for compliance that attempt to 
maintain vehicle attributes, utility, and performance. Using this 
approach allows the agencies to assess costs and benefits of potential 
standards under a scenario where consumers continue to get the similar 
vehicle attributes and features, other than changes in fuel economy. 
The purpose of constraining vehicle attributes is to simplify the 
analysis and reduce variance in other attributes that consumers value 
across the analyzed regulatory alternatives. This allows for a more 
streamlined accounting of costs and benefits by not requiring the 
values of other vehicle attributes that trade off with fuel economy.
    Several examples of vehicle attributes, utility and performance 
that could be impacted by adoption of fuel economy improving technology 
include the following.

[[Page 24331]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.113

    Consequences for the agencies not fully considering or accounting 
for potential changes in vehicle attributes, utility, and performance 
are degradation in vehicle attributes, utility, and performance that 
lead to consumer acceptance issues without accounting for the 
corresponding costs and/or not accounting for the costs of technology 
designs that maintain vehicle attributes, utility, and performance. The 
agencies incorporated changes in the NPRM analysis and that are carried 
into this final rule that address deficiencies in past analyses, 
including the Draft TAR and Proposed Determination analyses. These 
changes were discussed in the NPRM and are repeated in the discussion 
of individual technologies in this Preamble, the FRIA, and supporting 
documents. The following are several examples of technologies that did 
not maintain vehicle attributes, utility, and performance in the Draft 
TAR and Proposed Determination analyses.
    For the EPA Draft TAR and Proposed Determination analyses, HCR 
engine and downsized and turbocharged engine technologies effectiveness 
was estimated using Tier 2 certification fuel, which has a higher 
octane rating compared to regular octane fuel.486 487 This 
does not maintain functionality because consumers would incur higher 
costs for using premium fuel in order to achieve the modeled fuel 
economy improvements, compared to baseline engines that were replaced, 
which operated on lower cost regular octane fuel. By not maintaining 
the fuel octane functionality and vehicle attributes, the EPA Draft TAR 
and Proposed Determination analyses applied higher effectiveness for 
these technologies than could be achieved had regular octane fuel been 
assumed for the HCR and downsized turbocharged engines. The Draft TAR 
and Proposed Determination analyses also did not account for the higher 
costs that would be incurred by consumers to pay for high octane fuel. 
These issues were addressed in the

[[Page 24332]]

NPRM and this final rule analysis, and account for some of the 
effectiveness and cost differences between the Draft TAR/Proposed 
Determination and the NPRM/final rule.\488\
---------------------------------------------------------------------------

    \486\ Tier 2 fuel has an octane rating of 93. Typical regular 
grade fuel has an octane rating of 87 ((R+M)/2 octane.
    \487\ EPA Proposed Determination at 2-209 to 2-212.
    \488\ For more details, see Section VI.C.1 Engine Paths.
---------------------------------------------------------------------------

    Another example is mass reduction technology. As background, the 
agencies characterize mass reduction as either primary mass reduction 
or secondary mass reduction. Primary mass reduction involves reducing 
mass of components that can be done independently of the mass of other 
components. For example, the mass of a hood (e.g., replacing a steel 
hood with an aluminum hood) or reducing the mass of a seat are examples 
of primary mass reduction because each can be implemented 
independently. When there is a significant level of primary mass 
reduction, other components that are designed based on the mass of 
primary components, may be redesigned and have lower mass. An example 
of secondary mass reduction is the brake system. If the mass of primary 
components is reduced sufficiently, the resulting lighter weight 
vehicle could maintain braking performance and attributes, and safety 
with a lighter weight brake system. Mass reduction in the brake system 
is secondary mass reduction because it requires primary mass reduction 
before it can be incorporated. For the EPA Draft TAR and Proposed 
Determination analyses, secondary mass reduction was applied 
exclusively based on cost, with no regard to whether sufficient primary 
mass reduction was applied concurrently. The analyses did not account 
for the degraded functionality of the secondary components and systems 
and also understated the costs for lower levels of mass reduction.\489\ 
These issues were addressed in the NPRM and this final rule analysis, 
and account for some of the cost differences between the Draft TAR/
Proposed Determination and the NPRM/final rule.
---------------------------------------------------------------------------

    \489\ For more details, see Section VI.C.4 Mass Reduction.
---------------------------------------------------------------------------

    The agencies note that for some technologies it is not reasonable 
or practicable to match exactly the baseline vehicle's attributes, 
utility, and performance. For example, when engines are resized to 
maintain acceleration performance, if the agencies applied a criterion 
that allowed no shift in performance whatsoever, there would be an 
extreme proliferation of unique engine displacements. Manufacturers 
have repeatedly and consistently told the agencies that the high costs 
for redesign and the increased manufacturing complexity that would 
result from resizing engines for small technology changes preclude them 
from doing so. It would be unreasonable and unaffordable to resize 
powertrains for every unique combination of technologies, and 
exceedingly so for every unique combination technologies across every 
vehicle model due to the extreme manufacturing complexity that would be 
required to do so.\490\ For the NPRM and final rule analyses, engine 
resizing is limited to specific incremental technology changes that 
would typically be associated with a major vehicle or engine redesign 
to address product complexity and economies of scale considerations. 
The EPA Draft TAR and Proposed Determination analyses adjusted the 
effectiveness of every technology combination assuming performance 
could be held constant for every combination, and the analysis did not 
recognize or account for the extreme complexity nor the associated 
costs for that impractical assumption. The NPRM and final rule analyses 
account for these real-world practicalities and constraints, and doing 
so explains some of the effectiveness and cost differences between the 
Draft TAR/Proposed Determination and the NPRM/final rule.
---------------------------------------------------------------------------

    \490\ For more details, see Section VI.B.3.a)(6) Performance 
Neutrality.
---------------------------------------------------------------------------

    The subsections for individual technologies discuss the technology 
assumptions and constraints that were considered to maintain vehicle 
attributes, utility, and performance as closely as possible. The 
agencies believe that any minimal remaining differences, which may 
directionally either improve or degrade vehicle attributes, utility and 
performance are small enough to have de minimis impact on the analysis.
(6) How the Agencies Considered Performance Neutrality
    The CAFE model examines technologies that can improve fuel economy 
and reduce CO2 emissions. An improvement in efficiency can 
be realized by improving the powertrain that propels the vehicle (e.g., 
replacing a 6-cylinder engine with a smaller, turbocharged 4-cylinder 
engine), or by reducing the vehicle's loads or burdens (e.g., lowering 
aerodynamic drag, reducing vehicle mass and/or rolling resistance). 
Either way, these changes reduce energy consumption and create a range 
of choices for automobile manufacturers. At the two ends of the range, 
the manufacturer can choose either:
    (A) To design a vehicle that does same the amount of work as before 
but uses less fuel.
    For example, a redesigned pickup truck would receive a turbocharged 
V6 engine in place of the outgoing V8. The pickup would offer no 
additional towing capacity, acceleration, larger wheels and tires, 
expanded infotainment packages, or customer convenience features, but 
would achieve a higher fuel economy rating (and correspondingly lower 
CO2 emissions).
    (B) To design a vehicle that does more work and uses the same 
amount of fuel as before.
    For example, a redesigned pickup truck would receive a turbocharged 
V6 engine in place of the outgoing V8, but with engine efficiency 
improvements that allow the same amount of fuel to do more work. The 
pickup would offer improved towing capacity, improved acceleration, 
larger wheels and tires, an expanded (heavier) infotainment package, 
and more convenience features, while maintaining (not improving) the 
fuel economy rating of the previous year's model.
    In other words, automakers weigh the trade-offs between vehicle 
performance/utility and fuel economy, and they choose a blend of these 
attributes to balance meeting fuel economy and emissions standards and 
suiting the demands of their customers.
    Historically, vehicle performance has improved over the years. The 
average horsepower is the highest that it has ever been; all vehicle 
types have improved horsepower by at least 49 percent compared to the 
1975 model year, and pickup trucks have improved by 141 percent.\491\ 
Since 1978, the 0-60 acceleration time of vehicles has improved by 39-
47 percent depending on vehicle type.\492\ Also, to gain consumer 
acceptance of downsized turbocharged engines, manufacturers have stated 
they often offer an increase in performance.\493\ Fuel economy has also 
improved, but the horsepower and acceleration trends show that not 100 
percent of technological improvements have been applied to fuel 
savings. While future trends are uncertain, the past trends suggest 
vehicle performance is unlikely to decrease, as it seems reasonable to 
assume that customers

[[Page 24333]]

will at a minimum demand vehicles that offer the same utility as 
today's fleet.
---------------------------------------------------------------------------

    \491\ The 2018 EPA Automotive Trends Report (EPA-420-R-19-002 
March 2019) https://www.epa.gov/automotive-trends/download-automotive-trends-report.
    \492\ The 2018 EPA Automotive Trends Report (EPA-420-R-19-002 
March 2019) https://www.epa.gov/automotive-trends/download-automotive-trends-report.
    \493\ Alliance of Automobile Manufacturers, Attachment 
``Comment,'' Docket No. EPA-HQ-OAR-2015-0827-4089, at p. 122.
---------------------------------------------------------------------------

    For this rulemaking analysis, consistent with past CAFE and 
CO2 rulemakings, the agencies have analyzed technology 
pathways manufacturers could use for compliance that attempt to 
maintain vehicle attributes, utility and performance. NHTSA's analysis 
in the Draft TAR used the same approach for performance neutrality as 
was used for the NPRM and is being carried into this final rule. This 
approach is described throughout this section and further in FRIA 
Section VI. For the Draft TAR and Proposed Determination, the EPA 
analyses used an approach that maintained 0-60 mph acceleration time 
for every technology package. However, that approach did not account 
for the added development, manufacturing, assembly and service parts 
complexity and associated costs that would be incurred by manufacturers 
to produce the substantial number of engine variants that would be 
required to achieve those CO2 improvements.\494\ Using the 
NPRM approach, which is carried into this final rule, allows the 
agencies to assess costs and benefits of potential standards under a 
scenario where consumers continue to get the same vehicle attributes 
and features, other than changes in fuel economy (approaching the 
scenario in example ``A'' above). This approach also eliminates the 
need to assess the value of changes in vehicle attributes and features. 
As discussed later in this section, while some small level of 
performance increase is unavoidable when conducting this type of 
analysis, the added technology results almost exclusively in improved 
fuel economy. This allows the cost of these technologies to reflect 
almost entirely the cost of compliance with standards with nearly 
neutral vehicle performance.
---------------------------------------------------------------------------

    \494\ Each variant would require a unique engine displacement, 
requiring unique internal engine components, such as crankshaft, 
connecting rods and others.
---------------------------------------------------------------------------

    The CAFE model maintains the initial performance and utility levels 
of the analysis vehicle fleet, while considering real world constraints 
faced by manufacturers.
    To maintain performance neutrality when applying fuel economy 
technologies, it is first necessary to characterize the performance 
levels of each of the nearly 3000 vehicle models in the MY 2017 
baseline fleet. As discussed in Section VI.B.1.b) Assigning Vehicle 
Technology Classes, above, each individual vehicle model in the 
analysis fleet was assigned to one of ten vehicle ``technology 
classes''--the class that is most similar to the vehicle model. The 
technology classes include five standard class vehicles (compact car, 
midsize car, small SUV, midsize SUV, pickup) plus five ``performance'' 
versions of these same body styles.\495\ Each vehicle class has a 
unique set of attributes and characteristics, including vehicle 
performance metrics, that describe the typical characteristics of the 
vehicles in that class.
---------------------------------------------------------------------------

    \495\ Separate technology classes were created for high 
performance and low performance vehicles to better account for 
performance diversity across the fleet.
---------------------------------------------------------------------------

    The analysis used four criteria to characterize vehicle performance 
attributes and utility:

 Low-speed acceleration (time required to accelerate from 0-60 
mph)
 High-speed acceleration (time required to accelerate from 50-
80 mph)
 Gradeability (the ability of the vehicle to maintain constant 
65 miles per hour speed on a six percent upgrade)
 Towing capacity

    Low-speed and high-speed acceleration target times are typical of 
current production vehicles and range from 6 to 10 seconds depending on 
the vehicle class; for example, the midsize SUV performance class has a 
low- and high-speed acceleration target of 7 seconds.\496\ The 
gradeability criterion requires that the vehicle, given its attributes 
of weight, engine power, and transmission gearing, be capable of 
maintaining a minimum of 65 mph while going up a six percent grade. The 
towing criterion, which is applicable only to the pickup truck and 
performance pickup truck vehicle technology classes, is the same as the 
gradeability requirement but adds an additional payload/towing mass 
(3,000 lbs. for pickups, or 4,350 lbs for performance pickups) to the 
vehicle, essentially making the vehicle heavier.
---------------------------------------------------------------------------

    \496\ Note, for all vehicle classes, the low and high-speed 
acceleration targets use the same value. See section VI.B.1.b)(1) 
Assigning Vehicle Technology Classes for a list of low-speed 
acceleration target by vehicle technology class.
---------------------------------------------------------------------------

    In addition, to maintain the capabilities of certain electrified 
vehicles in the 2017 baseline fleet, the analysis required that those 
vehicles be capable of achieving the accelerations and speeds of 
certain standard driving cycles. The agencies use the US06 ``aggressive 
driving'' cycle and the UDDS ``city driving'' cycle to ensure that core 
capabilities of BEVs and PHEVs, such as driving certain speeds and/or 
distances in electric-only mode, are maintained. In addition to the 
four criteria discussed above, the following performance criteria are 
applied to these electrified vehicles:
     Battery electric vehicles (BEV) are sized to be capable of 
completing the US06 ``aggressive driving'' cycle.
     Plug-in hybrid vehicles with 50 mile all-electric range 
(PHEV50) are sized to be capable of completing the US06 ``aggressive 
driving'' cycle in electric-only mode.
     Plug-in hybrid vehicles with 20 mile all-electric range 
(PHEV20) are sized to be capable of completing the UDDS ``city 
driving'' cycle in electric-only (charge depleting) mode.\497\
---------------------------------------------------------------------------

    \497\ PHEV20's are blended-type plug-in hybrid vehicles, which 
are capable of completing the UDDS cycle in charge depleting mode 
without assistance from the engine. However, under higher loads, 
this charge depleting mode may use supplemental power from the 
engine.
---------------------------------------------------------------------------

    Together, these performance criteria are widely used by industry as 
metrics to quantify vehicle performance attributes that consumers 
observe and that are important for vehicle utility and customer 
satisfaction.\498\
---------------------------------------------------------------------------

    \498\ Conlon, B., Blohm, T., Harpster, M., Holmes, A. et al., 
``The Next Generation ``Voltec'' Extended Range EV Propulsion 
System,'' SAE Int. J. Alt. Power. 4(2):2015, doi:10.4271/2015-01-
1152. Kapadia, J., Kok, D., Jennings, M., Kuang, M., et al., 
``Powersplit or Parallel--Selecting the Right Hybrid Architecture,'' 
SAE Int. J. Alt. Power. 6(1):2017, doi:10.4271/2017-01-1154. Islam, 
E., A. Moawad, N. Kim, and A. Rousseau, 2018a, An Extensive Study on 
Vehicle Sizing, Energy Consumption and Cost of Advance Vehicle 
Technologies, Report No. ANL/ESD-17/17, Argonne National Laboratory, 
Lemont, Ill., Oct 2018.
---------------------------------------------------------------------------

    When certain fuel-saving technologies are applied that affect 
vehicle performance to a significant extent, such as replacing a pickup 
truck's V8 engine with a turbocharged V6 engine, iterative resizing of 
the vehicle powertrain (engine, electric motors, and/or battery) is 
performed in the Autonomie simulation such that the above performance 
criteria is maintained. For example, if the aforementioned engine 
replacement caused an improvement in acceleration, the engine may be 
iteratively resized until vehicle acceleration performance is shifted 
back to the initial target time for that vehicle technology class. For 
the low and high-speed acceleration criteria, engine resizing 
iterations continued until the acceleration time was within plus or 
minus 0.2 seconds of the target time,499 500 which is judged 
to balance

[[Page 24334]]

reasonably the precision of engine resizing with the number of 
simulation iterations needed to achieve performance within the 0.2 
second window, and the associated computer resources and time required 
to perform the iterative simulations. Engine resizing is explained 
further in Section VI.B.3.a)(4) How Autonomie Sizes Powertrains for 
Full Vehicle Simulation and the Argonne Model Documentation for the 
final rule analysis.
---------------------------------------------------------------------------

    \499\ For example, if a vehicle has a target 0-60 acceleration 
time of 6 seconds, a time within 5.8-6.2 seconds was accepted.
    \500\ With the exception of a few performance electrified 
vehicle types which, based on observations in the marketplace, use 
different criteria to maintain vehicle performance without battery 
assist. Performance PHEV20, and Performance PHEV50 resize to the 
performance of a conventional six-speed automatic (CONV 6AU). 
Performance SHEVP2, engines/electric-motors were resized if the 0-60 
acceleration time was worse than the target, but not resized if the 
acceleration time was better than the target time.
---------------------------------------------------------------------------

    The Autonomie simulation resizes until the least capable of the 
performance criteria is met, to ensure the pathways do not degrade any 
of the vehicle performance metrics. It is possible that as one 
criterion target is reached after the application of a specific 
technology or technology package, other criteria may be better than 
their target values. For example, if the engine size is decreased until 
the low speed acceleration target is just met, it is possible that the 
resulting engine size would cause high speed acceleration performance 
to be better than its target.\501\ Or, a PHEV50 may have an electric 
motor and battery appropriately sized to operate in all electric mode 
through the repeated accelerations and high speeds in the US06 driving 
cycle, but the resulting motor and battery size enables the PHEV50 
slightly to over-perform in 0-60 acceleration, which utilizes the power 
of both the electric motor and combustion engine.
---------------------------------------------------------------------------

    \501\ The Autonomie simulation databases include all of the 
estimated performance metrics for each combination of technology as 
modeled.
---------------------------------------------------------------------------

    To address product complexity and economies of scale, engine 
resizing is limited to specific incremental technology changes that 
would typically be associated with a major vehicle or engine 
redesign.\502\ Manufacturers have repeatedly and consistently told the 
agencies that the high costs for redesign and the increased 
manufacturing complexity that would result from resizing engines for 
small technology changes preclude them from doing so. It would be 
unreasonable and unaffordable to resize powertrains for every unique 
combination of technologies, and exceedingly so for every unique 
combination technologies across every vehicle model due to the extreme 
manufacturing complexity that would be required to do so. Engine 
displacements are further described in Section VI.C.1 Engine Paths.
---------------------------------------------------------------------------

    \502\ See 83 FR 43027 (Aug. 24, 2018).
---------------------------------------------------------------------------

    To address this issue, and consistent with past rulemakings, the 
NPRM simulation allowed engine resizing when mass reductions of 7.1 
percent, 10.7 percent, 14.2 percent (and 20 percent for the final rule 
analysis) were applied to the vehicle curb weight,\503\ and when one 
powertrain architecture was replaced with another architecture during a 
redesign cycle.\504\ At its refresh cycle, a vehicle may also inherit 
an already resized powertrain from another vehicle within the same 
engine-sharing platform. The analysis did not re-size the engine in 
response to adding technologies that have smaller effects on vehicle 
performance. For instance, if a vehicle's curb weight is reduced by 3.6 
percent (MR1), causing the 0-60 mile per hour time to improve slightly, 
the analysis would not resize the engine. The criteria for resizing 
used for the analysis better reflects what is feasible for 
manufacturers to do.\505\
---------------------------------------------------------------------------

    \503\ These correspond, respectively, to reductions of 10%, 15%, 
20%, and 28.2% of the vehicle glider mass. For more detail on glider 
mass calculation, see section VI.C.4 Mass Reduction.
    \504\ Some engine and accessory technologies may be added to an 
engine without an engine architecture change. For instance, 
manufacturers may adapt, but not replace engine architectures to 
include cylinder deactivation, variable valve lift, belt-integrated 
starter generators, and other basic technologies. However, switching 
from a naturally aspirated engine to a turbo-downsized engine is an 
engine architecture change typically associated with a major 
redesign and radical change in engine displacement.
    \505\ For instance, a vehicle would not get a modestly bigger 
engine if the vehicle comes with floor mats, nor would the vehicle 
get a modestly smaller engine without floor mats. This example 
demonstrates small levels of mass reduction. If manufacturers 
resized engines for small changes, manufacturers would have 
dramatically more part complexity, potentially losing economies of 
scale.
---------------------------------------------------------------------------

    Automotive manufacturers have commented that the CAFE model's 
consideration of the constraints faced in relation to vehicle 
performance and economies of scale are realistic.
    Industry associations and individual manufacturers widely supported 
the use of the performance metrics used in the NPRM analysis, the use 
of standard and higher performance technology classes, and the 
representation in the analysis of the real-world manufacturing 
complexity constraints and criteria for powertrain redesign.
    The Alliance of Automobile Manufacturers (Alliance), Ford, and 
Toyota stated that the inclusion of additional performance metrics such 
as gradeability are appropriate. Specifically in support of the 
gradeability performance criteria, the Alliance commented that 
``performance metrics related to vehicle operation in top gear are just 
as critical to customer acceptance as are performance metrics such as 
0-60 mph times that focus on performance in low-gear ranges.'' \506\ 
The Alliance also commented specifically on the relationship between 
gradeability and downsized engines, stating that as ``engine downsizing 
levels increase, top-gear gradeability becomes more and more 
important,'' and further that the consideration of gradeability ``helps 
prevent the inclusion of small displacement engines that are not 
commercially viable and that would artificially inflate fuel savings.'' 
\507\
---------------------------------------------------------------------------

    \506\ Alliance of Automobile Manufacturers, Attachment ``Full 
Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 139.
    \507\ Alliance of Automobile Manufacturers, Attachment ``Full 
Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 135.
---------------------------------------------------------------------------

    Ford and Toyota similarly commented in support of the CAFE model's 
consideration of multiple performance criteria. Ford stated that this 
model ``takes a more realistic approach to performance modeling'' and 
``better replicates OEM attribute-balancing practices.'' Ford stated 
furthermore that ``OEMs must ensure that each individual performance 
measure--and not an overall average--meets its customer's 
requirements,'' and that, in contrast, previous analyses did ``not 
align with product planning realities.'' \508\ Toyota commented in 
support of including gradeability as a performance metric ``to avoid 
underpowered engines and overestimated fuel savings.'' \509\
---------------------------------------------------------------------------

    \508\ Ford, Attachment 1, Docket No. NHTSA-2018-0067-11928, at 
8.
    \509\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at 
6.
---------------------------------------------------------------------------

    Toyota and the Alliance commented that the inclusion of performance 
vehicle classes addressed the market reality that some consumers will 
purchase vehicles for their performance attributes and will accept the 
corresponding reduction in fuel economy. Furthermore, Toyota commented 
that most consumers consider more than just fuel economy when 
purchasing a vehicle, and that ``dedicating all powertrain improvements 
to fuel efficiency is inconsistent with market reality.'' Toyota 
``supports the agencies' inclusion of performance classes in compliance 
modeling where a subset of certain models is defined to have higher 
performance and a commensurate reduction in fuel efficiency.'' \510\ 
Also in support of the addition of performance vehicle classes, the 
Alliance commented that ``vehicle categories have been increased to 10 
to better recognize the range of 0-60 performance

[[Page 24335]]

characteristics within each of the 5 previous categories, in 
recognition of the fact that many vehicles in the baseline fleet 
significantly exceeded the previously assumed 0-60 performance metrics. 
This provides better resolution of the baseline fleet and more accurate 
estimates of the benefits of technology.'' \511\
---------------------------------------------------------------------------

    \510\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at 
6.
    \511\ Alliance of Automobile Manufacturers, Attachment ``Full 
Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 135.
---------------------------------------------------------------------------

    Toyota also commented in support of various real-world 
manufacturing complexity constraints employed in the analysis for 
powertrain redesigns. Toyota commented that model parameters such as 
redesign cycles and engine sharing across vehicle models place a more 
realistic limit on the number of engines and transmissions that a 
manufacturer is capable of introducing. Toyota also commented in 
support of the constraints that the CAFE model placed on engine 
resizing, stating that ``there are now more realistic limits placed on 
the number of engines and transmissions in a powertrain portfolio which 
better recognizes [how] manufacturers must manage limited engineering 
resources and control supplier, production, and service costs. 
Technology sharing and inheritance between vehicle models tends to 
limit the rate of improvement in a manufacturer's fleet.'' Toyota 
pointed out that this is in contrast to previous analyses in which 
resizing was too unconstrained, which created an ``unmanageable number 
of engine configurations within a vehicle platform'' and spawned cases 
where ``engine downsizing and power reduction sometimes exceeded limits 
beyond basic acceleration requirements needed for vehicle safety and 
customer satisfaction.'' \512\
---------------------------------------------------------------------------

    \512\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at 
6.
---------------------------------------------------------------------------

    The above comments from the Alliance, Ford, and Toyota support the 
methodologies the agencies employed to conduct a performance neutral 
analysis. These methodologies helped to ensure that multiple 
performance criteria, including gradeability, are all individually 
accounted for and maintained when a vehicle powertrain is resized, and 
that real-world manufacturing complexity constraints are factored in to 
the agencies' analysis of feasible pathways manufacturers could take to 
achieve compliance with CAFE standards. The agencies continue to 
believe this is a reasonable approach for the aforementioned reasons.
    Environmental advocacy groups and CARB criticized the CAFE model's 
engine resizing constraints and how they affected the acceleration 
performance criteria.
    CARB, The International Council on Clean Transportation (ICCT), the 
Union of Concerned Scientists (UCS), and the American Council for an 
Energy-Efficient Economy (ACEEE) commented that the CAFE model was not 
performance neutral, allowing an improvement in performance which 
reduced the effectiveness of applied fuel-saving technologies and/or 
increased the cost of compliance. Specifically, ACEEE stated that there 
appeared to be a shortfall in the fuel economy effectiveness of 
technology packages, potentially resulting from the effectiveness being 
``consumed'' by additional vehicle performance rather than improvement 
of fuel economy. Several of these same commenters conducted analyses 
attempting to quantify the magnitude of these changes in vehicle 
performance for various vehicle technology classes.
    CARB commented on the performance shift of several vehicle types. 
Analyzing the 0-60 acceleration for the medium car non-performance 
technology class and looking at all cases with resized engines, CARB 
claimed that ``effectively half of the simulations resulted in improved 
performance.'' \513\ Focusing on electrified vehicles in that same 
technology class, CARB stated that ``the data from the Argonne 
simulations shows that 76 of the 88 strong electrified packages 
(including P2HPV, SHEVPS, BEV, FCEV, PHEV), where Argonne purposely 
resized the system to maintain performance neutrality, resulted in 
notably faster 0 to 60 mph acceleration times and passing times.'' 
Specifically regarding parallel hybrid electric vehicles (SHEVP2), CARB 
stated that all modeled packages resulted in improved performance.\514\ 
UCS commented that the NPRM analysis allowed too much change in vehicle 
performance, stating that ``while some performance creep may be 
reasonable'' many performance values show ``an overlap between 
performance and non-performance vehicles'' within the compact car 
technology class.\515\
---------------------------------------------------------------------------

    \513\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 180. Note that the target acceleration 
time for medium car non-performance is in fact 9.0 seconds, as 
indicated in ANL documentation, but was incorrectly reported as 9.4s 
in NPRM table II-7 in the NPRM.
    \514\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 186.
    \515\ Union of Concerned Scientists, Attachment 2, Docket No. 
NHTSA-2018-0067-12039, at 24.
---------------------------------------------------------------------------

    The agencies carefully considered these comments. For the NPRM 
analysis, the SHEVP2 engines/electric-motors were resized if the 0-60 
acceleration time was worse than the target, but not resized if the 
acceleration time was better than the target. This approach maintained 
vehicle performance with a depleted battery (without electric assist) 
in order to maintain fully the performance and utility characteristics 
under all conditions, and improved performance when electric assist was 
available (when the battery is not depleted), such as during the 0-60 
mph acceleration. The agencies found that this resulted in some 
parallel hybrid vehicles having improved 0-60 acceleration times. This 
approach was initially chosen for the NPRM because the resulting level 
of improved performance was consistent with observations of how 
industry had applied SHEVP2 technology. However, in assessing the CARB 
comment, the agencies balanced the NPRM approach for SHEVP2 performance 
with the agencies' criteria of maintaining vehicle functionality and 
performance when technology is applied. Both could not be fully 
achieved under all conditions for the case of the SHEVP2.
    The agencies concluded it is reasonable to maintain performance 
including electric assist when SHEVP2 technology is applied to a 
standard (non-performance) vehicle, and therefore the analysis for the 
final rule allows upsizing and downsizing of the parallel hybrid 
powertrain (SHEVP2) using the 0.2 seconds window around the 
target.\516\ For performance vehicles, the agencies concluded that it 
remains reasonable to maintain vehicle performance with a depleted 
battery (without electric assist) in order to maintain fully the 
performance characteristics under all conditions, and continued to use 
the NPRM methodology.
---------------------------------------------------------------------------

    \516\ To represent marketplace trends better, the performance 
class of SHEVP2's allow acceleration time below 0.2 seconds less 
than the target, and PHEV20's and PHEV50's inherit combustion engine 
size from the conventional powertrain they are replacing. Further 
discussion of resizing targets can be found in Chapter 8 of the ANL 
Model Documentation for the final rule analysis.
---------------------------------------------------------------------------

    The refinement for the standard performance SHEVP2 resolved the 
electrified packages issue identified by CARB, and also addressed most 
of the change in performance in the overall fleet, including with 
compact cars as mentioned by UCS. As explained further below, the 
agencies assessed performance among the alternatives for the final rule 
analysis. That assessment showed that, with the final rule refinements, 
245 out of 255 total resized vehicles (96 percent of vehicles) in the 
medium non-performance class (same

[[Page 24336]]

class focused on by CARB), had 0-60 mph acceleration times within the 
plus-or-minus 0.2 second window (8.8 to 9.2 seconds).\517\ The only 
vehicles outside the window were certain strong electrified vehicles 
which exceeded 0-60 the acceleration target as a result of achieving 
other performance criteria, such as the US06 driving cycles in all-
electric-mode.\518\
---------------------------------------------------------------------------

    \517\ This includes 135 strong electrified vehicles.
    \518\ As noted earlier, electrified vehicles had to be capable 
of successfully completing UDDS or US06 driving cycles in all-
electric mode, and in some cases the resulting motor size produced 
improved acceleration times.
---------------------------------------------------------------------------

    The assessment also showed that for the small car class (mentioned 
by UCS) the acceleration times of performance and non-performance 
vehicles do not go beyond each other's targets. For example, the 
vehicle in the small car class with the very best 0-60 mph time and a 
conventional powertrain achieves an 8.38 second 0-60 mph time, which is 
slower than the performance small car baseline of 8 seconds. This 
vehicle had multiple incremental technologies applied, including for 
example aerodynamic improvements, and has not reached the threshold for 
engine resizing.\519\ After engine resizing, the ``fastest'' 
conventional small car has a 0-60 mph time of 9.9 seconds, only 0.1 
seconds from the target of 10 seconds.\520\
---------------------------------------------------------------------------

    \519\ Discussion of engine resizing can be found in Section 
VI.B.3.a)(5).
    \520\ See NPRM Autonomie simulation database for Small cars, 
Docket ID NHTSA-2018-0067-1855.
---------------------------------------------------------------------------

    CARB also commented on the improvement of ``passing times,'' or 50-
80 mph high-speed acceleration times. As stated above, an improvement 
in one or more of the performance criteria is an expected outcome when 
using the rulemaking analysis methodology that resizes powertrains such 
that there is no degradation in any of the performance metrics. 
Consistent with past rulemakings, the agencies do not believe it is 
appropriate for the rulemaking analysis to show pathways that degrade 
vehicle performance or utility for one or more of the performance 
criteria, as doing so would adversely impact functional capability of 
the vehicle and could lead to customer dissatisfaction. The agencies 
agree there is very small increase in passing performance for some 
technology combinations, and believe this is an appropriate outcome. 
High-speed acceleration is rarely the least-capable performance 
criteria.
    CARB, ICCT, UCS, and H-D Systems (HDS), in an attempt to identify a 
potential cause for changes in performance, commented that the CAFE 
model should have placed fewer constraints on engine resizing. CARB and 
ICCT commented that engine resizing should have been allowed even at 
low levels of mass reduction. Comments from CARB, UCS, HDS, and ICCT 
stated that engine resizing should also have been allowed for other 
incremental technologies, and within their comments they conducted 
performance analysis of non-resized cases.
    CARB claimed that requiring a minimum of 7.1 percent curb weight 
reduction before engine resizing is a constraint that ``limits the 
optimization of the technologies being applied.'' \521\ UCS stated that 
``a significant share of the benefit of a few percent reduction in mass 
has gone towards improved performance rather than improved fuel 
economy, leaving a substantial benefit of mass reduction underutilized 
and/or uncounted.'' \522\ ICCT also commented that ``when vehicle 
lightweighting is deployed at up to a 7 percent mass reduction, the 
engine is not resized even though less power would be needed for the 
lighter vehicle, meaning any such vehicles inherently are higher 
performance.'' \523\
---------------------------------------------------------------------------

    \521\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 178. Note, a 7.1% curb weight reduction 
equates to the agencies' third level of mass reduction (MR3); 
additional discussion of engine resizing for mass reduction can be 
found in Section VI.B.3.a)(4) Autonomie Sizes Powertrains for Full 
Vehicle Simulation] and in the ANL Model Documentation for the final 
rule analysis.
    \522\ Union of Concerned Scientists, Attachment 2, Docket No. 
NHTSA-2018-0067-12039, at 11.
    \523\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-50.
---------------------------------------------------------------------------

    UCS and HDS commented on the lack of resizing for technologies 
other than mass reduction, with HDS stating that ``the Agencies 
incorrectly limited the efficacy of technologies that reduce tractive 
load because their modeling does not re-optimize engine performance 
after applying these technologies.'' \524\ CARB also commented that the 
lack of resizing when a BISG or CISG system is added ``results in a 
less than optimized system that does not take full advantage of the 
mild hybrid system.'' Similarly, ICCT noted a case in which a Dodge RAM 
``did not apply engine downsizing with the BISG system on that truck, 
so there are also significant performance benefits that should be 
accounted for, meaning that for constant-performance the fuel 
consumption reduction would be even greater.'' \525\
---------------------------------------------------------------------------

    \524\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
12395, at 4. For reference, technologies that reduce tractive road 
load include mass reduction, aerodynamic drag reduction, and tire 
rolling resistance reduction.
    \525\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-24.
---------------------------------------------------------------------------

    CARB further commented on the performance improvement in cases 
without engine resizing by stating that ``94 percent of the packages 
modeled result in improved performance,'' and that for these non-
resized cases that were actually adopted by a vehicle in the 
simulation, ``fewer than 20 percent maintained baseline performance 
with gains of 2 percent or less in acceleration time.'' \526\ Referring 
specifically to non-resized electrified vehicles, CARB also stated that 
``44,878 of the 53,818 packages, or greater than 83 percent, result in 
improved performance.'' \527\ CARB also commented that engine sharing 
across different vehicles within a platform, which in some cases may 
constrain resizing for a member of that platform, should not dictate 
that these engines must remain identical in all aspects, and that 
``this overly restrictive sharing of identical engines newly imposed in 
the CAFE Model is not consistent with today's industry practices and 
results in less optimal engine sizing and causes a systematic 
overestimation of technology costs to meet the existing standards.'' 
\528\
---------------------------------------------------------------------------

    \526\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 183.
    \527\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 187.
    \528\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 185.
---------------------------------------------------------------------------

    The agencies note broadly, in response to these comments, that when 
conducting an analysis which balances performance neutrality against 
the realities faced by manufacturers, such as manufacturing complexity, 
economies of scale, and maintaining the full range of performance 
criteria, it is inevitable to observe at least some minor shift in 
vehicle performance. For example, if a new transmission is applied to a 
vehicle, the greater number of gear ratios helps the engine run in its 
most efficient range which improves fuel economy, but also helps the 
engine to run in the optimal ``power band'' which improves performance. 
Thus, the technology can provide both improved fuel economy and 
performance. Another example is applying a small amount of mass 
reduction that improves both fuel economy and performance by a small 
amount. Resizing the engine to maintain performance in these examples 
would require a unique engine displacement that is only slightly 
different than the baseline engine. While engine resizing in these 
incremental cases could have some small benefit to fuel economy, the

[[Page 24337]]

gains may not justify the costs of producing unique niche engines for 
each combination of technologies. If manufacturers were to produce 
marginally downsized engines to complement every small increment of 
mass reduction or technology, the resulting large number of engine 
variants that would need to be manufactured would cause a substantial 
increase in manufacturing complexity, and require significant changes 
to manufacturing and assembly plants and equipment.\529\ The high costs 
would be economically infeasible.
---------------------------------------------------------------------------

    \529\ For example, each unique engine would require unique 
internal components such as crankshafts, pistons, and connecting 
rods, as well as unique engine calibrations for each displacement. 
Assembly plants would need to stock and feed additional unique 
engines to the stations where engines are dressed and inserted into 
vehicles.
---------------------------------------------------------------------------

    Also, as noted in the NPRM, the 2015 NAS report stated that ``[f]or 
small (under 5 percent [of curb weight]) changes in mass, resizing the 
engine may not be justified, but as the reduction in mass increases 
(greater than 10 percent [of curb weight]), it becomes more important 
for certain vehicles to resize the engine and seek secondary mass 
reduction opportunities.'' \530\ In consideration of both the NAS 
report and comments received from manufacturers, the agencies 
determined it would be reasonable to allow allows engine resizing upon 
adoption of 7.1 percent, 10.7 percent, 14.2 percent, and 20 percent 
curb weight reduction, but not at 3.6 percent and 5.3 percent.\531\ 
Resizing is also allowed upon changes in powertrain type or the 
inheritance of a powertrain from another vehicle in the same platform. 
The increments of these higher levels of mass reduction, or complete 
powertrain changes, more appropriately match the typical engine 
displacement increments that are available in a manufacturer's engine 
portfolio.
---------------------------------------------------------------------------

    \530\ National Research Council. 2011. Assessment of Fuel 
Economy Technologies for Light-Duty Vehicles. Washington, DC--The 
National Academies Press. http://nap.edu/12924.
    \531\ These curb weight reductions equate to the following 
levels of mass reduction as defined in the analysis: MR3, MR4, MR5 
and MR6, but not MR1 and MR2; additional discussion of engine 
resizing for mass reduction can be found in Section VI.B.3.a)(6) 
Autonomie Sizes Powertrains for Full Vehicle Simulation.
---------------------------------------------------------------------------

    The agencies point to the comments from manufacturers, discussed 
further above, which support the agencies' assertion that the CAFE 
model's resizing constraints are appropriate. As discussed previously, 
Toyota commented that this approach better considers the constraints of 
engineering resources and manufacturing costs and results in a more 
realistic number of engines and transmissions.\532\ The Alliance also 
commented on the benefit of constraining engine resizing, stating that 
``the platform and engine sharing methodology in the model better 
replicates reality by making available to each manufacturer only a 
finite number of engine displacements, helping to prevent 
unrealistically `over-optimized' engine sizing.'' \533\
---------------------------------------------------------------------------

    \532\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at 
6.
    \533\ Alliance of Automobile Manufacturers, Attachment ``Full 
Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 140.
---------------------------------------------------------------------------

    Another comment from CARB stated that engine resizing ``was only 
simulated for cases where those levels of mass reduction were applied, 
in the absence of virtually all other technology or efficiency 
improvements.'' \534\ The agencies do not agree that resizing should be 
simulated in all cases which involve small incremental technologies. In 
the final rule analysis, vehicles can have engines resized at four (out 
of six) levels of mass reduction technology, during a vehicle redesign 
cycle which changes powertrain architecture, and by inheritance during 
a vehicle refresh cycle. As discussed previously, the application of 
small incremental technologies such as reductions in aerodynamic drag 
or rolling resistance does not justify the high cost and complexity of 
producing additional varieties of engine sizes. Accordingly, for each 
curb weight reduction level of 7.1 percent or above and for each 
vehicle technology class, Autonomie sized a baseline engine by running 
a simulation of a vehicle without incremental technologies applied; 
then, those baseline engines were inherited by all other simulations 
using the same levels of curb weight reduction, which also added any 
variety of incremental technologies.\535\ For further clarification, in 
any case in which a vehicle adopts a 7.1 percent or more curb weight 
reduction, no matter what other technologies were already present or 
are added to the vehicle in conjunction with the mass reduction, that 
vehicle will receive an engine which has been appropriately sized for 
the newly applied mass reduction level.\536\ This can be observed in 
the Autonomie simulation databases by tracking the ``EngineMaxPower'' 
column (not the ``VehicleSized'' column).
---------------------------------------------------------------------------

    \534\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 178.
    \535\ In the Autonomie simulation database files, the 
simulations which establish baseline sized engines are marked 
``yes'' in the ``VehicleSized'' column, and the subsequent 
simulations which use this engine and add other incremental 
technologies are marked ``inherited.'' For a list of Autonomie 
simulation database files, see Table VI-4 Autonomie Simulation 
Database Output Files in Section VI.A.7 Structure of Model Inputs 
and Outputs.
    \536\ For example, if a vehicle possesses MR2, AERO1, and ROLL1 
and subsequently adopts MR3, AERO1, ROLL2, the vehicle will adopt 
the lower engine power level associated with MR3. As a counter 
example, if a vehicle possesses MR3, ROLL1, and AERO1 and 
subsequently adopts MR3, ROLL1, AERO2, the engine will not be 
resized and it will retain the power level associated with MR3.
---------------------------------------------------------------------------

    Finally, ICCT claimed that the agencies did not sufficiently report 
performance-related vehicle information. ICCT commented that the output 
files did not show data on ``engine displacement, the maximum power of 
each engine, the maximum torque of each engine, the initial and final 
curb weight of each vehicle (in absolute terms), and estimated 0-60 mph 
acceleration.'' ICCT claimed that because this data was not found, the 
agencies are ``showing that they have not even attempted to analyze 
accurately the future year fleet for their performance'' and that ``the 
agencies are intentionally burying a critical assumption, whereby their 
future fleet has not been appropriately downsized, and it therefore has 
greatly increased utility and performance characteristics.'' \537\
---------------------------------------------------------------------------

    \537\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-74.
---------------------------------------------------------------------------

    In fact, for the NPRM, and again for this final rule, the agencies 
did analyze vehicle performance and have made the data available to the 
public. An indication of the actual engine displacement change is 
available by noting the displacements used in Automonie simulation 
database for each of the technology states. The displacements reported 
in Autonomie are used by the full-vehicle-simulation within the 
Autonomie model, and while they do not directly represent each specific 
vehicle's actual engine sizes, they do fully reflect the relative 
change in engine size that is applied to each vehicle. It is the 
relative change in engine size that is relevant for the analysis. 
Similarly, the vehicle power and torque used by the full vehicle 
simulations are reported in the Autonomie simulation databases; their 
values and relative change across an engine resizing event can be 
observed. Initial and final curb weights for the analysis fleet are 
reported in Vehicles Report output file column titled ``CW Initial'' 
and ``CW,'' respectively. The time required for 0-60 mph acceleration 
is reported in the Autonomie simulation database files. A detailed 
description of the engine resizing methodology is available in the 
Argonne Model

[[Page 24338]]

Documentation, which explains how vehicle characteristics are used to 
calculate powertrain size.\538\ These data and information that are 
available in the Autonomie and CAFE model documentation provide the 
information needed to analyze performance, and in fact, this is 
evidenced by the statements of numerous commenters discussed in this 
section. The agencies have conducted their own performance analysis, 
which is discussed further below, using the same data documentation 
mentioned here.
---------------------------------------------------------------------------

    \538\ See Chapter 8 of the ANL Model documentation for the final 
rule analysis.
---------------------------------------------------------------------------

    Updates to the CAFE model have minimized performance shift over the 
simulated model years, and have eliminated performance differences 
between simulated standards.
    The Autonomie simulation updates, discussed previously, were 
included in the final rule analysis, and have resulted in average 
performance that is similar across the regulatory alternatives. Because 
the regulatory analysis compares differences in impacts among the 
alternatives, the agencies believe that having consistent performance 
across the alternatives is an important aspect of performance 
neutrality. If the vehicle fleet had performance gains which varied 
significantly depending on the alternative, performance differences 
would impact the comparability of the simulations. Using the NPRM CAFE 
model data, the agencies analyzed the sales-weighted average 0-60 
performance of the entire simulated vehicle fleet for MYs 2016 and 
2029, and identified that the Augural standards had 4.7 percent better 
0-60 mph acceleration time compared to the NPRM preferred alternative, 
which had no changes in standards in MYs 2021-2026.\539\ This 
assessment confirmed the observations of the various commenters. With 
the refinements that were incorporated for the final rule, similar 
analysis showed that the Augural standards had a negligible 0.1 percent 
difference in 0-60 mph acceleration time compared to the NPRM preferred 
alternative.\540\
---------------------------------------------------------------------------

    \539\ The agencies' analysis matched all MY 2016 and MY 2029 
vehicles in the NPRM Vehicles Report output file, under both the 
Augural standards and preferred alternative, with the appropriate 0-
60 mph acceleration time from the NPRM Autonomie simulation 
databases. This was done by examining each vehicle's assigned 
technologies, finding the Autonomie simulation with the 
corresponding set of technologies, and extracting that simulation's 
0-60 mph acceleration time. This process effectively assigned a 0-60 
time to every vehicle in the fleet for four scenarios: (1) MY 2016 
under augural standards, (2) MY 2016 under the preferred 
alternative, (3) MY 2029 under augural standards, and (4) MY 2029 
under the preferred alternative. For each scenario, an overall 
fleet-wide weighted average 0-60 time was calculated, using each 
vehicle's MY2016 sales volumes as the weight. For more information, 
see the FRIA Section VI.
    \540\ This updated analysis used the FRM CAFE Model Vehicles 
Report output file and the FRM Autonomie simulation databases. The 
final rule analysis introduced an updated MY 2017 fleet as a 
starting point, replacing the NPRM 2016MY fleet. For more 
information, see the FRIA Chapter VI.
---------------------------------------------------------------------------

    The updates applied to the final rule Autonomie simulations also 
resulted in further minimizing the performance change across model 
years. As the agencies attempted to minimize this performance shift 
occurring ``over time,'' it was also acknowledged that a small increase 
would be expected and would be reasonable. This increase is attributed 
to the analysis recognizing the practical constraints on the number of 
unique engine displacements manufacturers can implement, and therefore 
not resizing powertrains for every individual technology and every 
combination of technologies when the performance impacts are small. 
Perfectly equal performance with 0 percent change would not be 
achievable while accounting for these real world resizing constraints. 
The performance analysis in the 2011 NAS report shared a similar view 
on performance changes, stating that ``truly equal performance involves 
nearly equal values . . . within 5 percent.'' \541\ In response to 
comments, using NPRM CAFE model data, the agencies analyzed the sales-
weighted average 0-60 performance of the entire simulated vehicle 
fleet, and identified that the performance increase from MYs 2016 and 
2029 was 7.5 percent under Augural Standards and 3.1 percent under the 
NPRM preferred alternative standards. The agencies conducted a similar 
analysis using final rule data and found the performance increase over 
time from MYs 2017 to 2029 was 3.9 percent for Augural Standards and 
4.0 percent for the NPRM preferred alternative standards. The agencies 
determined this change in performance is reasonable and note it is 
within the 5 percent bound in discussed by NAS in its 2011 report.
---------------------------------------------------------------------------

    \541\ National Research Council. 2011. Assessment of Fuel 
Economy Technologies for Light-Duty Vehicles. Washington, DC--The 
National Academies Press, at 62. http://nap.edu/12924.
---------------------------------------------------------------------------

    This assessment shows that for the final rule analysis, performance 
is neutral across regulatory alternatives and across the simulated 
model years allowing for fair, direct comparison among the 
alternatives.
(7) How the Agencies Simulated Vehicle Models on Test Cycles
    After vehicle models are built for every combination of 
technologies and vehicle classes represented in the analysis, Autonomie 
simulates their performance on test cycles to calculate the 
effectiveness improvement of the fuel-economy-improving technologies 
that have been added to the vehicle. Discussed earlier, the agencies 
minimize the impact of potential variation in determining effectiveness 
by using a series of tests and procedures specified by federal law and 
regulations under controlled conditions.
    Autonomie simulates vehicles in a very similar process as the test 
procedures and energy consumption calculations that manufacturers must 
use for CAFE and CO2 compliance.542 543 544 
Argonne simulated each vehicle model on several test procedures to 
evaluate effectiveness. For vehicles with conventional powertrains and 
micro hybrids, Autonomie simulates the vehicles on EPA 2-cycle test 
procedures and guidelines.\545\ For mild and full hybrid electric 
vehicles and FCVs, Autonomie simulates the vehicles using the same EPA 
2-cycle test procedure and guidelines, and the drive cycles are 
repeated until the initial and final state of charge are within a SAE 
J1711 tolerance. For PHEVs, Autonomie simulates vehicles in similar 
procedures and guidelines as SAE J1711.\546\ For BEVs Autonomie 
simulates vehicles in similar procedures and guidelines as SAE 
J1634.\547\
---------------------------------------------------------------------------

    \542\ EPA, ``How Vehicles are Tested.'' https://www.fueleconomy.gov/feg/how_tested.shtml. Last accessed Nov 14, 
2019.
    \543\ ANL model documentation for final rule Chapter 6. Test 
Procedures and Energy Consumption Calculations.
    \544\ EPA Guidance Letter. ``EPA Test Procedures for Electric 
Vehicles and Plug-in Hybrids.'' Nov. 14, 2017. https://www.fueleconomy.gov/feg/pdfs/EPA%20test%20procedure%20for%20EVs-PHEVs-11-14-2017.pdf. Last accessed Nov. 7, 2019.
    \545\ 40 CFR part 600.
    \546\ PHEV testing is broken into several phased based on SAE 
J1711. Charge-Sustaining on the City cycle, Charge-Sustaining on the 
HWFET cycle, Charge-Depleting on the City and HWFET cycles.
    \547\ SAE J1634. ``Battery Electric Vehicle Energy Consumption 
and Range Test Procedure.'' July 12, 2017.
---------------------------------------------------------------------------

b) Selection of One Full-Vehicle Modeling and Simulation Tool
    The NPRM described tools that the agencies previously used to 
estimate technology effectiveness. For the analysis supporting the 2012 
final rule for MYs 2017 and beyond, the agencies used technology 
effectiveness estimates from EPA's lumped parameter model (LPM). The 
LPM was calibrated using data from vehicle simulation work performed by 
Ricardo Engineering.\548\

[[Page 24339]]

The agencies also used full vehicle simulation modeling data from 
Autonomie vehicle simulations performed by Argonne for mild hybrid and 
advanced transmission effectiveness estimates.549 550
---------------------------------------------------------------------------

    \548\ Response to Peer Review of: Ricardo Computer Simulation of 
Light-Duty Vehicle Technologies for Greenhouse Gas Emission 
Reduction in the 2020-2025 Timeframe, EPA-420-R-11-021 (December 
2011), available at https://nepis.epa.gov/Exe/ZyPDF.cgi/P100D5BX.PDF?Dockey=P100D5BX.PDF.
    \549\ Joint TSD: Final Rulemaking for 2017-2025 Light-Duty 
Vehicle Greenhouse Emission Standards and Corporate Average Fuel 
Economy Standards. August 2012. EPA-420-R-12-901.3.3.1.3 Argonne 
National Laboratory Simulation Study p. 3-69.
    \550\ Moawad, A. and Rousseau, A., ``Impact of Electric Drive 
Vehicle Technologies on Fuel Efficiency,'' Energy Systems Division, 
Argonne National Laboratory, ANL/ESD/12-7, August 2012.
---------------------------------------------------------------------------

    For the 2016 Draft TAR analysis, EPA and NHTSA used two different 
full system simulation programs for complementary but separate 
analyses. NHTSA used Argonne's Autonomie tool, described in detail 
above, with engine map inputs developed by IAV using GT-Power in 2014, 
and updated in 2016.551 552 553 Argonne, in coordination 
with NHTSA, developed a methodology for large scale simulation using 
Autonomie and distributed computing, thus overcoming one of the 
challenges to full vehicle simulation that the NAS committee outlined 
in its 2015 report and implementing a recommendation that the agencies 
use full-vehicle simulation to improve the analysis method of 
estimating technology effectiveness.\554\ EPA used a limited number of 
full-vehicle simulations performed using its ALPHA model, an EPA-
developed full-vehicle simulation model,\555\ to calibrate the LPM, 
used to estimate technology effectiveness. EPA also used the same 
modeling approach for its Proposed Determination analysis.\556\
---------------------------------------------------------------------------

    \551\ GT-Power Engine Simulation Software. https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software/. Last accessed Oct. 10, 2019.
    \552\ 2016 Draft TAR Engine Maps by IAV Automotive Engineering 
using GT-Power. https://www.nhtsa.gov/staticfiles/rulemaking/pdf/cafe/IAV_EngineMaps_Details.xlsx. Lass accessed Oct. 10, 2019.
    \553\ NHTSA-2018-0067-0003. ANL--Summary of Main Component 
Performance Assumptions NPRM.
    \554\ See National Research Council. 2015. Cost, Effectiveness, 
and Deployment of Fuel Economy Technologies for Light-Duty Vehicles. 
Washington, DC: The National Academies Press [hereinafter ``2015 NAS 
Report''] at p. 263, available at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles (last accessed June 21, 2018). See also A. 
Moawad, A. Rousseau, P. Balaprakash, S. Wild, ``Novel Large Scale 
Simulation Process to Support DOT's CAFE Modeling System,'' 
International Journal of Automotive Technology (IJAT), Paper No. 
220150349, Nov 2015; Pagerit, S., Sharper, P., Rousseau, A., Sun, Q. 
Kropinski, M. Clark, N., Torossian, J., Hellestrand, G., ``Rapid 
Partitioning, Automatic Assembly and Multicore Simulation of 
Distributed Vehicle Systems.'' ANL, General Motors, EST Embedded 
Systems Technology. 2015. https://www.autonomie.net/docs/5%20-%20Presentations/VPPC2015_ppt.pdf. Last accessed Dec. 9, 2019.
    \555\ See Lee, B., S. Lee, J. Cherry, A. Neam, J. Sanchez, and 
E. Nam. 2013. Development of Advanced Light-Duty Powertrain and 
Hybrid Analysis Tool. SAE Technical Paper 2013-01-0808. doi: 
10.4271/2013-01-0808.
    \556\ Proposed Determination on the Appropriateness of the Model 
Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards 
under the Midterm Evaluation, EPA-420-R-16-020 (November 2016), 
available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3DO.pdf; Final Determination on the 
Appropriateness of the Model Year 2022-2025 Light-Duty Vehicle 
Greenhouse Gas Emissions Standards under the Midterm Evaluation, 
EPA-420-R-17-001 (January 2017), available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100QQ91.pdf.
---------------------------------------------------------------------------

    In the subsequent August 2017 Request for Comment on 
Reconsideration of the Final Determination of the Mid-Term Evaluation 
of Greenhouse Gas Emissions Standards for MY 2022-2025 Light-Duty 
Vehicles, the agencies requested comments on whether EPA should use 
alternative methodologies and modeling, including the Autonomie full-
vehicle simulation tool and DOT's CAFE model, for the analysis that 
would accompany its revised Final Determination.\557\ As discussed in 
the NPRM, stakeholders questioned the efficacy of the combined outputs 
and assumptions of the LPM and ALPHA,\558\ especially as the tools were 
used to evaluate increasingly heterogeneous combinations of 
technologies in the vehicle fleet.\559\
---------------------------------------------------------------------------

    \557\ 82 FR 39551 (Aug. 21, 2017).
    \558\ 83 FR 43022 (``At NHTSA-2016-0068-0082, p. 49, FCA 
provided the following comments, ``FCA believes EPA is 
overestimating the benefits of technology. As the LPM is calibrated 
to those projections, so too is the LPM too optimistic.'' FCA also 
shared the chart, `LPM vs. Actual for 8 Speed Transmissions.' '').
    \559\ 83 FR 43022 (referencing Automotive News ``CAFE math gets 
trickier as industry innovates'' (Kulisch), March 26, 2018.).
---------------------------------------------------------------------------

    More specifically, the Auto Alliance noted that their previous 
comments to the midterm evaluation, in addition to comments from 
individual manufacturers, highlighted multiple concerns with EPA's 
ALPHA model that were unresolved, but addressed in Autonomie.\560\ 
First, the Alliance expressed concern over ALPHA modeling errors 
related to road load reductions, stating that an error derived from how 
mass and coast-down coefficients were updated when mass, tire and aero 
improvements were made resulted in benefits overstated by 3 percent to 
11 percent for all vehicle types. Next, the Alliance repeated its 
concern that EPA should consider top-gear gradeability as one of its 
performance metrics to maintain functionality, noting that EPA had 
acknowledged the industry's comments in the Proposed Determination, 
``but generally dismissed the auto industry concerns.'' Additional 
analysis by EPA in its Response to Comments document did not allay the 
Alliance's concerns,\561\ as the Alliance concluded that ``[c]onsistent 
with the National Academy of Sciences recommendation from 2011, EPA 
should monitor gradeability to ensure minimum performance.''
---------------------------------------------------------------------------

    \560\ EPA-HQ-OAR-2015-0827-9194, at p. 36-44.
    \561\ The Alliance noted that in higher-gear-count 
transmissions, like 8-speed automatics, modeled by ALPHA with an 
expanded ratio spread to achieve fuel economy, are concerning for 
gradeability. Additionally, infinite engine downsizing along with 
expanded ratio spread transmission, in real world gradeability may 
cause further deteriorate as modeled in ALPHA, which leads to 
inflated effectiveness values for powertrains that would not meet 
customer demands.
---------------------------------------------------------------------------

    Furthermore, the Alliance stated that ALPHA vehicle technology 
walks provided in response to manufacturer comments on the Proposed 
Determination did not correctly predict cumulative effectiveness when 
compared to technologies in real world applications. The Alliance 
stated that many of the individual technologies and assumptions used by 
ALPHA overestimated technology effectiveness and were derived from 
questionable sources. As an example, the Alliance referenced an engine 
map used by EPA to represent the Honda L15B7 engine, where the engine 
map data was collected by ``(1) taking a picture of an SAE document 
containing an image of the engine map, and then (2) `digitizing' the 
image by `tracing image contours' '' (citing EPA's ALPHA 
documentation). The Alliance could not definitively state whether the 
``digitization'' process, lack of detail in the source image, or 
another factor were the reasons that some regions of overestimated 
efficiency were observed in the engine map, but concluded that ``the 
use of this map should be discontinued within ALPHA,'' and ``any 
analysis conducted with it is highly questionable.'' Based on these 
concerns and others, the Alliance recommended that Autonomie be used to 
inform the downstream cost optimization models (i.e., the CAFE model 
and/or OMEGA).
    Global Automakers argued that NHTSA's CAFE model, which 
incorporates data from Autonomie simulations, provided a more 
transparent and discrete step through each of the modeling 
scenarios.\562\ Global pointed out that the LPM is ``of particular 
concern due to its simplified technology projection processes,'' and it 
``propagates fundamentally flawed

[[Page 24340]]

content into the ALPHA and OMEGA models and therefore cannot accurately 
assess the efficacy of fuel economy technologies.'' Global did note 
that EPA ``plans to abandon its reliance on LPM in favor of another 
modeling approach,'' referring to the RSE,\563\ but stated that ``EPA 
must provide stakeholders with adequate time to evaluate the updated 
modeling approach, ensure it is analytically robust, and provide 
meaningful feedback.'' Global Automakers concluded that EPA's engine 
mapping and tear-down analyses have played an important role in 
generating publicly-available information, and stated that the data 
should be integrated into the Autonomie model.
---------------------------------------------------------------------------

    \562\ EPA-HQ-OAR-2015-0827-9728, at 14.
    \563\ See Moskalik, A., Bolon, K., Newman, K., and Cherry, J. 
``Representing GHG Reduction Technologies in the Future Fleet with 
Full Vehicle Simulation,'' SAE Technical Paper 2018-01-1273, 2018, 
doi:10.4271/2018-01-1273. Since 2018, EPA has employed vehicle-
class-specific response surface equations automatically generated 
from a large number of ALPHA runs to more readily apply large-scale 
simulation results, which eliminated the need for manual calibration 
of effectiveness values between ALPHA and the LPM.
---------------------------------------------------------------------------

    On the other hand, other stakeholders commented that EPA's ALPHA 
modeling should continue to be used, for procedural reasons like, 
``[i]t would appear arbitrary for EPA now, after five years of modeling 
based on ALPHA, to declare it can no longer use its internally 
developed modeling tools and must rely solely on the Autonomie model,'' 
and ``[t]he ALPHA model is inextricably built into the regulatory and 
technical process. It will require years of new analysis to replace the 
many ALPHA and OMEGA modeling inputs and outputs that permeate the 
entire rulemaking process, should EPA suddenly decide to change its 
models.'' \564\ Commenters also cited technical reasons to use ALPHA, 
like EPA's progress benchmarking and validating the ALPHA model to over 
fifteen various MY 2013-2015 vehicles,\565\ and that technologies like 
the ``Atkinson 2'' engine technology were not considered in NHTSA's 
compliance modeling.\566\ Commenters also cited that ALPHA was created 
to be publicly available, open-sourced, and peer-reviewed, ``to allow 
for transparency to both automakers and public stakeholders, without 
hidden and proprietary aspects that are present in commercial modeling 
products.'' \567\
---------------------------------------------------------------------------

    \564\ EPA-HQ-OAR-2015-9826, at 39-40.
    \565\ EPA-HQ-OAR-2015-9826, at 40.
    \566\ EPA-HQ-OAR-2015-9197, at 28.
    \567\ EPA-HQ-OAR-2015-9826, at 38.
---------------------------------------------------------------------------

    The agencies described in the NPRM that after having reviewed 
comments about whether EPA should use alternative methodologies and 
modeling, and after having considered the matter fully, the agencies 
determined it was reasonable and appropriate to use Autonomie for full-
vehicle simulation.\568\ The agencies stated that nothing in Section 
202(a) of the Clean Air Act (CAA) mandated that EPA use any specific 
model or set of models for analysis of potential CO2 
standards for light duty vehicles. The agencies also distinguished the 
models and the inputs used to populate them; specifically, comments 
presented as criticisms of the models, such as ``Atkinson 2'' engine 
technology not considered in the compliance modeling, actually 
concerned model inputs.\569\
---------------------------------------------------------------------------

    \568\ 83 FR 43001.
    \569\ 83 FR 43002.
---------------------------------------------------------------------------

    With regards to modeling technology effectiveness, the agencies 
concluded that, although the CAFE model requires no specific approach 
to developing effectiveness inputs, the National Academy of Sciences 
recommended, and stakeholders have commented, that full-vehicle 
simulation provides the best balance between realism and practicality. 
As stated above, Argonne has spent several years developing, applying, 
and expanding means to use distributed computing to exercise its 
Autonomie full-vehicle simulation tool at the scale necessary for 
realistic analysis of technologies that could be used to comply with 
CAFE and CO2 standards, and this scalability and related 
flexibility (in terms of expanding the set of technologies to be 
simulated) makes Autonomie well-suited for developing inputs to the 
CAFE model.
    In response to the NPRM, the Auto Alliance commented that NHTSA's 
modeling and analysis tools are superior to EPA's, noting that NHTSA's 
tools have had a significant lead in their development.\570\ The 
Alliance pointed out that Autonomie was developed from the beginning to 
address the complex task of combining two power sources in a hybrid 
powertrain, while EPA's ALPHA model had not been validated or used to 
simulate hybrid powertrains. While both models are physics-based 
forward looking vehicle simulators, the Alliance commented that 
Autonomie is fully documented with available training, while ALPHA 
``has not been documented with any instructions making it difficult for 
users outside of EPA to run and interpret the model.'' The Alliance 
also mentioned specific improvements in the Autonomie simulations since 
the Draft TAR, including expanded performance classes to better 
consider vehicle performance characteristics, the inclusion of 
gradeability as a performance metric, as recommended by the NAS, the 
inclusion of new fuel economy technologies, and the removal of unproven 
technologies.
---------------------------------------------------------------------------

    \570\ NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    The Alliance, Global Automakers, and other automakers writing 
separately all stated that the agencies should use one simulation and 
modeling tool for analysis.571 572 The Alliance stated that 
since both the Autonomie and ALPHA modeling systems answer essentially 
the same questions, using both systems leads to inconsistencies and 
conflicts, and is inefficient and counterproductive.
---------------------------------------------------------------------------

    \571\ NHTSA-2018-0067-12073; NHTSA-2018-0067-12032. Comments of 
the Association of Global Automakers, Inc. on the Safer Affordable 
Fuel-Efficient Vehicles Rule Docket ID Numbers: NHTSA-2018-0067 and 
EPA-HQ-OAR-2018-0283 October 26, 2018.
    \572\ NHTSA-2018-0067-11943. FCA Comments on The Safer 
Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-
2026 Passenger Cars and Light Trucks Notice of Proposed Rulemaking.
---------------------------------------------------------------------------

    The agencies agree with the Alliance that the fully developed and 
validated Autonomie model fulfills the agencies' analytical needs for 
full-vehicle modeling and simulation. The agencies also agree that it 
is counterintuitive to have two separate models conducting the same 
work.
    Some commenters stated that broadly, EPA was required to conduct 
its own technical analysis and rely on its own models to do so.\573\ 
Those comments are addressed in Section IV.
---------------------------------------------------------------------------

    \573\ NHTSA-2018-0067-12000; NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    Regarding the merits of EPA's models, and based on previous inputs 
and assumptions used to populate those models, ICCT commented that 
``[b]ased on the ICCT's global analysis of vehicle regulations, the 
EPA's physics-based ALPHA modeling offers the most sophisticated and 
thorough modeling of the applicable technologies that has ever been 
conducted.'' ICCT listed several reasons for this, including that the 
EPA modeling is based on systematic modeling of technologies and their 
synergies; it was built and improved upon by extensive modeling by and 
with Ricardo (an engineering consulting firm); it incorporated National 
Academies input at multiple stages; it has included many peer reviews 
at many stages of the modeling and the associated technical reports 
published by engineers in many technical journal articles and 
conference proceedings; and EPA's Draft TAR analysis, which used ALPHA, 
used state-of-the-art engine maps based on benchmarked high-efficiency 
engines. ICCT concluded

[[Page 24341]]

that ``[d]espite these rigorous advances in vehicle simulation 
modeling, it appears that the agencies have inexplicably abandoned this 
approach, expressly disregarding the EPA benchmarked engines, ALPHA 
modeling, and all its enhancements since the last rulemaking.''
    The hallmarks ICCT lists regarding the ALPHA modeling are equally 
applicable to Autonomie.\574\ Autonomie is also based on systematic 
modeling of technologies and their synergies when combined as packages. 
The U.S. Department of Energy created Autonomie, and over the past two 
decades, helped to develop and mature the processes and inputs used to 
represent real-world vehicles using continuous feedback from the tool's 
worldwide user base of vehicle manufacturers, suppliers, government 
agencies, and other organizations. Moreover, using Autonomie brings the 
agencies closer to the NAS Committee's stated goal of ``full system 
simulation modeling for every important technology pathway and for 
every vehicle class.'' \575\ While the NAS Committee originally thought 
that full vehicle simulation modeling would not be feasible for the 
thousands of vehicles in the analysis fleets because the technologies 
present on the vehicles might differ from the configurations used in 
the simulation modeling,\576\ Argonne has developed a process to 
simulate explicitly every important technology pathway for every 
vehicle class. Moreover, although separate from the Autonomie model 
itself, the Autonomie modeling for this rulemaking incorporated other 
NAS committee recommendations regarding full vehicle simulation inputs 
and input assumptions, including using engine-model-generated maps 
derived from a validated baseline map in which all parameters except 
the new technology of interest are held constant.\577\
---------------------------------------------------------------------------

    \574\ See Theo LeSieg, Ten Apples Up On Top! (1961), at 4-32.
    \575\ 2015 NAS Report at 358.
    \576\ 2015 NAS Report at 359.
    \577\ NAS Recommendation 2.1.
---------------------------------------------------------------------------

    As discussed further below and in VI.C.1 Engine Paths, this is one 
reason why the IAV maps were used instead of the EPA maps, and the 
agencies instead referenced EPA's engine maps to corroborate the 
Autonomie effectiveness results. The IAV maps are engine-model-
generated maps derived from a validated baseline map in which all 
parameters except the new technology of interest are held constant. 
While EPA's engine maps benchmarking specific vehicles' engines 
incorporate multiple technologies, for example including improvements 
in engine friction and reduction in accessory parasitic loads, 
comparisons presented in Section VI.C.1 showed that engine maps 
developed by IAV, while not exactly the same, are representative of 
EPA's engine benchmarking data.
    In addition, both ALPHA and Autonomie have been used to support 
analyses that have been published in technical journal articles and 
conference proceedings, but those analyses differ fundamentally because 
of the nature of the tools. ALPHA was developed as a tool to be used by 
EPA's in-house experts.\578\ As EPA stated in the ALPHA model peer 
review,\579\ ``ALPHA is not intended to be a commercial product or 
supported for wide external usage as a development tool.'' \580\ 
Accordingly, EPA experts have published several peer-reviewed journal 
articles using ALPHA and have presented the results of those papers at 
conference proceedings.\581\
---------------------------------------------------------------------------

    \578\ ALPHA Peer Review, at 4-1.
    \579\ ICCT's comments intimate that ALPHA has been peer reviewed 
at many stages of the modeling; although EPA has published several 
peer-reviewed technical papers, the ALPHA model itself has been 
subject to one peer review. See Peer Review of ALPHA Full Vehicle 
Simulation Model, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
    \580\ ALPHA Peer Review, at 4-2.
    \581\ See, e.g., Dekraker, P., Kargul, J., Moskalik, A., Newman, 
K. et al., ``Fleet-Level Modeling of Real World Factors Influencing 
Greenhouse Gas Emission Simulation in ALPHA,'' SAE Int. J. Fuels 
Lubr. 10(1):2017, doi:10.4271/2017-01-0899.
---------------------------------------------------------------------------

    To explore ICCT's comments on the importance of peer review 
further, it is important to take the actual substantive content of the 
ALPHA peer review into account.\582\ One reviewer raised significant 
questions over the availability of ALPHA documentation, stating 
``[t]here is an overall lack of detail on key technical features that 
are new in the model,'' and ``[w]e were not able to find any 
information on how the model handles component weight changes.'' 
Reviewers also raised questions related to model readiness, stating 
``[a]ccording to the documentation review, ALPHA's stop/start modeling 
appears to be very simplistic.'' Moreover, when running ALPHA 
simulations, the reviewer noted the results ``strongly suggest that the 
model has errors in the underlying equations or coding with respect to 
all of the load reductions.'' Also, one reviewer said the following of 
ALPHA: ``A specific simulation runtime is significantly high, more than 
10 mins. without providing any indication to the user progress made so 
far. A fairly more complicated model such as Autonomie available even 
with enhanced capabilities is significantly faster today.'' \583\
---------------------------------------------------------------------------

    \582\ EPA. ``Peer Review of ALPHA Full Vehicle Simulation 
Model.'' EPA-420-R-16-013. October 2016. https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf. Last accessed Nov 18, 2019.
    \583\ Peer Review of ALPHA Full Vehicle Simulation Model, at C-
4, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
---------------------------------------------------------------------------

    The peer reviewer's assessment of Autonomie as a more complicated 
model with enhanced capabilities is not surprising, given Autonomie's 
history of development. Autonomie is a commercial tool with more than 
275 worldwide organizational users, including vehicle manufacturers, 
suppliers, government agencies, and nonprofit organizations having 
licensed and used Autonomie. Both Autonomie's creators and user base 
unaffiliated with Argonne have published over 100 papers, including 
peer-reviewed papers in journals, related to Autonomie validation and 
other studies.584 585 One could even argue that the tool has 
been continuously peer reviewed by these thousands of experts over the 
past two decades.
---------------------------------------------------------------------------

    \584\ At least 15 peer-reviewed papers authored by ANL experts 
have been referenced throughout this Section, and others can be 
found at SAE International's website, https://www.sae.org/, using 
the search bar for ``Autonomie.''
    \585\ See, e.g., Haupt, T., Henley, G., Card, A., Mazzola, M. et 
al., ``Near Automatic Translation of Autonomie-Based Power Train 
Architectures for Multi-Physics Simulations Using High Performance 
Computing,'' SAE Int. J. Commer. Veh. 10(2):483-488, 2017, https://doi.org/10.4271/2017-01-0267; Samadani, E., Lo, J., Fowler, M., 
Fraser, R. et al., ``Impact of Temperature on the A123 Li-Ion 
Battery Performance and Hybrid Electric Vehicle Range,'' SAE 
Technical Paper 2013-01-1521, 2013, https://doi.org/10.4271/2013-01-1521.
---------------------------------------------------------------------------

    In fact, in responding to a peer review comment on the ALPHA 
model's underlying equations and coding with respect to road load 
reductions, EPA noted that Autonomie had been used as a reference 
system simulation tool to validate ALPHA model results.\586\
---------------------------------------------------------------------------

    \586\ Peer Review of ALPHA Full Vehicle Simulation Model, at 4-
14 and 4-15, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
---------------------------------------------------------------------------

    Outside of formal peer-reviewed studies, Autonomie has been used by 
organizations like ICCT to support policy documents, position briefs, 
and white papers assessing the potential of future efficiency 
technologies to meet potential regulatory requirements,\587\

[[Page 24342]]

just as the agencies did in this rulemaking.
---------------------------------------------------------------------------

    \587\ See, e.g., Oscar Delgado and Nic Lutsey, Advanced Tractor-
Trailer Efficiency Technology Potential in the 2020-2030 Timeframe 
(April 2015), available at https://theicct.org/sites/default/files/publications/ICCT_ATTEST_20150420.pdf; Ben Sharpe, Cost-
Effectiveness of Engine Technologies for a Potential Heavy-Duty 
Vehicle Fuel Efficiency Regulation in India (June 2015), available 
at https://theicct.org/sites/default/files/publications/ICCT_position-brief_HDVenginetech-India_jun2015.pdf; Ben Sharpe and 
Oscar Delgado, Engines and tires as technology areas for efficiency 
improvements for trucks and buses in India (working paper published 
March 2016), available at https://theicct.org/sites/default/files/publications/ICCT_HDV-engines-tires_India_20160314.pdf.
---------------------------------------------------------------------------

    Similarly to ICCT, UCS stated that in contrast to Autonomie, ALPHA 
had been thoroughly peer-reviewed and is constantly being updated to 
reflect the latest technology developments based on work performed by 
the National Vehicle and Fuel Emissions Laboratory.\588\ UCS also 
stated that because EPA has direct control over the model and its 
interface to OMEGA, EPA can better ensure that the inputs into OMEGA 
reflect the most up-to-date data, unlike the Autonomie work, which 
effectively has to be ``locked in'' before it can be deployed in the 
CAFE model. UCS also stated that ALPHA is based on the GEM model (used 
to simulate compliance with heavy-duty vehicle regulations) which was 
been updated with feedback from heavy-duty vehicle manufacturers and 
suppliers, and in fact, ``NHTSA has such confidence in the GEM model 
that they accept its simulation-based results as compliance with the 
heavy-duty fuel economy regulations.''
---------------------------------------------------------------------------

    \588\ NHTSA-2018-0067-12039 (UCS).
---------------------------------------------------------------------------

    Again, the agencies believe that it is important to note that 
Autonomie not only meets, but also exceeds, UCS' listed metrics. 
Autonomie's models, sub-models, and controls are constantly being 
updated to reflect the latest technology developments based on work 
performed by Argonne National Laboratory's Advanced Mobility Technology 
Laboratory (AMTL) (formerly Advanced Powertrain Research Facility, or 
ARPF).589 590 The Autonomie validation has included nine 
validation studies with accompanying reports for software, six 
validation studies and reports for powertrains, nine validation studies 
and reports for advanced components, ten validation studies and reports 
for advanced controls, and overall model validation using test data 
from over 50 vehicles.\591\
---------------------------------------------------------------------------

    \589\ See NPRM PRIA. The agencies cited a succinctly-summarized 
presentation of Autonomie vehicle validation procedures based on 
AMTL test data in the NPRM ANL modeling documentation and PRIA 
docket for stakeholders to review at NHTSA-2018-0067-1972 and NHTSA-
2018-0067-0007.
    \590\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A., 
``Analysis and Model Validation of the Toyota Prius Prime,'' SAE 
2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N, 
Jeong, J., Rousseau, A. & Lohse-Busch, H. ``Control Analysis and 
Thermal Model Development of PHEV,'' SAE 2015-01-1157, SAE World 
Congress, Detroit, April 15; Kim, N., Rousseau, A. & Lohse-Busch, H. 
``Advanced Automatic Transmission Model Validation Using Dynamometer 
Test Data,'' SAE 2014-01-1778, SAE World Congress, Detroit, Apr. 
14.; Lee, D. Rousseau, A. & Rask, E. ``Development and Validation of 
the Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE World 
Congress, Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., & Duoba, 
M. ``Validating Volt PHEV Model with Dynamometer Test Data using 
Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit, Apr. 
13.; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model Validation 
with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040, SAE World 
Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A, Pagerit, S., 
& Sharer, P. ``Plug-in Vehicle Control Strategy--From Global 
Optimization to Real Time Application,'' 22th International Electric 
Vehicle Symposium (EVS22), Yokohama, (October 2006).
    \591\ Rousseau, A. Moawad, A. Kim, Namdoo. ``Vehicle System 
Simulation to Support NHTSA CAFE standards for the Draft Tar.'' 
https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/anl-nhtsa-workshop-vehicle-system-simulation.pdf. Last accessed Nov 20, 2019.
---------------------------------------------------------------------------

    In fact, using Autonomie, which has validated data based on test 
data from over 50 vehicles, alleviates other stakeholder concerns about 
the level of model validation in past analyses. For example, Global 
Automakers expressed concerns about whether the effectiveness values 
used in past EPA analysis, generated from ALPHA full-vehicle model 
simulations, were properly validated, stating that ``[a]lthough EPA 
claims that the LPM was calibrated based on thorough testing and 
modeling with the ALPHA model, the materials provided with the Proposed 
and Final Determination only cover 18 percent of the projected vehicle 
fleet with regards to specific combinations of powertrain technology 
presented by EPA in the MY 2025 OMEGA pathway. It is unclear how EPA 
calibrated the LPM for the remaining 82 percent of the projected 
vehicles. EPA's failure to publicly share the data for such a large 
percentage of vehicles raises questions about the quality of data.'' 
\592\ While simple modeled parameters like single dimensional linear 
systems, such as engine dynamometer torque measurements can be 
validated through other models,\593\ full vehicle systems are complex 
multi-dimensional non-linear systems that need to be developed with 
multiple data sets, and validated with other fully independent data 
sets. Autonomie's models and sub-models have undergone extensive 
validation that has proven the models' agreement with empirical data 
and the principles of physics.
---------------------------------------------------------------------------

    \592\ Docket ID EPA-HQ-OAR-2015-0827-9728. Global later repeated 
that ``only 18% of all vehicle data used as inputs to the ALPHA 
modeling was made available in the EPA's public sources. Additional 
data had to be specifically requested subsequent to the publication 
of the Draft TAR and Proposed Determination. This lack of publicly 
available data highlights transparency concerns, which Global 
Automakers has raised on several previous occasions.''
    \593\ Section 89.307 Dynamometer calibration.
---------------------------------------------------------------------------

    In addition, the agencies disagree with UCS' comment that EPA's 
direct control over its effectiveness modeling and interface to OMEGA 
results in a more up-to-date analysis. Argonne's participation in 
developing inputs for the rulemaking analysis allowed the agencies 
access to vehicle benchmarking data from more vehicles than if the 
agencies were limited by their own resources, and access to the Argonne 
staff's extensive experience based on direct coordination with vehicle 
manufacturers, suppliers, and researchers that all actively use 
Autonomie for their own work. In addition to Autonomie's continuous 
updates to incorporate the latest fuel-economy-improving technologies, 
discussed throughout this section, the data supplied to and generated 
by Autonomie for use in the CAFE model was continuously updated during 
the analysis process. This is just one part of the iterative quality 
assurance (QA) and quality check (QC) process that the agencies 
developed when Argonne's large-scale simulation modeling based in 
Autonomie was first used for the Draft TAR.
    In addition to Argonne's team constantly updating Autonomie, 
Argonne's use of high performance computing (HPC) allowed for constant 
update of the analysis during the rulemaking process. Argonne's HPC 
platform allows a full set of simulations--over 750,000 modeled 
vehicles that incorporate over 50 different fuel-economy-improving 
technologies--to be simulated in one week. Subsets of the simulations 
can be re-run should issues come up during QA/QC in a day or less. 
Tools like the internet and high performance computers have allowed the 
agencies to evaluate technology effectiveness with up-to-date inputs 
without the proximity of the computers and the people running them 
working as a detriment the analysis.
    Finally, GEM, ALPHA, and Autonomie were all developed in the MATLAB 
computational environment as forward-looking physics-based vehicle 
models. Just as ALPHA has roots in GEM, created in 2010 to accompany 
the agencies' heavy-duty vehicle CO2 emissions and fuel 
consumption standards, Autonomie has its origins in the software PSAT, 
developed over 20 years ago. While this information is useful, as 
implied by UCS' comment, the origin of the software was less important 
than the capabilities the software could provide for today's analysis. 
NHTSA's acceptance of GEM

[[Page 24343]]

results for compliance with heavy-duty fuel economy regulations had no 
bearing on the decision to use Autonomie to assess the effectiveness of 
light-duty fuel economy and CO2 improving technologies. GEM 
was developed to serve as the compliance model for heavy-duty 
vehicles,\594\ and GEM serves that limited scope very well.
---------------------------------------------------------------------------

    \594\ Newman, K., Dekraker, P., Zhang, H., Sanchez, J. et al., 
``Development of Greenhouse Gas Emissions Model (GEM) for Heavy- and 
Medium-Duty Vehicle Compliance,'' SAE Int. J. Commer. Veh. 
8(2):2015, doi:10.4271/2015-01-2771.
---------------------------------------------------------------------------

    UCS did comment that full vehicle simulation could significantly 
improve the estimates of technology effectiveness, but thought it 
critical that the process be as open and transparent as possible. UCS 
pointed to ALPHA results published in peer-reviewed journals as an 
example of how transparency has provided the ALPHA modeling effort with 
significant and valuable feedback, and contrasted what they 
characterized as Autonomie's ``black box'' approach, which they stated 
``does not lend itself to similar dialog, nor does it make it easy to 
assess the validity of the results.'' Specifically, UCS stated that it 
is ``impossible to verify, replicate, or alter the work done by 
Autonomie due to the expensive nature of the tools used and lack of 
open source or peer-reviewed output.'' In contrast, UCS stated that 
EPA's ALPHA model has been thoroughly peer reviewed, and is readily 
``downloadable, editable, and accessible to anyone with a MATLAB 
license.''
    The agencies responses on the merits of how ALPHA and Autonomie 
were peer-reviewed are discussed above. Regarding UCS' comment that it 
is impossible to verify, replicate, or alter the work done by 
Autonomie, the agencies disagree. All inputs, assumptions, model 
documentation--including of component models and individual control 
algorithms--and outputs for the NPRM Autonomie modeling were submitted 
to the docket for review.\595\ Commenters were able to provide a robust 
analysis of Autonomie's technology effectiveness inputs, input 
assumptions, and outputs, as shown by their comments on specific 
vehicle technology effectiveness assumptions, discussed throughout this 
section and in the individual technology sections below.
---------------------------------------------------------------------------

    \595\ NHTSA-2018-0067-1855. ANL Autonomie Compact Car Vehicle 
Class Results. Aug 21, 2018. NHTSA-2018-0067-1856. ANL Autonomie 
Performance Compact Car Vehicle Class Results. Aug 21, 2018. NHTSA-
2018-0067-1494. ANL Autonomie Midsize Car Vehicle Class Results. Aug 
21, 2018. NHTSA-2018-0067-1487. ANL Autonomie Performance Pick-Up 
Truck Vehicle Class Results. Aug 21, 2018. NHTSA-2018-0067-1663. ANL 
Autonomie Performance Midsize Car Vehicle Class Results. Aug 21, 
2018. NHTSA-2018-0067-1486. ANL Autonomie Small SUV Vehicle Class 
Results. Aug 21, 2018 NHTSA-2018-0067-1662. ANL Autonomie 
Performance Midsize SUV Vehicle Class Results. Aug 21, 2018. NHTSA-
2018-0067-1661. ANL Autonomie Pickup Truck Vehicle Class Results. 
Aug 21, 2018. NHTSA-2018-0067-1485. ANL Autonomie Small Performance 
SUV Vehicle Class Results. Aug 21, 2018 NHTSA-2018-0067-1492. ANL 
Autonomie Midsize SUV Vehicle Class Results. Aug. 21, 2018. NHTSA-
2018-0067-0005. ANL Autonomie Model Assumptions Summary. Aug 21, 
2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main Component 
Assumptions. Aug 21, 2018. NHTSA-2018-0067-0007. Islam, E. S, 
Moawad, A., Kim, N, Rousseau, A. ``A Detailed Vehicle Simulation 
Process To Support CAFE Standards 04262018--Report'' ANL Autonomie 
Documentation. Aug 21, 2018. NHTSA-2018-0067-0004. ANL Autonomie 
Data Dictionary. Aug 21, 2018. NHTSA-2018-0067-1692. ANL BatPac 
Model 12 55. Aug 21, 2018. NHTSA-2018-0067-12299. Preliminary 
Regulatory Impact Analysis (July 2018). Posted July 2018 and updated 
August 23 and October 16, 2018.
---------------------------------------------------------------------------

    The agencies also disagree with UCS' assessment of Autonomie as 
``expensive.'' While Autonomie is a commercial product, the biggest 
financial barrier to entry for both ALPHA and Autonomie is the same: A 
MathWorks license.596 597 Regardless, Argonne has made the 
version of Autonomie used for this final rule analysis available upon 
request, including the individual runs used to generate each technology 
effectiveness estimate.\598\
---------------------------------------------------------------------------

    \596\ Autonomie. Frequently Asked Questions. ``Which version of 
matlab can I use?'' https://www.autonomie.net/faq.html#faq2. Last 
accessed Nov. 19, 2019.
    \597\ EPA ALPHA v2.2 Technology Walk Samples. ``Running this 
version of ALPHA requires Matlab/Simulink with StateFlow 2016b.'' 
https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha.
    \598\ Argonne Nationally Laboratory. Autonomie License 
Information. https://www.autonomie.net/asp/LicenseRequest.aspx. Last 
accessed Nov, 18, 2019.
---------------------------------------------------------------------------

    Next, ICCT supplanted its statement that the agencies 
``inexplicably'' abandoned ALPHA, commenting that the agencies' 
explanation and justification for relying on Autonomie rather than 
ALPHA failed to discuss ALPHA in detail, and the agencies did not 
compare and contrast the two models. ICCT continued, ``the EPA cannot 
select its modeling tool arbitrarily, yet it appeared that the EPA has 
whimsically shifted from an extremely well-vetted, up-to-date, 
industry-grade modeling tool to a less-vetted, academic-grade framework 
with outdated inputs without even attempt to scrutinize the change.'' 
ICCT also stated that the agencies are legally obligated to acknowledge 
and explain when they change position, and ``cannot simply ignore that 
EPA previously concluded that the ALPHA modeling accurately projected 
real-world effects of technologies and technology packages.''
    The agencies disagree that a more in-depth discussion of ALPHA was 
required in the NPRM. In acknowledging the transition to using 
Autonomie for effectiveness modeling and the CAFE model for analysis of 
regulatory alternatives,\599\ the agencies described several analytical 
needs that using a single analysis from the CAFE model--with inputs 
from the Autonomie tool--addressed. These included that Autonomie 
produced realistic estimates of fuel economy levels and CO2 
emission rates through consideration of real-world constraints, such as 
the estimation and consideration of performance, utility, and 
drivability metrics (e.g., towing capability, shift busyness, frequency 
of engine on/off transitions).\600\ That EPA previously concluded the 
ALPHA modeling accurately projected real-world effects of technologies 
and technology packages has no bearing on Autonomie's ability to 
fulfill the analytical needs that the agencies articulated in the NPRM, 
including that Autonomie also accurately projects real-world effects of 
technologies and technology packages.
---------------------------------------------------------------------------

    \599\ 83 FR 43000 (Aug. 24, 2018).
    \600\ 83 FR 43001 (Aug. 24, 2018).
---------------------------------------------------------------------------

    The agencies also disagree with ICCT's characterization of ALPHA as 
``an extremely well-vetted, up-to-date, industry-grade modeling tool'' 
and Autonomie as a ``less-vetted, academic-grade framework with 
outdated inputs.'' Again, Autonomie has been used by government 
agencies, vehicle manufacturers (and by agencies and manufacturers 
together in the collaborative government-industry partnership U.S. 
DRIVE program), suppliers, and other organizations because of its 
ability to simulate many powertrain configurations, component 
technologies, and vehicle-level controls over numerous drive cycles. 
Characterizing ALPHA as an ``industry-grade modeling tool'' contravenes 
EPA's own description of its tool--an in-house vehicle simulation model 
used by EPA, not intended to be a commercial product.\601\
---------------------------------------------------------------------------

    \601\ See, e.g., Overview of ALPHA Model, https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha; ALPHA Effectiveness Modeling: 
Current and Future Light-Duty Vehicle & Powertrain Technologies 
(Jan. 20, 2016), available at https://www.epa.gov/sites/production/files/2016-10/documents/alpha-model-sae-govt-ind-mtg-2016-01-20.pdf 
(``ALPHA is not a commercial product (e.g. there are no user 
manuals, tech support hotlines, graphical user interfaces, or full 
libraries of components).''). See also Peer Review of ALPHA Full 
Vehicle Simulation Model, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf. While ALPHA peer reviewers found the 
model to be a ``fairly simple transparent model . . . [t]he model 
execution requires an expert MatLab/Simulink user since no user-
friendly interface currently exists.'' Indeed, EPA noted in response 
to this comment that ``[a]s with any internal tool, EPA does not 
have the need for a ``user-friendly interface'' like one that would 
normally accompany a commercial product which is available for 
purchase and fully supported for wide external usage.''

---------------------------------------------------------------------------

[[Page 24344]]

    That characterization also contravenes documentation from the 
automotive industry indicating that manufacturers consider ALPHA to 
generate overly optimistic effectiveness values, to be unrepresentative 
of real-world constraints, and a difficult tool to 
use.602 603 The Alliance commented to the MTE 
reconsideration that ``[p]revious comments from the Alliance and 
individual manufacturers to the MTE docket have highlighted multiple 
concerns with EPA's ALPHA model. Many of these concerns remain 
unresolved.'' \604\ Furthermore, the Alliance commented that ALPHA 
``has not been documented with any instructions making it difficult for 
users outside of EPA to run and interpret the model.'' \605\ Global 
Automakers further stated that the ``lack of publicly available data 
[related to inputs used in the ALPHA modeling] highlights transparency 
concerns, which Global Automakers has raised on several previous 
occasions.'' \606\ In fact, both the Alliance of Automobile 
Manufacturers and Global Automakers, the two trade organizations that 
represent the automotive industry, concluded that Autonomie should be 
used to generate effectiveness inputs for the CAFE model.\607\
---------------------------------------------------------------------------

    \602\ See EPA-HQ-OAR-2015-0827-10125, at 7. As part of their 
assessment that known technologies could not meet the original MY 
2022-2025 standards, Toyota noted that the ALPHA conversion of 
Toyota's MY 2015 to MY 2025 performance ``appears to yield overly 
optimistic results because the powertrain efficiency curves 
represent best-case targets and not the average vehicle, the imposed 
performance constraints are unmarketable, and the generated credits 
are out of sync with product cadence and design cycles.'' See also 
NHTSA-2018-0067-12431, at 7. More recently, Toyota stated in their 
comments to the NPRM that ``Toyota's position [on the efficacy of 
the OMEGA and LPM models] has been clearly represented by comments 
previously submitted by the Alliance of Automobile Manufacturers, 
Global Automakers, and Novation Analytics. Those comments identify 
the LPM and OMEGA models as sources of inaccuracy in EPA technology 
evaluations and provide suggested improvements. Neither model is 
transparent, intuitive, or user friendly.''
    \603\ EPA-HQ-OAR-2015-0827-9194.
    \604\ EPA-HQ-OAR-2015-0827-9194, at 33.
    \605\ EPA-HQ-OAR-2015-0827-9194.
    \606\ EPA-HQ-OAR-2015-0827-9728.
    \607\ EPA-HQ-OAR-2015-0827-9163 at 5. (``EPA should abandon the 
lumped-parameter model and instead use NHTSA's Autonomie and Volpe 
models to support the Revised Final Determination.''). See also EPA-
HQ-OAR-2015-0827-9728 at 15 (stating the EPA's engine mapping and 
tear down analyses ``should be integrated into the Autonomie model, 
which then feeds into the Volpe modeling process.''); EPA-HQ-OAR-
2015-0827-9194 at 33.
---------------------------------------------------------------------------

    In addition, Autonomie contains up-to-date sub-models to represent 
the latest electrification and advanced transmission and advanced 
engine technologies. As summarized by the Alliance, ``Autonomie was 
developed from the start to address the complex task of combining 2 
power sources in a hybrid powertrain.'' \608\ Autonomie has 
continuously improved over the years by adopting new technologies into 
its modeling framework. Even a small sampling of SAE papers shows how 
Autonomie has been validated to simulate the latest fuel-economy-
improving technologies like hybrid vehicles and PHEVs.\609\
---------------------------------------------------------------------------

    \608\ Alliance, Docket ID NHTSA-2018-0067-12073 at 135.
    \609\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A., 
``Analysis and Model Validation of the Toyota Prius Prime,'' SAE 
2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N, 
Jeong, J. Rousseau, A. & Lohse-Busch, H. ``Control Analysis and 
Thermal Model Development of PHEV,'' SAE 2015-01-1157.
---------------------------------------------------------------------------

    Moreover, Autonomie effectively considers other real-world 
constraints faced by the automotive industry. Vehicle manufacturers and 
suppliers spend significant time and effort to ensure technologies are 
incorporated into vehicles in ways that will balance consumer 
acceptance for attributes such as driving quality,\610\ noise-
vibration-harshness (NVH), and meeting other regulatory mandates, like 
EPA's and CARB's On-Board Diagnostics (OBD) requirements,\611\ and 
EPA's and CARB's criteria exhaust emissions standards.\612\ The 
implementation of new fuel economy improving technologies have at times 
raised consumer acceptance issues.\613\ As discussed earlier, there are 
diminishing returns for modeling every vehicle attribute and tradeoff, 
as each takes time and incurs cost; however, Autonomie sub-models are 
designed to account for a number of the key attributes and tradeoffs, 
so the resulting effectiveness estimates reflect these real world 
constraints.
---------------------------------------------------------------------------

    \610\ An example of a design requirement is accommodating the 
``lag'' in torque delivery due to the spooling of a turbine in a 
turbocharged downsized engine. This affects real-world vehicle 
performance, as well as the vehicle's ability to shift during normal 
driving and test cycles.
    \611\ EPA adopted and incorporated by reference current OBD 
regulations by the California ARB, effective for MY 2017, that cover 
all vehicles except those in the heavier fraction of the heavy-duty 
vehicle class.
    \612\ Tier 3 emission standards for light-duty vehicles were 
proposed in March 2013 78 FR 29815 (May 21, 2013) and signed into 
law on March 3, 2014 79 FR 23413 (June 27, 2014). The Tier 3 
standards--closely aligned with California LEV III standards--are 
phased-in over the period from MY2017 through MY2025. The regulation 
also tightens sulfur limits for gasoline.
    \613\ Atiyeh, C. ``What you need to know about Ford's PowerShift 
Transmission Problems'' Car and Driver. July 11, 2019. https://www.caranddriver.com/news/a27438193/ford-powershift-transmission-problems/.
---------------------------------------------------------------------------

    Furthermore, aside from the fact that Autonomie represents the 
structural state-of-the-art in full-vehicle modeling and simulation, 
Autonomie can be populated with any inputs that could be populated in 
the ALPHA model.\614\ The agencies chose to use specific inputs for 
this rulemaking because, as discussed further in Sections VI.C below, 
they best represent the technologies that manufacturers could 
incorporate in the rulemaking timeframe, in a way that balanced 
important concerns like consumer acceptance. Some other examples of how 
Autonomie inputs have been updated with the latest vehicle technology 
data specifically for this analysis include test data incorporated from 
both Argonne and NHTSA-sponsored vehicle benchmarking, including an 
updated automatic transmission skip-shifting feature,\615\ additional 
application of cylinder deactivation for turbocharged downsized 
engines, and as discussed above, new modeling and simulation that 
includes variable compression ratio and Miller Cycle engines.
---------------------------------------------------------------------------

    \614\ For example, Autonomie used the HCR1 and HCR2 engine maps 
used as inputs to ALHPA in the Draft TAR and Proposed Determination.
    \615\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford F-
150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812 520.
---------------------------------------------------------------------------

    Finally, ICCT commented that the agencies must conduct a systematic 
comparison of the Autonomie modeling system and ALPHA modeling in 
several respects, including the differences in technical inputs and 
resulting efficiency estimates, to explain how the choice of model 
altered the regulatory technology penetration and compliance cost 
estimations, and the differences in modeling methodologies, including 
regarding the relative level of experience of the teams conducting the 
effectiveness modeling, to demonstrate that the choice to use Autonomie 
was not ``due to convenience and easier access by the NHTSA research 
team, rather than for any technical improvement.'' ICCT stated that 
without performing this comparison, ``it otherwise appears that the 
agencies switched from a better-vetted model and system of inputs with 
more recent input data to a less-vetted model and system of inputs as a 
way to bury many dozens of changes without transparency or expert 
assessment (as illustrated in the

[[Page 24345]]

above errors and invalidated data on individual technologies).'' Each 
issue is discussed below in turn.
    First, regarding technical inputs, technology pathways, and 
resulting outputs, ICCT stated that the agencies must compare (1) 
whether the models have been routinely strengthened by incorporating 
cutting edge 2020-2025 automotive technologies to ensure they reflect 
the available improvements; (2) every efficiency technology in the 2016 
Draft TAR and original EPA TSD and Proposed and Final Determination 
analysis against the NPRM; (3) all the major technology package 
pathways (i.e., all combinations with high uptake in the Adopted and 
Augural 2025 standards) in the current NPRM versus the 2016 Draft TAR 
and the 2016 TSD and original Final Determination analysis; (4) each of 
the major 2025 technology package synergies; (5) the modeling work of 
EPA's, Ricardo's, and Argonne's 2014-2018 model year engine 
benchmarking and modeling of top engine and transmission models; and 
``defend why they appear to have chosen to dismiss the superior and 
better vetted technology modeling approach.''
    ICCT stated that the agencies must make these comparisons because, 
``[o]therwise, it seems obvious that the agencies have subjectively 
decided to use the modeling that increases the modeled cost, providing 
further evidence of a high degree of bias without an objective 
accounting of the methodological differences and the sensitivity of the 
results to their new decision.'' Moreover, ICCT stated that ``[b]ecause 
ALPHA is the dominant, preferred, and better-vetted modeling and was 
used in the original Proposed and Final Determination, the agencies are 
responsible for assessing and describing how the use of the ALPHA 
modeling would result in a different regulatory result for their 
analysis of the 2017-2025 adopted [CO2] and Augural CAFE 
standards.''
    The agencies do not believe that it is necessary to conduct a 
retrospective comparison of ALPHA/LPM and Autonomie effectiveness for 
every technology in the Draft TAR and Proposed Determination to the 
NPRM and final rule analyses, between the two models for technologies 
and packages used in the NPRM and final rule analysis, or to explain 
where and why Autonomie provided different results from ALPHA and the 
LPM, to assess and describe how the use of the ALPHA modeling would 
result in a different regulatory result of CAFE and CO2 
standards, per ICCT's request. While it is anticipated that different 
values will be produced using different tools in an analysis, it is not 
appropriate to select the tool for use based on preferred results. The 
selection of an analysis tool should be based on an evaluation of the 
tool's capabilities and appropriateness for the analysis task. The 
analysis tool should support the full extent of the analysis and 
support the level of input and output resolution required. To compare 
the output of the two models for the purpose of selecting a tool for 
the analysis would likely be biased and disingenuous to the purpose of 
the analysis. In this case, Autonomie was selected for this analysis 
for the reasons discussed throughout this section, and accordingly the 
agencies believe that it was reasonable to consider effectiveness 
estimates developed with Autonomie.
    That said, comparison of how the tools behave is discussed here to 
further support the agencies' decision process. To demonstrate, in 
addition to everything discussed previously in this section, 
differences in how each model handles powertrain systems modeling with 
specific examples are discussed below as a reference, and differences 
between the agencies' approaches to effectiveness modeling for specific 
technologies is discussed in Section VI.C where appropriate. While the 
improved approach to estimating technology effectiveness estimates 
certainly impacted the regulatory technology penetration, compliance 
cost estimates, and ``major 2025 technology packages and synergies,'' 
how technologies are applied in the compliance modeling and the 
associated costs of the technologies is equally as important to 
consider when examining factors that might impact the regulatory 
analysis; that consideration goes beyond the scope of simply 
considering which full vehicle simulation model better performs the 
functions required of this analysis.
    The agencies have discussed updates to the technologies considered 
in the Autonomie modeling throughout this section, in addition to 
Autonomie's models and sub-models that control advanced technologies 
like hybrid and electrified powertrains. Autonomie's explicit models, 
sub-models, and controls for hybrid and electric vehicles have been 
continuously validated over the past several years,\616\ as Autonomie 
was developed from the beginning to address the complex task of 
combining two power sources in a hybrid powertrain.
---------------------------------------------------------------------------

    \616\ Karbowski, D., Kwon, J., Kim, N., & Rousseau, A., 
``Instantaneously Optimized Controller for a Multimode Hybrid 
Electric Vehicle,'' SAE paper 2010-01-0816, SAE World Congress, 
Detroit, April 2010; Sharer, P., Rousseau, A., Karbowski, D., & 
Pagerit, S. ``Plug-in Hybrid Electric Vehicle Control Strategy--
Comparison between EV and Charge-Depleting Options,'' SAE paper 
2008-01-0460, SAE World Congress, Detroit (April 2008); and 
Rousseau, A., Shidore, N., Carlson, R., & Karbowski, D. ``Impact of 
Battery Characteristics on PHEV Fuel Economy,'' AABC08; Jeong, J., 
Kim, N., Stutenberg, K., Rousseau, A., ``Analysis and Model 
Validation of the Toyota Prius Prime,'' SAE 2019-01-0369, SAE World 
Congress, Detroit, April 2019; Kim, N, Jeong, J. Rousseau, A. & 
Lohse-Busch, H. ``Control Analysis and Thermal Model Development of 
PHEV,'' SAE 2015-01-1157, SAE World Congress, Detroit, April 15; 
Lee, D. Rousseau, A. & Rask, E. ``Development and Validation of the 
Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE World Congress, 
Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., & Duoba, M. 
``Validating Volt PHEV Model with Dynamometer Test Data using 
Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit, Apr. 
13.; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model Validation 
with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040, SAE World 
Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A, Pagerit, S., 
& Sharer, P. ``Plug-in Vehicle Control Strategy--From Global 
Optimization to Real Time Application,'' 22th International Electric 
Vehicle Symposium (EVS22), Yokohama, (October 2006).
---------------------------------------------------------------------------

    Also regarding the inputs to both models, as highlighted in Section 
VI.C.3.a), and discussed above, inputs and assumptions for the ALPHA 
modeling used for the EPA Draft TAR and Proposed Determination analysis 
were projected from benchmarking testing. While it is straightforward 
to measure engine fuel consumption and create an engine fuel map, it is 
extremely challenging to identify the specific technologies and levels 
of technologies present on a benchmarking engine. Attributing changes 
in the overall engine fuel consumption to the individual engine 
technologies that make up the complete engine involves significant 
uncertainty.
    The fixed-point model approach used by the ALPHA model does not 
develop an effectiveness function and assigns a single value to a 
technology. The single value is derived from benchmark testing, which 
often does not isolate the effect of a single technology from the 
effects of other technologies on the tested vehicle. To isolate a 
single technology's effect for use in fixed point modeling properly, 
the agencies would need to benchmark multiple versions of a single 
vehicle, carefully controlling changes to the vehicles' fuel efficiency 
technologies. This process would need to be repeated for a large 
portion of the vehicle fleet and would require significant funding and 
thousands of lab hours to complete. Without this level of data, fixed-
point effectiveness estimates tend to be too high, as they are unable 
to account for synergetic effects of multiple technologies. 
Specifically, when EPA benchmarks vehicles like the 2018 Toyota Camry, 
the resulting fuel map captures the benefits of many

[[Page 24346]]

technologies associated with that engine. This data can be helpful when 
developing controls and validating component operations in modeling, 
but it is inaccurate to conclude that fuel consumption is directly 
related to individual engine technologies, such as lubrication and 
friction reduction, and geometric improvements in efficiency.
    Contrasted, the NPRM and final rule Autonomie analyses selected 
specific base engine maps and applied technologies incrementally, both 
individually and in known combinations, to better isolate the impacts 
of the technologies. As discussed above, this also implemented NAS 
Recommendation 2.1, to use engine-model-generated maps in the full 
vehicle simulations derived from a validated baseline map in which all 
parameters except the new technology of interest are held 
constant.\617\ While the different methods are valid for different 
purposes, the method selected for the analysis presented today was more 
useful for measuring the incremental effectiveness increments as 
opposed to the absolute values of technology effectiveness, e.g., that 
could be measured by benchmarking a technology package.
---------------------------------------------------------------------------

    \617\ 2015 NAS Report at p. 82.
---------------------------------------------------------------------------

    To provide an example of another difference in behavior between the 
simulation tools, a comparison between ALPHA and Autonomie 
transmissions shifting behavior was conducted. The comparison 
highlighted the differences in how each simulation tool approaches 
transmission shift logic. The ALPHA simulation tool used ALPHAShift. 
ALPHAShift is an optimization algorithm that uses numerous vehicle 
characteristics to find a best shifting strategy. The primary inputs 
for the algorithm includes the fuel consumption (or cost) map for the 
vehicle engine.\618\ Although a public version of ALPHA is available 
for evaluation, the ALPHAShift algorithm used by the tool is hard coded 
with fixed values.619 620 This is an issue, because despite 
peer reviewed documentation on how to tune the algorithm,\621\ no 
documentation of how the algorithm logic works is available for review. 
This is confounding for the use of the software, particularly when the 
observed behavior of the model departs from expected behavior. Figure 
VI-6 below shows simulated gear shift (left) versus actual gear shift 
(right), demonstrating an unexpected shift to neutral before shifting 
to the requested gear.
---------------------------------------------------------------------------

    \618\ Newman, K., Kargul, J., and Barba, D., ``Development and 
Testing of an Automatic Transmission Shift Schedule Algorithm for 
Vehicle Simulation,'' SAE Int. J. Engines 8(3):2015, doi:10.4271/
2015-01-1142.
    \619\ Aymeric, R. Islam, E. S. ``Analysis of EPA's ALPHA Shift 
Model--ALPHAShift.'' ANL. March 9, 2020.
    \620\ ALPHA v2.2 Technology Walk Samples. EPA. January 2017. 
https://www.epa.gov/sites/production/files/2017-01/alpha-20170112.zip. Last Accessed March 9, 2020.
    \621\ Newman, K., Kargul, J., and Barba, D., ``Development and 
Testing of an Automatic Transmission Shift Schedule Algorithm for 
Vehicle Simulation,'' SAE Int. J. Engines 8(3):2015, doi:10.4271/
2015-01-1142.
---------------------------------------------------------------------------

    By contrast, and discussed further in VI.C.2 Transmission Paths, 
Autonomie uses a fully documented algorithm to develop a best shifting 
strategy for each unique vehicle configuration. The algorithm develops 
shifting strategies unique to each individual vehicle based on gear 
ratio, final drive ratio, engine BSFC and other vehicle 
characteristics. This is one example of model behavior, in addition to 
the availability of more transparency on this behavior for greater 
stakeholder review, that led the agencies to determine it was 
reasonable and appropriate to use Autonomie for this analysis.
---------------------------------------------------------------------------

    \622\ ALPHA v2.2 Technology Walk Samples. Jan. 12, 2017. https://www.epa.gov/sites/production/files/2017-01/alpha-20170112.zip. Last 
accessed Dec 9, 2019.
[GRAPHIC] [TIFF OMITTED] TR30AP20.114

    Regarding the technical expertise of the team conducting the 
---------------------------------------------------------------------------
effectiveness modeling, ICCT commented:

    [T]he agencies should also disclose how much commercial business 
is conducted by the Ricardo, IAV, and Argonne Autonomie teams that 
underpin the modeling of EPA and NHTSA, respectively, including how 
much related research they have done for auto industry clients over 
the past ten years. We mention this because we strongly suspect that 
Ricardo, upon which EPA built its ALPHA model, has done at least an 
order of magnitude (in number of projects, person-hours, and budget) 
more work with and for the automotive industry than the IAV and 
Autonomie teams have in direct work for

[[Page 24347]]

automotive industry clients. A conventional government procurement 
effort that competitively vets potential research expert teams would 
presumably have selected for such automotive industry credentials 
and experience, yet it appears that the agencies are wholly 
deferring to Autonomie's less rigorous research-grade modeling 
framework and data due to convenience and easier access by the NHTSA 
research team, rather than for any technical improvement, and this 
is to the detriment of showing clear understanding of real-world 
automotive engineering developments (as demonstrated by many 
erroneous technology combination results throughout these comments).

    First, NHTSA follows Federal Acquisition Regulation (FAR) to award 
contracts and Interagency Agreements (IAAs),\623\ and any awarded 
contracts and IAAs must follow the FAR requirements. Importantly, FAR 
3.101-1 includes key aspects of conduct and ethics that NHTSA must 
follow in awarding a contract or IAA:
---------------------------------------------------------------------------

    \623\ Federal Acquisition Regulation (FAR). https://www.acquisition.gov/.

    Government business shall be conducted in a manner above 
reproach and, except as authorized by statute or regulation, with 
complete impartiality and with preferential treatment for none. 
Transactions relating to the expenditure of public funds require the 
highest degree of public trust and an impeccable standard of 
conduct. The general rule is to avoid strictly any conflict of 
interest or even the appearance of a conflict of interest in 
Government-contractor relationships. While many Federal laws and 
regulations place restrictions on the actions of Government 
personnel, their official conduct must, in addition, be such that 
they would have no reluctance to make a full public disclosure of 
their actions.\624\
---------------------------------------------------------------------------

    \624\ FAR 3.101-1.

    While some factors are more relevant than others in considering 
whether to award a contract or enter into an IAA, the amount of work 
that an organization has performed, characterized by projects, person-
hours, and budget, is only one of a multitude of factors that is 
considered (if it is even considered at all--an agency might not 
request this information and an organization might decline to provide 
it because of contractual clauses or to protect commercial business 
interests) when assessing whether an organization meets the agency's 
needs for a specific task. Other factors, such as the federal budget, 
also set boundaries for the scope of work that can be performed under 
any competitive government procurement effort.
    As discussed throughout this section, the team at Argonne National 
Laboratory behind Autonomie has developed and refined a state-of-the-
art tool that is used by the automotive industry, government agencies, 
and research or other nongovernmental institutions around the world. 
The tool has been and continues to be validated to production vehicles, 
and updated to include models, sub-models, and controls representing 
the state-of-the-art in fuel economy improving technology. To the 
extent that ICCT believes that ``research done for auto industry 
clients,'' ``work with and for the automotive industry,'' and 
``automotive industry credentials and experience,'' are metrics upon 
which to base this type of important decision, the agencies point ICCT 
to the statements from the automotive industry, above, recommending 
Autonomie be used for technology effectiveness modeling.
    ICCT concluded that ``[w]hile the agencies are in their process of 
conducting a proper vetting of their NPRM's foundational Autonomie-
based modeling, we recommend that they rely on what appears to be the 
superior and better vetted technology modeling approach with more 
thorough and state-of-the-art advanced powertrain systems modeling and 
engine maps from the EPA ALPHA modeling.''
    The agencies properly vetted the Autonomie modeling and decided 
that Autonomie represented a reasonable and appropriate tool to provide 
technology effectiveness estimates for this rulemaking. To the extent 
that commenters' concerns were more about the effectiveness results 
than the tools used to model technology effectiveness, modeling updates 
detailed in the Section VI.B.3.c), below, address those comments. While 
some commenters may still be dissatisfied with Autonomie's technology 
effectiveness estimates, the agencies believe that the refinement of 
inputs and input assumptions, and associated explanation of why those 
refinements are appropriate and reasonable, have appropriately 
addressed comments on these issues. Importantly, none of these 
refinements have led either agency to reconsider using Autonomie for 
this rulemaking analysis.
    Additional discussion of the agencies' decision to rely on one set 
of modeling tools for this rulemaking is located in Section VI.A of 
this preamble.
c) Technology Effectiveness Values Implementation in the CAFE Model
    While the Autonomie model produces a large amount of information 
about each simulation run--for a single technology combination, in a 
single technology class--the CAFE model only uses two elements of that 
information: Battery costs and fuel consumption on the city and highway 
cycles. The agencies combine the fuel economy information from the two 
cycles to produce a composite fuel economy for each vehicle, on each 
fuel. Plug-in hybrids, being the only dual-fuel vehicles in the 
Autonomie simulation, require efficiency estimates of operation on both 
gasoline and electricity--as well as an estimate of the utility factor, 
or the number of miles driven on each fuel. The fuel economy 
information for each technology combination, for each technology class, 
is converted into a single number for use in the CAFE model.
    As described in greater detail below, each Autonomie simulation 
record represents a unique combination of technologies, and the 
agencies create a technology ``key'' or technology state vector that 
describes all the technology content associated with a record. The 2-
cycle fuel economy of each combination is converted into fuel 
consumption (gallons per mile) and then normalized relative to the 
starting point for the simulations. In each technology class, the 
combination with the lowest technology content is the VVT (only) 
engine, with a 5-speed transmission, no electrification, and no body-
level improvements (mass reduction, aerodynamic improvements, or low 
rolling resistance tires). This is the reference point (for each 
technology class) for all the effectiveness estimates in the CAFE 
model. The improvement factors that the model uses are a given 
combination's fuel consumption improvement relative to the reference 
vehicle in its technology class.
    For the majority of the technologies analyzed within the CAFE 
Model, the fuel economy improvements were derived from the database of 
Autonomie's detailed full-vehicle modeling and simulation results. In 
addition to the technologies found in the Autonomie simulation 
database, the CAFE modeling system also incorporated a handful of 
technologies that were required for CAFE modeling, but were not 
explicitly simulated in Autonomie. The total effectiveness of these 
technologies either could not be captured on the 2-cycle test, or there 
was no robust data that could be used as an input to the full-vehicle 
modeling and simulation, like with emerging technologies such as 
advanced cylinder deactivation (ADEAC). These additional technologies 
are discussed further in Sections VI.B.3 Technology Effectiveness and 
individual technologies sections. For calculating fuel economy 
improvements attributable to these additional technologies, the model 
used defined fuel consumption improvement factors that are constant

[[Page 24348]]

across all technology combinations in the database and scale 
multiplicatively when applied together. The Autonomie-simulated and 
additional technologies were then externally combined, forming a single 
dataset of simulation results (referred to as the vehicle simulation 
database, or simply, database), which may then be utilized by the CAFE 
modeling system.
    To incorporate the results of the combined database of Autonomie-
simulated and additional technologies, while still preserving the basic 
structure of the CAFE Model's technology subsystem, it was necessary to 
translate the points in this database into corresponding locations 
defined by the technology pathways. By recognizing that most of the 
pathways are unrelated, and are only logically linked to designate the 
direction in which technologies are allowed to progress, it is possible 
to condense the paths into a smaller number of groups based on the 
specific technology. In addition, to allow for technologies present on 
the Basic Engine and Dynamic Road Load (DLR--i.e., MASS, AERO, and 
ROLL) paths to be evaluated and applied in any given combination, a 
unique group was established for each of these technologies.
    As such, the following technology groups are defined within the 
modeling system: Engine cam configuration (CONFIG), VVT engine 
technology (VVT), VVL engine technology (VVL), SGDI engine technology 
(SGDI), DEAC engine technology (DEAC), non-basic engine technologies 
(ADVENG), transmission technologies (TRANS), electrification and 
hybridization (ELEC), low rolling resistance tires (ROLL), aerodynamic 
improvements (AERO), mass reduction levels (MR), EFR engine technology 
(EFR), electric accessory improvement technologies (ELECACC), LDB 
technology (LDB), and SAX technology (SAX). The combination of 
technologies along each of these groups forms a unique technology state 
vector and defines a unique technology combination that corresponds to 
a single point in the database for each technology class evaluated 
within the modeling system.
    As an example, a technology state vector describing a vehicle with 
a SOHC engine, variable valve timing (only), a 6-speed automatic 
transmission, a belt-integrated starter generator, rolling resistance 
(level 1), aerodynamic improvements (level 2), mass reduction (level 
1), electric power steering, and low drag brakes, would be specified as 
``SOHC; VVT; AT6; BISG; ROLL10; AERO20; MR1; EPS; LDB.'' \625\ By 
assigning each unique technology combination a state vector such as the 
one in the example, the CAFE Model can then assign each vehicle in the 
analysis fleet an initial state that corresponds to a point in the 
database.
---------------------------------------------------------------------------

    \625\ In the example technology state vector, the series of 
semicolons between VVT and AT6 correspond to the engine technologies 
which are not included as part of the combination, while the gap 
between MR1 and EPS corresponds to EFR and the omitted technology 
after LDB is SAX. The extra semicolons for omitted technologies are 
preserved in this example for clarity and emphasis, and will not be 
included in future examples.
---------------------------------------------------------------------------

    Once a vehicle is assigned (or mapped) to an appropriate technology 
state vector (from one of approximately three million unique 
combinations, which are defined in the vehicle simulation database as 
CONFIG; VVT; VVL; SGDI; DEAC; ADVENG; TRANS; ELEC; ROLL; AERO; MR; EFR; 
ELECACC; LDB; SAX), adding a new technology to the vehicle simply 
represents progress from a previous state vector to a new state vector. 
The previous state vector simply refers to the technologies that are 
currently in use on a vehicle. The new state vector, however, is 
computed within the modeling system by adding a new technology to the 
combination of technologies represented by the previous state vector, 
while simultaneously removing any other technologies that are 
superseded by the newly added one.
    For example, consider the vehicle with the state vector described 
as: SOHC; VVT; AT6; BISG; ROLL10; AERO20; MR1; EPS; LDB. Assume the 
system is evaluating PHEV20 as a candidate technology for application 
on this vehicle. The new state vector for this vehicle is computed by 
removing SOHC, VVT, AT6, and BISG technologies from the previous state 
vector,\626\ while also adding PHEV20, resulting in the following: 
PHEV20; ROLL10; AERO20; MR1; EPS; LDB.
---------------------------------------------------------------------------

    \626\ For more discussion of how the CAFE Model handles 
technology supersession, see Section VI.A.7.
---------------------------------------------------------------------------

    From here, it is relatively simple to obtain a fuel economy 
improvement factor for any new combination of technologies and apply 
that factor to the fuel economy of a vehicle in the analysis fleet. The 
formula for calculating a vehicle's fuel economy after application of 
each successive technology represented within the database is defined, 
simply put, as the difference between the fuel economy improvement 
factor associated with the technology state vector before application 
of a candidate technology, and after the application of a candidate 
technology.\627\ This is applied to the original compliance fuel 
economy value for a discrete vehicle in the MY 2017 analysis fleet, as 
discussed previously in Section VI.B.3 Technology Effectiveness.
---------------------------------------------------------------------------

    \627\ For more discussion of how the CAFE Model calculates a 
vehicle's fuel economy where the vehicle switches from one type of 
fuel to another, for example, from gasoline operation to diesel 
operation or from gasoline operation to plug-in hybrid/electric 
vehicle operation, see Section VI.A CAFE Model.
---------------------------------------------------------------------------

    The fuel economy improvement factor is defined in a way that 
captures the incremental improvement of moving between points in the 
database, where each point is defined uniquely as a combination of up 
to 15 distinct technologies describing, as mentioned above, the 
engine's cam configuration, multiple distinct combinations of engine 
technologies, transmission, electrification type, and various vehicle 
body level technologies.
    Unlike the preceding versions of the modeling system, the current 
version of the CAFE Model relies entirely on the vehicle simulation 
database for calculating fuel economy improvements resulting from all 
technologies available to the system. The fuel economy improvements are 
derived from the factors defined for each unique technology combination 
or state vector. Each time the improvement factor for a new state 
vector is added to a vehicle's existing fuel economy, the factor 
associated with the old technology combination is entirely removed. In 
that sense, application of technologies obtained from the Autonomie 
database is ``self-correcting'' within the model. As such, special-case 
adjustments defined by the previous version of the model are not 
applicable to this one.
    Meszler Engineering Services, commenting on behalf of Natural 
Resources Defense Council, commented that ``[w]ith very limited 
exception, technology is not included in the NPRM CAFE model if it was 
not included in the simulation modeling that underlies the Argonne 
database,'' citing the ``add-on'' technologies and technologies with 
fixed effectiveness values.\628\ Meszler continued, ``[t]his same 
limitation controls the coupling of technologies, and by extension the 
definition of the CAFE model technology pathways. If a combination of 
technologies were not modeled during the development of the Argonne 
database, that package (or combination) of technologies is not 
available for adoption in the CAFE model. Both of these design 
constraints serve to limit the slate of technologies available to 
respond to fuel economy

[[Page 24349]]

standards. The slate of available technologies is basically constrained 
to those included in NHTSA's research activity. If a technology or 
technology combination was not in the NHTSA research planning process, 
it is not available in the model.'' Finally, Meszler stated that 
``because of the constrained model architecture and the reliance on the 
Argonne database for impact estimates, independently expanding the 
model to include additional technologies or technology combinations is 
not trivial.''
---------------------------------------------------------------------------

    \628\ NHTSA-2018-0067-11723, at 4-5.
---------------------------------------------------------------------------

    We agree that expanding the database to include new technologies is 
not trivial. However, it is possible. The set of available technologies 
is part of the model code, and the code is made public upon each 
release of the model. Many commenters made modifications to the model 
code, conducted additional tests of their own, and presented their 
results to the agencies in the form of public comments before the end 
of the public comment period. A user could add the new technology, 
identify the associated engineering restrictions that determine 
combinations for which that technology should not be considered, and 
add the relevant rows (representing possible technology combinations 
that include the new technology) in the database (which exists locally 
on every computer that runs the model). An enterprising user could also 
take an existing technology along a given path and replace the 
efficiency values with new values--presumably from their own full 
vehicle simulations for each technology combination that contains the 
technology in question. Given the length of time and computing power 
required to simulate vehicle fuel economy on the test cycle for every 
possible combination that could be considered by the CAFE model, using 
a pre-defined database that represents a large ensemble of simulated 
technology combinations is preferable to the alternative of fully 
integrating a vehicle simulation model that would be required to run in 
real-time during the compliance simulation to evaluate the 
effectiveness of every combination considered (not just applied) by the 
model.
4. Technology Costs
    In the proposal, the agencies estimated present and future costs 
for fuel-saving technologies, taking into consideration the type of 
vehicle, or type of engine if technology costs vary by application. 
These cost estimates are based on three main inputs. First, the 
agencies estimated direct manufacturing costs (DMCs), or the component 
and labor costs of producing and assembling the physical parts and 
systems, with estimated costs assuming high volume production. DMCs 
generally do not include the indirect costs of tools, capital 
equipment, financing costs, engineering, sales, administrative support 
or return on investment. Second, the agencies accounted for these 
indirect costs via a scalar markup of direct manufacturing costs (the 
retail price equivalent, or RPE). Finally, costs for technologies may 
change over time as industry streamlines design and manufacturing 
processes. The agencies therefore estimated potential cost improvements 
with learning effects (LE). The retail cost of equipment in any future 
year is estimated to be equal to the product of the DMC, RPE, and LE. 
Considering the retail cost of equipment, instead of merely direct 
manufacturing costs, is important to account for the real-world price 
effects of a technology, as well as market realities. Absent a 
government mandate, motor vehicle manufacturers will not undertake 
expensive development and production efforts to implement technologies 
without realistic prospects of consumers being willing to pay enough 
for such technology to allow for the manufacturers to recover their 
investment.
a) Direct Manufacturing Costs
    Direct manufacturing costs (DMCs) are the component costs of the 
physical parts and systems that make up a complete vehicle. The 
analysis used agency-sponsored tear-down studies of vehicles and parts 
to estimate the DMCs of individual technologies, in addition to 
independent tear-down studies, other publications, and confidential 
business information. In the simplest cases, the agency-sponsored 
studies produced results that confirmed third-party industry estimates, 
and aligned with confidential information provided by manufacturers and 
suppliers. In cases with a large difference between the tear-down study 
results and credible independent sources, study assumptions were 
scrutinized, and sometimes the analysis was revised or updated 
accordingly.
    Due to the variety of technologies and their applications, and the 
cost and time required to conduct detailed tear-down analyses, the 
agencies did not sponsor teardown studies for every technology. In 
addition, many fuel-saving technologies were considered that are pre-
production, or sold in very small pilot volumes. For those 
technologies, a tear-down study could not be conducted to assess costs 
because the product is not yet in the marketplace for evaluation. In 
these cases, the agencies relied upon third-party estimates and 
confidential information from suppliers and manufacturers were relied 
upon; however, there are some common pitfalls with relying on 
confidential business information to estimate costs. The agencies and 
the source may have had incongruent or incompatible definitions of 
``baseline.'' The source may have provided DMCs at a date many years in 
the future, and assumed very high production volumes, important caveats 
to consider for agency analysis. In addition, a source, under no 
contractual obligation to the agencies, may provide incomplete and/or 
misleading information. In other cases, intellectual property 
considerations and strategic business partnerships may have contributed 
to a manufacturer's cost information and could be difficult to account 
for in the model as not all manufacturer's may have access to 
proprietary technologies at stated costs. The agencies carefully 
evaluated new information in light of these common pitfalls, especially 
regarding emerging technologies.
    Specifically, the analysis used third-party, forward-looking 
information for advanced cylinder deactivation and variable compression 
ratio engines. While these cost estimates may be preliminary (as is the 
case with many emerging technologies prior to commercialization), the 
agencies consider them to be reasonable estimates of the likely costs 
of these technologies.
    While costs for fuel-saving technologies reflect the best estimates 
available today, technology cost estimates will likely change in the 
future as technologies are deployed and as production is expanded. For 
emerging technologies, the best information available at the time of 
the analysis was utilized, and cost assumptions will continue to be 
updated for any future analysis. Below, discussion of each category of 
technologies (e.g., engines, transmissions, electrification) summarizes 
comments on corresponding direct cost estimates, and reviews estimates 
the agencies have applied for today's analysis.
Indirect Costs
    As discussed above, direct costs represent the cost associated with 
acquiring raw materials, fabricating parts, and assembling vehicles 
with the various technologies manufacturers are expected to use to meet 
future CAFE and CO2 standards. They include materials, 
labor, and variable energy costs required to produce and assemble the 
vehicle. However, they do not

[[Page 24350]]

include overhead costs required to develop and produce the vehicle, 
costs incurred by manufacturers or dealers to sell vehicles, or the 
profit manufacturers and dealers make from their investments. All of 
these items contribute to the price consumers ultimately pay for the 
vehicle. These components of retail prices are illustrated in Table VI-
23 below. 
[GRAPHIC] [TIFF OMITTED] TR30AP20.115

    In addition to direct manufacturing costs, the agencies estimated 
and considered indirect manufacturing costs. To estimate indirect 
costs, direct manufacturing costs are multiplied by a factor to 
represent the average price for fuel-saving technologies at retail.
    In the Draft TAR and preceding CAFE and safety rulemaking analyses, 
NHTSA relied on a factor, referred to as the retail price equivalent 
(RPE), to account for indirect manufacturing costs. The RPE accounts 
for indirect costs like engineering, sales, and administrative support, 
as well as other overhead costs, business expenses, warranty costs, and 
return on capital considerations. In the Draft TAR (and subsequent 
Determination) as well as the 2012 rulemaking analysis, EPA applied an 
``Indirect Cost Multiplier'' (ICM) approach that it first applied in 
the 2010 rulemaking regarding standards for MYs 2012-2016, which also 
accounted for indirect manufacturing costs, albeit in a different way 
than the RPE approach.
    Some commenters recommended the agencies rely on the ICM approach 
for the current rulemaking, citing EPA's prior peer review and use of 
this approach.\629\ Others supported the agencies' reliance on the RPE 
approach, citing the National Research Council's observations in 2015 
that the ICM approach lacks an empirical basis.\630\ The agencies have 
carefully considered these comments, and conclude that while the ICM 
approach has conceptual merit, its application requires a range of 
specific estimates, and data to support such estimates is scant and, in 
some cases, nonexistent. The agencies have, therefore, applied the RPE 
approach for this final rule, as in the NPRM analysis and other 
rulemaking analyses. The following sections discuss both approaches in 
detail to explain why the RPE approach was chosen for this final rule.
---------------------------------------------------------------------------

    \629\ See, e.g., ICCT, NHTSA-2018-0067-11741, Attachment 3, at 
I-83. See also CFA, NHTSA-2018-0067-12005, Attachment B, at p.189.
    \630\ See, e.g., Alliance, NHTSA-2018-0067-12073, at 143. See 
also National Research Council, ``Cost, Effectiveness, and 
Deployment of Fuel Economy Technologies for Light-Duty Vehicles,'' 
2015, available at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-lightduty-vehicles (``. . . the empirical basis for such multipliers 
is still lacking, and, since their application depends on expert 
judgment, it is not possible for to determine whether the Agencies' 
ICMs are accurate or not'').
---------------------------------------------------------------------------

(1) Retail Price Equivalent
    Historically, the method most commonly used to estimate indirect 
costs of producing a motor vehicle has been the retail price equivalent 
(RPE). The RPE markup factor is based on an examination of historical 
financial data contained in 10-K reports filed by manufacturers with 
the Securities and Exchange Commission (SEC). It represents the ratio 
between the retail price of motor vehicles and the direct costs of all 
activities that manufacturers engage in, including the design, 
development, manufacturing, assembly,

[[Page 24351]]

and sales of new vehicles, refreshed vehicle designs, and modifications 
to meet safety or fuel economy standards.
    Figure VI-7 indicates that for more than three decades, the retail 
price of motor vehicles has been, on average, roughly 50 percent above 
the direct cost expenditures of manufacturers. This ratio has been 
remarkably consistent, averaging roughly 1.5 with minor variations from 
year to year over this period. At no point has the RPE markup exceeded 
1.6 or fallen below 1.4.\631\ During this time frame, the average 
annual increase in real direct costs was 2.5 percent, and the average 
annual increase in real indirect costs was also 2.5 percent. Figure VI-
7 illustrates the historical relationship between retail prices and 
direct manufacturing costs.\632\
---------------------------------------------------------------------------

    \631\ Based on data from 1972-1997 and 2007. Data were not 
available for intervening years, but results for 2007 seem to 
indicate no significant change in the historical trend.
    \632\ Rogozhin, A., Gallaher, M., & McManus, W., 2009, 
Automobile Industry Retail Price Equivalent and Indirect Cost 
Multipliers. Report by RTI International to Office of Transportation 
Air Quality. U.S. Environmental Protection Agency, RTI Project 
Number 0211577.002.004, February, Research Triangle Park, N.C. 
Spinney, B.C., Faigin, B., Bowie, N., & St. Kratzke, 1999, Advanced 
Air Bag Systems Cost, Weight, and Lead Time analysis Summary Report, 
Contract NO. DTNH22-96-0-12003, Task Orders--001, 003, and 005. 
Washington, DC, U.S. Department of Transportation.
---------------------------------------------------------------------------

    An RPE of 1.5 does not imply that manufacturers automatically mark 
up each vehicle by exactly 50 percent. Rather, it means that, over 
time, the competitive marketplace has resulted in pricing structures 
that average out to this relationship across the entire industry. 
Prices for any individual model may be marked up at a higher or lower 
rate depending on market demand. The consumer who buys a popular 
vehicle may, in effect, subsidize the installation of a new technology 
in a less marketable vehicle. But, on average, over time and across the 
vehicle fleet, the retail price paid by consumers has risen by about 
$1.50 for each dollar of direct costs incurred by manufacturers.
[GRAPHIC] [TIFF OMITTED] TR30AP20.116

    It is also important to note that direct costs associated with any 
specific technology will change over time as some combination of 
learning and resource price changes occurs. Resource costs, such as the 
price of steel, can fluctuate over time and can experience real long-
term trends in either direction, depending on supply and demand. 
However, the normal learning process generally reduces direct 
production costs as manufacturers refine production techniques and seek 
out less costly parts and materials for increasing production volumes. 
By contrast, this learning process does not generally influence 
indirect costs. The implied RPE for any given technology would thus be 
expected to grow over time as direct costs decline relative to indirect 
costs. The RPE for any given year is based on direct costs of 
technologies at different stages in their learning cycles, and which 
may have different implied RPEs than they did in previous years. The 
RPE averages 1.5 across the lifetime of technologies of all ages, with 
a lower average in earlier years of a technology's life, and, because 
of learning effects on direct costs, a higher average in later years.
    The RPE has been used in all NHTSA safety and most previous CAFE 
rulemakings to estimate costs. The National Academy of Sciences 
recommends RPEs of 1.5 for suppliers and 2.0 for in-house production be 
used to estimate total costs. The Alliance of Automobile Manufacturers 
also advocates these values as appropriate markup factors for 
estimating costs of technology changes. An RPE of 2.0 has also been 
adopted by a coalition of environmental and research groups (NESCCAF, 
ICCT, Southwest Research Institute, and TIAX-LLC) in a report on 
reducing heavy truck emissions, and 2.0 is recommended by the U.S. 
Department of Energy for estimating the cost of hybrid-electric and 
automotive fuel cell costs ((see Vyas et al. (2000) in Table VI-24, 
below).
    Table VI-24 below lists other estimates of the RPE. Note that all 
RPE estimates vary between 1.4 and 2.0, with most in the 1.4 to 1.7 
range.

[[Page 24352]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.117

    The RPE has thus enjoyed widespread use and acceptance by a variety 
of governmental, academic, and industry organizations. The RPE has been 
the most commonly used basis for indirect cost markups in regulatory 
analyses. However, as noted above, the RPE is an aggregate measure 
across all technologies applied by manufacturers and is not technology 
specific. A more detailed examination of these technologies is possible 
through an alternative measure, the indirect cost multiplier, which was 
developed to focus more specifically on technologies used to meet CAFE 
and CO2 standards.
---------------------------------------------------------------------------

    \633\ Duleep, K.G. ``2008 Analysis of Technology Cost and Retail 
Price.'' Presentation to Committee on Assessment of Technologies for 
Improving Light Duty Vehicle Fuel Economy, January 25, Detroit, MI.; 
Jack Faucett Associates, September 4, 1985. Update of EPA's Motor 
Vehicle Emission Control Equipment Retail Price Equivalent (RPE) 
Calculation Formula. Chevy Chase, MD--Jack Faucett Associates; 
McKinsey & Company, October 2003. Preface to the Auto Sector Cases. 
New Horizons--Multinational Company Investment in Developing 
Economies, San Francisco, CA.; NRC (National Research Council), 
2002. Effectiveness and Impact of Corporate Average Fuel Economy 
Standards, Washington, DC--The National Academies Press; NRC, 2011. 
Assessment of Fuel Economy Technologies for Light Duty Vehicles. 
Washington, DC--The National Academies Press; Sierra Research, Inc., 
November 21, 2007, Study of Industry-Average Mark-Up Factors used to 
Estimate Changes in Retail Price Equivalent (RPE) for Automotive 
Fuel Economy and Emissions Control Systems, Sacramento, CA--Sierra 
Research, Inc.; Vyas, A. Santini, D., & Cuenca, R. 2000. Comparison 
of Indirect Cost Multipliers for Vehicle Manufacturing. Center for 
Transportation Research, Argonne National Laboratory, April. 
Argonne, Ill.
---------------------------------------------------------------------------

(2) Indirect Cost Multiplier
    A second approach to accounting for indirect costs is the indirect 
cost multiplier (ICM). ICMs specifically evaluate the components of 
indirect costs likely to be affected by vehicle modifications 
associated with environmental regulation. EPA developed the ICM concept 
to enable the application of markups more specific to each technology. 
For example, the indirect cost implications of using tires with better 
rolling resistance would not be the same as those for developing an 
entire new hybrid vehicle technology, which would require far more R&D, 
capital investment, and management oversight. With more than 80 
different technologies available to incrementally achieve fuel economy 
improvements,\634\ a wide range of indirect cost effects might be 
expected. ICMs attempt to isolate only those indirect costs that would 
have to change to develop a specific technology. Thus, for example, if 
a company were to hire additional staff to sell vehicles equipped with 
fuel economy improving technology, or to search the technology 
requirements of new CO2 or CAFE standards, the cost of these 
staff would be included in ICMs. However, if these functions were 
accomplished by existing staff, they would not be included. For 
example, if an executive who normally devoted 10 percent of his time to 
fuel economy standards compliance were to devote 50 percent of his time 
in response to new more stringent requirements, his salary would not be 
included in ICMs because he would be paid the same salary regardless of 
whether he devoted his time to addressing CAFE requirements, developing 
new performance technologies, or improving the company's market share. 
ICMs thus do not account for the diverted resources required for 
manufacturers to meet these standards, but rather for the net change in 
costs manufacturers might experience because of hiring additional 
personal or acquiring additional assets or services.
---------------------------------------------------------------------------

    \634\ There are roughly 40 different basic unique technologies, 
but variations among these technologies roughly double the possible 
number of different technology applications.
---------------------------------------------------------------------------

    For past rulemakings EPA developed both short-term and long-term 
ICMs. Long-term ICMs are lower than short-term ICMs. This decline 
reflects the belief that many indirect costs will decline over time. 
For example, research is initially required to develop a new technology 
and apply it throughout the vehicle fleet, but a lower level of 
research will be required to improve, maintain, or adapt that new 
technology to subsequent vehicle designs.
    While the RPE was derived from data in financial statements 
(reflecting real-world operating and financial results), no similar 
data sources were available to estimate ICMs. ICMs are based on the 
RPE, broken into its components, as shown in Table VI-25. Adjustment 
factors were then developed for those components, based on the 
complexity and time frame of low-, medium-, and high-complexity 
technologies. The adjustment factors were developed from two panels of 
engineers with background in the automobile industry. Initially, a 
group of engineers met and developed an estimate of ICMs for three 
different technologies. This ``consensus'' panel examined one low 
complexity technology, one medium complexity technology, and one high 
complexity technology, with the initial intent of using these 
technologies to represent ICM factors for all technologies falling in 
those categories. At a later date, a second panel was convened to 
examine three more technologies (one low, one medium, and one high 
complexity), using a modified Delphi approach to estimate indirect cost 
effects. The results from the second panel identified the same pattern 
as those of the original report--the indirect cost multipliers increase 
with the

[[Page 24353]]

complexity of the technology and decrease over time. The values derived 
in process are higher than those in the RPE/IC Report by values ranging 
from 0.09 (that is, the multiplier increased from 1.20 to 1.29) to 0.19 
(the multiplier increased from 1.45 to 1.64). This variation may be due 
to differences in the technologies used in each panel. The results are 
shown in Figure VI-8, together with the historical average RPE.
[GRAPHIC] [TIFF OMITTED] TR30AP20.118

    In subsequent CAFE and CO2 analyses for MYs 2011, as 
well as for the 2012-2016 rulemaking, a simple average of the two 
resulting ICMs in the low and medium technology complexity categories 
was applied to direct costs for all unexamined technologies in each 
specific category. For high complexity technologies, the lower 
consensus-based estimate was used for high complexity technologies 
currently being produced, while the higher modified Delphi-based 
estimate was used for more advanced technologies, such as plug-in 
hybrid or electric vehicles, which had little or no current market 
penetration. Note that ICMs originally did not include profit or 
``return on capital,'' a fundamental difference from the RPE. However, 
prior to the 2012-2016 CAFE analysis, ICMs were modified to include 
provision for return on capital.
(3) Application of ICMs in the 2017-2025 Analysis
    For the model year 2017-2025 rulemaking analysis, NHTSA and EPA 
revisited technologies evaluated by EPA staff and reconsidered their 
method of application. The agencies were concerned that averaging 
consensus and modified Delphi ICMs might not be the most accurate way 
to develop an estimate for the larger group of unexamined technologies. 
Specifically, there was concern that some technologies might not be 
representative of the larger groups they were chosen to represent. 
Further, the agencies were concerned that the values developed under 
the consensus method were not subject to the same analytical discipline 
as those developed from the modified Delphi method. As a result, the 
agencies relied primarily on the modified Delphi-based technologies to 
establish their revised distributions. Thus, for the MY 2017-2025 
analysis, the agencies used the following basis for estimating ICMs:
     All low complexity technologies were estimated to equal 
the ICM of the modified Delphi-based low technology-passive aerodynamic 
improvements.
     All medium complexity technologies were estimated to equal 
the ICM of the modified Delphi-based medium technology-engine turbo 
downsizing.
     Strong hybrids and non-battery plug-in hybrid electric 
vehicles (PHEVs) were estimated to equal the ICM of the high complexity 
consensus-based high technology-hybrid electric vehicle.
     PHEVs with battery packs and full electric vehicles were 
estimated to equal the ICM of the high complexity modified Delphi-based 
high technology-plug-in hybrid electric vehicle.
    In addition to shifting the proxy basis for each technology group, 
the agencies reexamined each technology's complexity designation in 
light of the examined technologies that would serve as the basis for 
each group. The resulting designations together with the associated 
proxy technologies are shown in Table VI-25.

[[Page 24354]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.119

    Many basic technologies noted in Table VI-25 have variations 
sharing the same complexity designation and ICM estimate. Table VI-26 
lists each technology used in the CAFE model together with their ICM 
category and the year through which the short-term ICM would be 
applied. Note that the number behind each ICM category designation 
refers to the source of the ICM estimate, with 1 indicating the 
consensus panel and 2 indicating the modified Delphi panel.
BILLING CODE 4910-59-P

[[Page 24355]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.120


[[Page 24356]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.121


[[Page 24357]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.122


[[Page 24358]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.123

BILLING CODE 4910-59-C
    An additional adjustment was made to ICMs to account for the fact 
that they were derived from the RPE analysis for a specific year 
(2007). The agencies believed it would be more appropriate to base ICMs 
on the expected long-term average RPE rather than that of one specific 
year. To account for this, ICMs were normalized to an average RPE 
multiplier level of 1.5.
    Table VI-27 lists values of ICMs by technology category used in the 
previous MY 2017-2025 rulemaking. As noted previously, the Low 1 and 
Medium 1 categories, which were derived using the initial consensus 
panel, are not used. Short-term values applied to CAFE technologies 
thus range from 1.24 for Low complexity technologies, 1.39 for Medium 
complexity technologies, 1.56 for High1 complexity technologies, and 
1.77 for High2 complexity technologies. When long-term ICMs are applied 
in the year following that noted in the far-right column of Table VI-
27, these values will drop to 1.19 for Low, 1.29 for Medium, 1.35 for 
High1 and 1.50 for High2 complexity technologies.
[GRAPHIC] [TIFF OMITTED] TR30AP20.124

    Note that ICMs for warranty costs are listed separately in Table 
VI-27. This was done because warranty costs are treated differently 
than other indirect costs. In some previous analyses (prior to MY 2017-
2025), learning was applied directly to total costs. However, the 
agencies believe learning curves are more appropriately applied only to 
direct costs, with indirect costs established up front based on the ICM 
and held constant while direct costs are reduced by learning. 
Warranties are an exception to this because warranty costs involve 
future replacement of defective parts, and the cost of these parts 
would reflect the effect of learning. Warranty costs were thus treated 
as being subject to learning along with direct costs.\635\
---------------------------------------------------------------------------

    \635\ Note that warranty costs also involve labor costs for 
installation. This is typically done at dealerships, and it is 
unlikely labor costs would be subject to learning curves that affect 
motor vehicle parts or assembly costs. However, the portion of these 
costs that is due to labor versus that due to parts is unknown, so 
for this analysis, learning is applied to the full warranty cost.
---------------------------------------------------------------------------

    The effect of learning on direct costs, together with the eventual 
substitution of lower long-term ICMs, causes the effective markup from 
ICMs to differ from the initial ICM on a yearly basis. An example of 
how this occurs is provided in Table VI-28.\636\ This table, which was 
originally developed for the MY 2017-2025 analysis, traces the effect 
of learning on direct costs and its implications for both total costs 
and the ICM-based markup. Direct costs are assigned a value 
(proportion) of 1 to facilitate analysis on the same basis as ICMs (in 
an ICM markup factor, the proportion of direct costs is represented by 
1 while the proportion of indirect costs is represented by the fraction 
of 1 to the right of the decimal.) Table VI-28 examines the effects of 
these factors on turbocharged downsized engines, one of the more 
prevalent CAFE technologies.
---------------------------------------------------------------------------

    \636\ Table VI-22 illustrates the learning process from the base 
year consistent with the direct cost estimate obtained by the 
agencies. It is a mature technology well into the flat portion of 
the learning curve. Note that costs were actually applied in this 
rulemaking example beginning with MY 2017.

---------------------------------------------------------------------------

[[Page 24359]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.125

    The second column of Table VI-28 lists the learning schedule 
applied to turbocharged downsized engines. Turbocharged downsized 
engines are a mature technology, so the learning schedule captures the 
relatively flat portion of the learning curve occurring after larger 
decreases have already reduced direct costs. The cost basis for 
turbocharged downsized engines in the analysis was effective in 2012, 
so this is the base year for this calculation when direct costs are set 
to 1. The third column shows the progressive decline in direct costs as 
the learning schedule in column 2 is applied to direct costs. Column 4 
contains the value of all indirect costs except warranty. Turbocharged 
downsized engines are a medium-complexity technology, so this value is 
taken from the Medium2 row of Table VI-27. The initial value in 2012 is 
the short-term value, which is used through 2018. During this time, 
these indirect costs are not affected by learning, and they remain 
constant. Beginning in 2019, the long-term ICM from Table VI-27 is 
applied.
    The fifth column contains warranty costs. As previously mentioned, 
these costs are considered to be affected by learning like direct 
costs, so they decline steadily until the long-term ICM is applied in 
2019, at which point they drop noticeably before continuing their 
gradual decline. In the sixth column, direct and indirect costs are 
totaled. Results indicate a decline in total costs of roughly 30 
percent during this 14-year period. The last column shows the effective 
ICM-based markup, which is derived by dividing total costs by direct 
costs. Over this period, the ICM-based markup rose from the initial 
short-term ICM level of 1.39 to 1.45 in 2018. It then declined to 1.35 
in 2019 when the long-term ICM was applied to the 2019 direct cost. 
Over the remaining years, it gradually rises back up to 1.41 as 
learning continues to degrade direct costs.
    There are thus two somewhat offsetting processes affecting total 
costs derived from ICMs. The first is the learning curve, which reduces 
direct costs, which raises the effective ICM-based markup. As noted 
previously, learning reflects learned efficiencies in assembly methods 
as well as reduced parts and materials costs. The second is the 
application of a long-term ICM, which reduces the effective ICM-based 
markup. This represents the reduced burden needed to maintain new 
technologies once they are fully developed. In this case, the two 
processes largely offset one another and produce an average real ICM 
over this 14-year period that roughly equals the original short-term 
ICM.
    Figure VI-9 illustrates this process for each of the 4 technologies 
used to represent the universe of fuel economy and CO2 
improving technologies. As with the turbocharged engines, aerodynamic 
improvements and mild hybrid vehicles show a gradual increase in the 
effective ICM-based markup through the point where the long-term ICM is 
applied. At that time, the ICM-based markup makes an abrupt decline 
before beginning a gradual rise. The decline due to application of 
long-term ICMs is particularly pronounced in the case of the mild 
hybrid--even more so than for the advanced hybrid. The advanced hybrid 
ICM behaves somewhat differently because it is shown through its 
developing stages when more radical learning is applied, but only every 
few years. This produces a significant step-up in ICM levels concurrent 
with each learning

[[Page 24360]]

application, followed by a sharp decline when the long-term ICM is 
applied. After that, it begins a gradual rise as more moderate learning 
is applied to reflect its shift to a mature technology. Note that as 
with the turbocharged downsized engine example above, for the 
aerodynamic improvements and mild hybrid technologies, the offsetting 
processes of learning and long-term ICMs result in an average ICM over 
the full time frame that is roughly equal to the initial short-term 
ICM. However, the advanced hybrid ICM rose to a level significantly 
higher than the initial ICM. This is a direct function of the rapid 
learning schedule applied in the early years to this developing 
technology. Brand new technologies might thus be expected to have 
effective lifetime ICM markups exceeding their initial ICMs, while more 
mature technologies are more likely to experience ICMs over their 
remaining life span that more closely approximate their initial ICMs.
[GRAPHIC] [TIFF OMITTED] TR30AP20.126

    ICMs for these 4 technologies would drive the indirect cost markup 
rate for the analysis. However, the effect on total costs is also a 
function of the relative incidence of each of the 50+ technologies 
shown in Table VI-26 which are assumed to have ICMs similar to one of 
these 4 technologies. The net effect on costs of these ICMs is also 
influenced by the learning curve appropriate to each technology, 
creating numerous different and unique ICM paths. The average ICM 
applied by the model is also a function of each technology's direct 
cost and because ICMs are applied to direct costs, the measured 
indirect cost is proportionately higher for any given ICM when direct 
costs are higher. The average ICM applied to the fleet for any given 
model year is calculated as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.127

where:

D = direct cost of each technology
A = application rate for each technology
ICM = average ICM applied to each technology
and n = 1, 2 . . . . 88

    The CAFE model predicts technology application rates assuming 
manufacturers will apply technologies

[[Page 24361]]

to meet standards in a logical fashion based on estimated costs and 
benefits. The application rates will thus be different for each model 
year and for each alternative scenario examined. For the MY 2017-2025 
FRIA, to illustrate the effects of ICMs on total technology costs, 
NHTSA calculated the weighted average ICM across all technologies for 
the preferred alternative.\637\ This was done separately for each 
vehicle type and then aggregated based on predicted sales of each 
vehicle type used in the model. Results are shown in Table VI-29.
---------------------------------------------------------------------------

    \637\ For each alternative, this rulemaking examined numerous 
scenarios based on different assumptions, and these assumptions 
could influence the relative frequency of selection of different 
technologies, which in turn could affect the average ICM. The 
scenario examined here assumed a 3 percent discount rate, a 1-year 
payback period, real world application of expected civil penalties, 
and reflects expected voluntary over-compliance by manufacturers.
[GRAPHIC] [TIFF OMITTED] TR30AP20.128

    The ICM-based markups in Table VI-29 were derived in a manner 
consistent with the way the RPE is measured, that is, they reflect 
combined influences of direct cost learning and changes in indirect 
cost requirements weighted by both the incidence of each technology's 
adaptation and the relative direct cost of each technology. The results 
indicate generally higher ICMs for passenger cars than for light 
trucks. This is a function of the technologies estimated to be adopted 
for each respective vehicle type, especially in later years when 
hybrids and electric vehicles become more prevalent in the passenger 
car fleet. The influence of these advanced vehicles is driven primarily 
by their direct costs, which greatly outweigh the costs of other 
technologies. This results in the application of much more weight to 
their higher ICMs. This is most notable in MYs 2024 and 2025 for 
passenger cars, when electric vehicles begin to enter the fleet. The 
average ICM increased 0.013 in 2024 primarily because of these 
vehicles. It immediately dropped 0.017 in 2025 because both an 
additional application of steep (20 percent) learning is applied to the 
direct cost of these vehicles (which reduces their relative weight), 
and the long-term ICM becomes effective in that year (which decreases 
the absolute ICM factor). Both influences occur one year after these 
vehicles begin to enter the fleet because of CAFE requirements.
    ICMs also change over time, again, reflecting the different mix of 
technologies present during earlier years but that are often replaced 
with more expensive technologies in later years. Across all model 
years, the wide-ranging application of diverse technologies required to 
meet CAFE and CO2 standards produced an average ICM-based 
markup (or RPE equivalent) of approximately 1.34, applying only 67 
percent of the indirect costs found in the RPE and implying total costs 
11 percent below those predicted by the RPE-based calculation.
(4) Uncertainty
    As noted above, the RPE and ICM assign different markups over 
direct manufacturing costs, and thus imply different total cost 
estimates for CAFE and CO2 technologies. While there is a 
level of uncertainty associated with both markups, this uncertainty 
stems from different issues. The RPE is derived from financial 
statements and is thus grounded in historical data. Although 
compilation of this data is subject to some level of interpretation, 
the two independent researchers who derived RPE estimates from these 
financial reports each reached essentially identical conclusions, 
placing the RPE at roughly 1.5. All other estimates of the RPE fall 
between 1.4 and 2.0, and most are between 1.4 and 1.7. There is thus a 
reasonable level of consistency among researchers that RPEs are 1.4 or 
greater. In addition, the RPE is a measure of the cumulative effects of 
all operations manufacturers undertake in the course of producing their 
vehicles, and is thus not specific to individual technologies, nor of 
CAFE or CO2 technologies in particular. Because this 
provides only a single aggregate measure, using the RPE multiplier 
results in the application of a common incremental markup to all 
technologies. This assures the aggregate cost effect across all 
technologies is consistent with empirical data, but it does not allow 
for indirect cost discrimination among different technologies or over 
time. Because it is applied across all changes, this implies the markup 
for some technologies is likely to be understated, and for others it is 
likely to be overstated.
    By contrast, the ICM process derives markups specific to several 
CAFE and CO2 technologies, but these markups

[[Page 24362]]

have no basis in empirical data. They are based on informed judgment of 
a panel of engineers with auto industry experience regarding cost 
effects of a small sample (roughly 8 percent) of the 50+ technologies 
applied to achieve compliance with CAFE and CO2 standards. 
Uncertainty regarding ICMs is thus based both on the accuracy of the 
initial assessments of the panel on the examined technologies and on 
the assumption that these 4 technologies are representative of the 
remaining technologies that were not examined. Both agencies attempted 
to categorize these technologies in the most representative way 
possible. However, while this represented the best judgment of EPA and 
NHTSA's engineering staffs at that time, the actual effect on indirect 
costs remains uncertain for most technologies. As with RPEs, this means 
that even if ICMs were accurate for the specific technologies examined, 
indirect cost will be understated for some technologies and overstated 
for others.
    There was considerable uncertainty demonstrated in the ICM panel's 
assessments, as illustrated by the range of estimates among the 14 
modified Delphi panel members surrounding the central values reported 
by the panel. These ranges are shown in Table VI-30 and Figure VI-10, 
Figure VI-11, and Figure VI-12 below. For the low complexity 
technology, passive aerodynamic improvements, panel responses ranged 
from a low of basically no indirect costs (1.001 short term and 1.0 
long term), to a high of roughly a 40 percent markup (1.434 and 1.421). 
For the medium complexity technology, turbo charged and downsized 
engines, responses ranged from a low estimate implying almost no 
indirect cost (1.018 and 1.011), to a high estimate implying that 
indirect costs for this technology would roughly equal the average RPE 
(1.5) for all technologies (1.527 and 1.445). For the high complexity 
technology, plug-in hybrid electric vehicles, responses ranged from a 
low estimate that these vehicles would require significantly less 
indirect cost than the average RPE (1.367 and 1.121) to a high estimate 
implying they would require more indirect costs than the average RPE 
(2.153 and 1.691). There was considerable diversity of opinion among 
the panel members.\638\ This is apparent in Figure VI-10, Figure VI-11, 
and Figure VI-12, which show the 14 panel members' final estimates for 
short-term ICMs as scatter plots.
---------------------------------------------------------------------------

    \638\ Sample confidence intervals, which mitigate the effect of 
outlying opinions, indicate a less extreme but still significant 
range of ICMs. Applying mean ICMs helps mitigate these potential 
differences, but there is clearly a significant level of uncertainty 
regarding indirect costs. A t-distribution is used to estimate 
confidence intervals because of the small sample size (14 panel 
members).
[GRAPHIC] [TIFF OMITTED] TR30AP20.129


[[Page 24363]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.130

[GRAPHIC] [TIFF OMITTED] TR30AP20.131


[[Page 24364]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.132

    Although these results were based on modified Delphi panel 
techniques, it is apparent the goal of the Delphi process, an eventual 
consensus or convergence of opinion among panel experts, was not 
achieved. Given this lack of consensus and the divergence of ICM-based 
results from the only available empirical measure (the RPE), there is 
considerable uncertainty that current ICM estimates provide a realistic 
basis of estimating indirect costs. ICMs have not been validated 
through a direct accounting of actual indirect costs for individual 
technologies, and they produce results that conflict with the only 
available empirical evidence of indirect cost markups. Further, they 
are intended to represent indirect costs specifically associated with 
the most comprehensive redesign effort ever undertaken by the auto 
industry, with virtually every make/model requiring ground-up design 
modifications to comply. This includes entirely new vehicle design 
concepts, extensive material substitution, and complete drivetrain 
redesigns, all of which require significant research efforts and 
assembly plant redesign. Under these circumstances, one might expect 
indirect costs to equal or possibly increase above the historical 
average, but not to decrease, as implied by estimated ICMs. For 
regulations, such as the CAFE and CO2 emission standards 
under consideration, that drive changes to nearly every vehicle system, 
the overall average indirect costs should align with the RPE value. 
Applying RPE to the cost for each technology assures that alignment.
    In the 2015 NAS study, the Committee stated a conceptual agreement 
with the ICM method because ICM takes into account design challenges 
and the activities required to implement each technology. However, 
although endorsing ICMs as a concept, the NAS Committee stated ``the 
empirical basis for such multipliers is still lacking, and, since their 
application depends on expert judgment, it is not possible to determine 
whether the Agencies' ICMs are accurate or not.'' \639\ NAS also stated 
``the specific values for the ICMs are critical because they may affect 
the overall estimates of costs and benefits for the overall standards 
and the cost effectiveness of the individual technologies.'' \640\ The 
Committee encouraged continued research into ICMs given the lack of 
empirical data for them to evaluate ICMs used by the agencies in past 
analyses. On balance, and considering the relative merits of both 
approaches for realistically estimating indirect costs, the agencies 
consider the RPE method to be a more reliable basis for estimating 
indirect costs.
---------------------------------------------------------------------------

    \639\ National Research Council of the National Academies 
(2015). Cost, Effectiveness, and Deployment of Fuel Economy 
Technologies for Light-Duty Vehicles. https://www.nap.edu/resource/21744/deps_166210.pdf.
    \640\ Ibid.
---------------------------------------------------------------------------

(5) Using RPE To Evaluate Indirect Costs in This Analysis
    To ensure overall indirect costs in the analysis align with the 
historical RPE value, the primary analysis has been developed based on 
applying the RPE value of 1.5 to each technology. As noted previously, 
the RPE is the ratio of aggregate retail prices to aggregate direct 
manufacturing costs. The ratio already reflects the mixture of learned 
costs of technologies at various stages of maturity. Therefore, the RPE 
is applied directly to the learned direct cost for each technology in 
each year. This was previously done in the MY 2017-2025 FRIA for the 
preferred alternative for that rulemaking, used in the above analysis 
of average ICMs. Results are shown in Table VI-31.
    Recognizing there is uncertainty in any estimate of indirect costs, 
a sensitivity analyses of indirect costs has also been conducted by 
applying a lower RPE value as a proxy for the ICM approach. This value 
was derived from a direct comparison of incremental technology costs 
determined in the MY 2017-2025 FRIA.\641\ This analysis is summarized 
in Table VI-31 below. From this table, total costs were estimated to be 
roughly 18 percent lower using ICMs compared to the RPE. As previously 
mentioned, there are two different reasons for these differences. The 
first is the direct effect of applying a higher retail markup. The 
second is an indirect effect resulting from the influence these 
differing markups have on the order of the selection of technologies in 
the CAFE model, which can change as different direct cost levels 
interact with altered retail markups, shifting their relative overall 
effectiveness.
---------------------------------------------------------------------------

    \641\ See Table 5-9a in Final Regulatory Impact Analysis, 
Corporate Average Fuel Economy for MY 2017-MY 2025 Passenger Cars 
and Light Trucks.
---------------------------------------------------------------------------

    The relative effects of ICMs may vary somewhat by scenario, but in 
this case, the application of ICMs produces total

[[Page 24365]]

technology cost estimates roughly 18 percent lower than those that 
would result from applying a single RPE factor to all technologies, or, 
conversely, the RPE produces estimates that averaged 21 percent higher 
than the ICM. Under the CAFE model construct, which will apply an 
alternate RPE to the same base technology profile to represent ICMs, 
this implies an RPE equivalent of 1.24 would produce similar net 
impacts [1.5/(1 + x) = 1.21, x = 0.24]. This value is applied for the 
ICM proxy estimate. Additional values were also examined over a range 
of 1.1-2.0. The results, as well as the reference case using the 1.5 
RPE, are summarized in Table VI-32.
[GRAPHIC] [TIFF OMITTED] TR30AP20.133

[GRAPHIC] [TIFF OMITTED] TR30AP20.134

    Several responders submitted comments on the issue of indirect 
costs. The International Council on Clean Transportation (ICCT) stated 
that ``The agencies abandoned their previously-used indirect cost 
multiplier method for estimating total costs, which was vetted with 
peer review, and more complexly handled differing technologies with 
different supply chain and manufacturing aspects. The agencies have, at 
this point, opted to use a simplistic retail price equivalent method, 
which crudely assumes all technologies have a 50 percent markup from 
the direct manufacturing technology cost. We recommend the agencies 
revert back to the previously-used and better substantiated ICM 
approach.'' \642\
---------------------------------------------------------------------------

    \642\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    A private commenter, Thomas Stephens, noted that ``In Section II. 
Technical Foundation for NPRM Analysis, under 1. Data Sources and 
Processes for Developing Individual Technology Assumptions, the 
agencies state that indirect costs are estimated using a Retail Price 
Equivalent (RPE) factor. Concerns with RPE factors and the difficulty 
of accounting for differences in indirect costs of different 
technologies when using this approach were identified by the EPA 
(Rogozhin et al., Using indirect cost multipliers to estimate the total 
cost of adding new technology in the automobile industry, International 
Journal of Production Economics 124, 360-368, 2010), which suggested 
using indirect cost (IC) multipliers instead of RPE factors. The EPA 
developed and updated IC multipliers for relevant vehicle technologies 
with automotive industry input and review. The agencies should consider 
using these IC multipliers to estimate indirect manufacturing costs 
instead of RPE factors.'' \643\
---------------------------------------------------------------------------

    \643\ NHTSA-2018-0067-12067.
---------------------------------------------------------------------------

    By contrast, the Alliance of Automobile Manufacturers (The 
Alliance) ``supports the use of retail

[[Page 24366]]

price equivalents in the compliance cost modeling to estimate the 
indirect costs associated with the additional added technology required 
to meet a given future standard. The alternative indirect cost 
multiplier (``ICM'') approach is not sufficiently developed for use in 
rulemaking. As noted by the National Research Council, the indirect 
cost multipliers previously developed by EPA have not been validated 
with empirical data.\644\ Furthermore, in reference to the memorandum 
documenting the development of ICMs previously used by EPA, Exponent 
Failure Analysis Associates found that,
---------------------------------------------------------------------------

    \644\ Cost, Effectiveness, and Development of Fuel Economy 
Technologies for Light-Duty Vehicles, pages 248-49, National 
research Council, the National Academies Press (2015).
---------------------------------------------------------------------------

Past Toyota Comments on Atkinson-Cycle Benefits Have Addressed Only 
Those Derived From Variable Valve Timing
    In response to these comments the agencies continue to find the RPE 
approach preferable to the ICM approach, at least at this stage in the 
development ICM estimates, for the reasons discussed both above and 
previously in the NPRM. The agencies note that the concerns are not 
with the concept of ICMs, but rather with the judgment-based values 
suggested for use as ICMs, which have not been validated, and which 
conflict with the empirically derived RPE value. The agencies will 
continue to monitor any developments in ICM methodologies as part of 
future rulemakings.
c) Stranded Capital Costs
    Past analyses accounted for costs associated with stranded capital 
when fuel economy standards caused a technology to be replaced before 
its costs were fully amortized. The idea behind stranded capital is 
that manufacturers amortize research, development, and tooling expenses 
over many years, especially for engines and transmissions. The 
traditional production life-cycles for transmissions and engines have 
been a decade or longer. If a manufacturer launches or updates a 
product with fuel-saving technology, and then later replaces that 
technology with an unrelated or different fuel-saving technology before 
the equipment and research and development investments have been fully 
paid off, there will be unrecouped, or stranded, capital costs. 
Quantifying stranded capital costs accounts for such lost investments.
    In the Draft TAR and NPRM analyses, only a few technologies for a 
few manufacturers were projected to have stranded capital costs. As 
more technologies are included in this analysis, and as the CAFE model 
has been expanded to account for platform and engine sharing and 
updated with redesign and refresh cycles, accounting for stranded 
capital has become increasingly complex. Separately, manufacturers may 
be shifting their investment strategies in ways that may alter how 
stranded capital calculations were traditionally considered. For 
example, some suppliers sell similar transmissions to multiple 
manufacturers. Such arrangements allow manufacturers to share in 
capital expenditures, or amortize expenses more quickly.
    Manufacturers share parts on vehicles around the globe, achieving 
greater scale and greatly affecting tooling strategies and costs. Given 
these trends in the industry and their uncertain effect on capital 
amortization, and given the difficulty of handling this uncertainty in 
the CAFE model, this analysis does not account for stranded capital. 
The agencies' analysis continues to rely on the CAFE model's explicit 
year-by-year accounting for estimated refresh and redesign cycles, and 
shared vehicle platforms and engines, to moderate the cadence of 
technology adoption and thereby limit the implied occurrence of 
stranded capital and the need to account for it explicitly. The 
agencies will monitor these trends to assess the role of stranded 
capital moving forward.
d) Cost Learning
    Manufacturers make improvements to production processes over time, 
which often result in lower costs. ``Cost learning'' reflects the 
effect of experience and volume on the cost of production, which 
generally results in better utilization of resources, leading to higher 
and more efficient production. As manufacturers gain experience through 
production, they refine production techniques, raw material and 
component sources, and assembly methods to maximize efficiency and 
reduce production costs. Typically, a representation of this cost 
learning, or learning curves, reflect initial learning rates that are 
relatively high, followed by slower learning as additional improvements 
are made and production efficiency peaks. This eventually produces an 
asymptotic shape to the learning curve, as small percent decreases are 
applied to gradually declining cost levels. These learning curve 
estimates are applied to various technologies that are used to meet 
CAFE standards.
    For the NPRM and this final rule, the agencies estimated cost 
learning by considering methods established by T.P. Wright \645\ and 
later expanded upon by J.R. Crawford. Wright, examining aircraft 
production, found that every doubling of cumulative production of 
airplanes resulted in decreasing labor hours at a fixed percentage. 
This fixed percentage is commonly referred to as the progress rate or 
progress ratio, where a lower rate implies faster learning as 
cumulative production increases. J.R. Crawford expanded upon Wright's 
learning curve theory to develop a single unit cost model,\646\ that 
estimates the cost of the nth unit produced given the following 
information is known: (1) Cost to produce the first unit; (2) 
cumulative production of n units; and (3) the progress ratio.
---------------------------------------------------------------------------

    \645\ Wright, T.P., Factors Affecting the Cost of Airplanes. 
Journal of Aeronautical Sciences, Vol. 3 (1936), pp. 124-125. 
Available at http://www.uvm.edu/pdodds/research/papers/others/1936/wright1936a.pdf.
    \646\ Crawford, J.R., Learning Curve, Ship Curve, Ratios, 
Related Data, Burbank, California-Lockheed Aircraft Corporation 
(1944).
---------------------------------------------------------------------------

    As pictured in Figure VI-13, Wright's learning curve shows the 
first unit is produced at a cost of $1,000. Initially cost per unit 
falls rapidly for each successive unit produced. However, as production 
continues, cost falls more gradually at a decreasing rate. For each 
doubling of cumulative production at any level, cost per unit declines 
20 percent, so that 80 percent of cost is retained. The CAFE model uses 
the basic approach by Wright, where cost reduction is estimated by 
applying a fixed percentage to the projected cumulative production of a 
given fuel economy technology.

[[Page 24367]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.135

    The analysis accounts for learning effects with model year-based 
cost learning forecasts for each technology that reduce direct 
manufacturing costs over time. The agencies evaluated the historical 
use of technologies, and reviewed industry forecasts to estimate future 
volumes for the purpose of developing the model year-based technology 
cost learning curves.
    The following section discusses the agencies' development of model 
year-based cost learning forecasts, including how the approach has 
evolved from the 2012 rulemaking for MY 2017-2025 vehicles, and how the 
progress ratios were developed for different technologies considered in 
the analysis. Finally, the agencies discuss how these learning effects 
are applied in the CAFE Model.
(1) Time Versus Volume-Based Learning
    For the 2012 joint CAFE/CO2 rulemaking, the agencies 
developed learning curves as a function of vehicle model year.\647\ 
Although the concept of this methodology is derived from Wright's 
cumulative production volume-based learning curve, its application for 
CAFE and CO2 technologies was more of a function of time. 
More than a dozen learning curve schedules were developed, varying 
between fast and slow learning, and assigned to each technology 
corresponding to its level of complexity and maturity. The schedules 
were applied to the base year of direct manufacturing cost and 
incorporate a percentage of cost reduction by model year declining at a 
decreasing rate through the technology's production life. Some newer 
technologies experience 20 percent cost reductions for introductory 
model years, while mature or less complex technologies experience 0-3 
percent cost reductions over a few years.
---------------------------------------------------------------------------

    \647\ CAFE 2012 Final Rule, NHTSA DOT, 77 FR 62624.
---------------------------------------------------------------------------

    In their 2015 report to Congress, the National Academy of Sciences 
(NAS) recommended the agencies should ``continue to conduct and review 
empirical evidence for the cost reductions that occur in the automobile 
industry with volume, especially for large-volume technologies that 
will be relied on to meet the CAFE/GHG standards.'' \648\
---------------------------------------------------------------------------

    \648\ Cost, Effectiveness, and Deployment of Fuel Economy 
Technologies for Light-Duty Vehicles, National Research Council of 
the National Academies (2015), available at https://www.nap.edu/resource/21744/deps_166210.pdf.
---------------------------------------------------------------------------

    In response, the agencies have incorporated statically projected 
cumulative volume production data of fuel economy improving 
technologies, representing an improvement over the previously used 
time-based method. Dynamic projections of cumulative production are not 
feasible with current CAFE model capabilities, so one set of projected 
cumulative production data for most vehicle technologies was developed 
for the purpose of determining cost impact. For many technologies 
produced and/or sold in the U.S., historical cumulative production data 
was obtained to establish a starting point for learning schedules. 
Groups of similar technologies or technologies of similar complexity 
may share identical learning schedules.
    The slope of the learning curve, which determines the rate at which 
cost reductions occur, has been estimated using research from an 
extensive literature review and automotive cost tear-down reports (see 
below). The slope of the learning curve is derived from the progress 
ratio of manufacturing automotive and other mobile source technologies.
(2) Deriving the Progress Ratio Used in This Analysis
    Learning curves vary among different types of manufactured 
products. Progress ratios can range from 70 to 100 percent, where 100 
percent indicates no learning can be achieved.\649\ Learning effects 
tend to be greatest in operations where workers often touch the 
product, while effects are less substantial in operations consisting of 
more automated processes. As automotive manufacturing plant processes 
become increasingly automated, a progress ratio towards the higher end 
would seem more suitable. The agencies incorporated findings from 
automotive cost-teardown studies with EPA's literature review of 
learning-related studies to estimate a progress ratio used to determine 
learning schedules of fuel economy improving technologies.
---------------------------------------------------------------------------

    \649\ Martin, J., ``What is a Learning Curve?'' Management and 
Accounting Web, University of South Florida, available at: https://www.maaw.info/LearningCurveSummary.htm.
---------------------------------------------------------------------------

    EPA's literature review examined and summarized 20 studies related 
to learning in manufacturing industries and mobile source 
manufacturing.\650\

[[Page 24368]]

The studies focused on many industries, including motor vehicles, 
ships, aviation, semiconductors, and environmental energy. Based on 
several criteria, EPA selected five studies providing quantitative 
analysis from the mobile source sector (progress ratio estimates from 
each study are summarized in Table VI-33, below). Further, those 
studies expand on Wright's Learning Curve function by using cumulative 
output as a predictor variable, and unit cost as the response variable. 
As a result, EPA determined a best estimate of 84 percent as the 
progress ratio in mobile source industries. However, of those five 
studies, EPA at the time placed less weight on the Epple et al. (1991) 
study, because of a disruption in learning due to incomplete knowledge 
transfer from the first shift to introduction of a second shift at a 
North American truck plant. While learning may have decelerated 
immediately after adding a second shift, the agencies note that unit 
costs continued to fall as the organization gained experience operating 
with both shifts. The agencies now recognize that disruptions are an 
essential part of the learning process and should not, in and of 
themselves, be discredited. For this reason, the analysis uses a re-
estimated average progress ratio of 85 percent from those five studies 
(equally-weighted).
---------------------------------------------------------------------------

    \650\ Cost Reduction through Learning in Manufacturing 
Industries and in the Manufacture of Mobile Sources, United States 
Environmental Protection Agency (2015). Prepared by ICF 
International and available at https://19january2017snapshot.epa.gov/sites/production/files/2016-11/documents/420r16018.pdf.
[GRAPHIC] [TIFF OMITTED] TR30AP20.136

    In addition to EPA's literature review, this progress ratio 
estimate was informed based on NHTSA's findings from automotive cost-
teardown studies. NHTSA routinely performs evaluations of costs of 
previously issued Federal Motor Vehicle Safety Standards (FMVSS) for 
new motor vehicles and equipment. NHTSA's engages contractors to 
perform detailed engineering ``tear-down'' analyses for representative 
samples of vehicles, to estimate how much specific FMVSS add to the 
weight and retail price of a vehicle. As part of the effort, cost and 
production volume are examined for automotive safety technologies. In 
particular, the agency estimated costs from multiple cost tear-down 
studies for technologies with actual production data from the Cost and 
weight added by the Federal Motor Vehicle Safety Standards for MY 1968-
2012 passenger cars and LTVs (2017).\656\
---------------------------------------------------------------------------

    \651\ Argote, L., Epple, D., Rao, R. D., & Murphy, K., The 
acquisition and depreciation of knowledge in a manufacturing 
organization--Turnover and plant productivity, Working paper, 
Graduate School of Industrial Administration, Carnegie Mellon 
University (1997).
    \652\ Benkard, C. L., Learning and Forgetting--The Dynamics of 
Aircraft Production, The American Economic Review, Vol. 90(4), pp. 
1034-54 (2000).
    \653\ Epple, D., Argote, L., & Devadas, R., Organizational 
Learning Curves--A Method for Investigating Intra-Plant Transfer of 
Knowledge Acquired through Learning by Doing, Organization Science, 
Vol. 2(1), pp. 58-70 (1991).
    \654\ Epple, D., Argote, L., & Murphy, K., An Empirical 
Investigation of the Microstructure of Knowledge Acquisition and 
Transfer through Learning by Doing, Operations Research, Vol. 44(1), 
pp. 77-86 (1996).
    \655\ Levitt, S. D., List, J. A., & Syverson, C., Toward an 
Understanding of Learning by Doing--Evidence from an Automobile 
Assembly Plant, Journal of Political Economy, Vol. 121 (4), pp. 643-
81 (2013).
    \656\ Simons, J. F., Cost and weight added by the Federal Motor 
Vehicle Safety Standards for MY 1968-2012 Passenger Cars and LTVs 
(Report No. DOT HS 812 354). Washington, DC--National Highway 
Traffic Safety Administration (November 2017), at pp. 30-33.
---------------------------------------------------------------------------

    NHTSA chose five vehicle safety technologies with sufficient data 
to estimate progress ratios of each, because these technologies are 
large-volume technologies and are used by almost all vehicle 
manufacturers. Table VI-34 below includes these five technologies and 
yields an average progress rate of 92 percent:
[GRAPHIC] [TIFF OMITTED] TR30AP20.137


[[Page 24369]]


    For a final progress ratio used in the CAFE model, the five 
progress rates from EPA's literature review and five progress rates 
from NHTSA's evaluation of automotive safety technologies results were 
averaged. This resulted in an average progress rate of approximately 89 
percent. Equal weight was placed on progress ratios from all 10 
sources. More specifically, equal weight was placed on the Epple et al. 
(1991) study, because disruptions have more recently been recognized as 
an essential part in the learning process, especially in an effort to 
increase the rate of output. Further discussion of how the progress 
ratios were derived for this analysis is located in FRIA Section 9.
    ICCT commented that the choice to use safety technology as a model 
for fuel efficiency led to lower learning rates in the NPRM analysis 
compared to prior analyses.\657\ ICCT stated that safety technologies 
were chosen for the NPRM because they are used by almost every 
manufacturer, in contrast to fuel efficiency technologies, where not 
every manufacturer will use them, particularly when they are first 
introduced. ICCT stated that to show the impact of changing learning 
rates, the agencies should run a sensitivity analysis using the 
learning rates in the TAR, as well as EPA's learning rates in its Final 
Determination. ICCT concluded that ``[w]ithout doing so and without 
conducting a peer review of the change in approach, it appears clear 
the agencies have decided to switch to a new costing method that 
affects all future costs, but without any significant research 
justification, vetting, or review.''
---------------------------------------------------------------------------

    \657\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    The agencies' selection of a progress rate of 0.89 is based on an 
average of findings across research and literature reviews conducted by 
NHTSA and EPA. The EPA cited rates were derived from five studies 
selected from a sample of 20 transportation modal learning studies that 
were examined by an EPA contractor, ICF International.\658\ One of 
these 5 studies (Benkard (2000) examines learning in the commercial 
aircraft industry, which the author notes has many unique features that 
influence marginal costs. It also has the lowest progress rate. The 
agencies note that EPA regulates all mobile sources, and while the 
inclusion of non-passenger vehicle studies in their report was 
justified, it may have biased the estimate of learning attributable to 
the motor vehicle industry. Notably, nearly all of the other studies 
included in the ICF International study found progress rates higher 
than the 0.84 rate selected by the authors at that time. In reviewing 
the ICF study, NHTSA found many other studies not included in the 
report, including many specific to the motor vehicle and environmental 
technology industries. Over 90 percent of those studies indicated 
higher progress ratios than ICF recommended.\659\ The agencies' current 
approach includes a broader and more representative sample of these 
studies rather than the narrow sample selected by ICF.
---------------------------------------------------------------------------

    \658\ Cost Reduction through Learning in Manufacturing 
Industries and in the Manufacture of Mobile Sources. United States 
Environmental Protection Agency. Prepared by ICF International and 
available at: https://19january2017snapshot.epa.gov/sites/production/files/2016-11/documents/420r16018.pdf.
    \659\ See, for example, progress ratios of multiple technologies 
referenced in The Carbon Productivity Challenge: Curbing Climate 
Change and Sustaining Economic Growth, McKinsey Climate Change 
Special Initiative, McKinsey Global Institute, June 2008 (quoting 
from UC Berkeley Energy Resource Group, Navigant Consulting) and 
Technology Innovation for Climate Mitigation and its Relation to 
Government Policies, Edward S. Rubin, Carnegie Mellon University, 
Presentation to the UNFCCC Workshop on Climate Change Mitigation, 
Bonn, Germany, June 19, 2004.
---------------------------------------------------------------------------

    The agencies do not agree that safety technologies are adopted by 
all manufacturers at an early stage. Most safety technologies are 
initially offered as options or standard equipment on only a small 
segment of the vehicle fleet, typically luxury vehicles. After a number 
of years, these technologies may be adopted on less expensive vehicles, 
and eventually they will become required equipment on all vehicles, but 
the production process is gradual, as it is with fuel efficiency 
technologies. FMVSS are necessarily established as performance 
standards--and automakers are free to develop or choose from existing 
technologies to achieve such performance requirements--much like 
automakers can develop or choose from a number of established fuel 
efficiency technologies to achieve fuel economy requirements. Further, 
the derivation of progress ratios is based on the concept of a doubling 
of cumulative production, not time. Therefore, even if production 
continues at a different pace, it should not disqualify non-fuel 
efficiency studies. Moreover, the derivation of the progress ratio used 
in the TAR and Final Determination document were not confined to fuel 
efficiency technologies. In fact, as noted above, they even included at 
least one entirely unrelated study of the aircraft industry.
    Finally, the agencies note that the previous learning schedules 
used in the TAR and EPA's Final Determination were only developed 
through 2025, whereas this final rule projects learning through 2050. 
The previous learning schedules are thus not directly compatible with 
the analysis conducted in this Final Rule, making a sensitivity 
analysis problematic.
(3) Obtaining Appropriate Baseline Years for Direct Manufacturing Costs 
To Create Learning Curves
    Direct manufacturing costs for each fuel economy improving 
technology were obtained from various sources, as discussed above. To 
establish a consistent basis for direct manufacturing costs in the 
rulemaking analysis, each technology cost is adjusted to MY 2018 
dollars. For each technology, the DMC is associated with a specific 
model year, and sometimes a specific production volume, or cumulative 
production volume. The base model year is established as the MY in 
which direct manufacturing costs were assessed (with learning factor of 
1.00). With the aforementioned data on cumulative production volume for 
each technology and the assumption of a 0.89 progress ratio for all 
automotive technologies, the agencies can solve for an implied cost for 
the first unit produced. For some technologies, the agencies used 
modestly different progress ratios to match detailed cost projections 
if available from another source (for instance, batteries for plug-in 
hybrids and battery electric vehicles).
    This approach produced reasonable estimates for technologies 
already in production, and some additional steps were required to set 
appropriate learning rates for technologies not yet in production. 
Specifically, for technologies not yet in production in MY 2017 (the 
baseline analysis fleet), the cumulative production volume in MY 2017 
is zero, because manufacturers have not yet produced the technologies. 
For pre-production cost estimates in the NPRM, the agencies often 
relied on confidential business information sources to predict future 
costs. Many sources for pre-production cost estimates include 
significant learning effects, often providing cost estimates assuming 
high volume production, and often for a timeframe late in the first 
production generation or early in the second generation of the 
technology. Rapid doubling and re-doubling of a low cumulative volume 
base with Wright's learning curves can provide unrealistic cost 
estimates. In addition, direct manufacturing cost projections can vary 
depending on the initial production volume assumed. Accordingly, the 
agencies carefully examined direct costs with learning, and made 
adjustments to the starting point for those technologies on the 
learning curve to better align

[[Page 24370]]

with the assumptions used for the initial direct cost estimate.
(4) Cost Learning as Applied in the CAFE Model
    For the NPRM analysis, the agencies updated the manner in which 
learning effects apply to costs. In the Draft TAR analysis, the 
agencies had applied learning curves only to the incremental direct 
manufacturing costs or costs over the previous technology on the 
technology tree. In practice, two things were observed: (1) If the 
incremental direct manufacturing costs were positive, technologies 
could not become less expensive than their predecessors on the 
technology tree, and (2) absolute costs over baseline technology 
depended on the learning curves of root technologies on the technology 
tree. For the NPRM and final rule analysis, the agencies applied 
learning effects to the incremental cost over the null technology state 
on the applicable technology tree. After this step, the agencies 
calculated year-by-year incremental costs over preceding technologies 
on the tech tree to create the CAFE model inputs. As discussed below, 
for the final rule, the agencies revised the CAFE model to replace 
incremental cost estimates with absolute estimates, each specified 
relative to the null technology state on the applicable technology 
tree. This change facilitated quality assurance and is expected to make 
cost inputs more transparently relatable to detailed model output. 
Likewise, this change made it easier to apply learning curves in the 
course of developing inputs to the CAFE model.
    The agencies grouped certain technologies, such as advanced 
engines, advanced transmissions, and non-battery electric components 
and assigned them to the same learning schedule. While these grouped 
technologies differ in operating characteristics and design, the 
agencies chose to group them based on their complexity, technology 
integration, and economies of scale across manufacturers. The low 
volume of certain advanced technologies, such as hybrid and electric 
technologies, poses a significant issue for suppliers and prevents them 
from producing components needed for advanced transmissions and other 
technologies at more efficient high scale production. The technology 
groupings were carried over from the NPRM analysis for the final rule 
analysis.\660\ Like the NPRM, this final rule analysis uses the same 
groupings that considers market availability, complexity of technology 
integration, and production volume of the technologies that can be 
implemented by manufacturers and suppliers. For example, technologies 
like ADEAC and VCR are grouped together; these technologies were not in 
production or were only in limited introduction in MY 2017, and are 
planned to be introduced in limited production by a few manufacturers. 
The details of these technologies are discussed in Section VI.C.
---------------------------------------------------------------------------

    \660\ See PRIA Chapter 6 for technology groupings.
---------------------------------------------------------------------------

    In addition, for the final rule, as discussed in Section VI.A.4 
Compliance Simulation, the agencies expanded model inputs to extend the 
explicit simulation of technology application through MY 2050, in 
response to comments on the NPRM. Accordingly, the agencies updated the 
learning curves for each technology group to cover MYs through 2050. 
For MYs 2017-2032, the agencies expect incremental improvements in all 
technologies, particularly in electrification technologies because of 
increased production volumes, labor efficiency, improved manufacturing 
methods, specialization, network building, and other factors. While 
these and other factors contribute to continual cost learning, the 
agencies believe that many fuel economy improving technologies 
considered in this rule will approach a flat learning level by the 
early 2030s. Specifically, older and less complex internal combustion 
engine technologies and transmissions will reach a flat learning curve 
sooner when compared to electrification technologies, which have more 
opportunity for improvement. For batteries and non-battery 
electrification components, the agencies estimated a steeper learning 
curve that will gradually flatten after MY 2040. For a more detailed 
discussion of the electrification learning curves used for the final 
rule analysis, see Section VI.C.3.e) Electrification Costs. The 
following Table VI-35 and Table VI-36 show the learning curve schedules 
for CAFE model technologies for MYs 2017-2033 and MYs 2034-2050.
BILLING CODE 4910-59-P

[[Page 24371]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.138


[[Page 24372]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.139


[[Page 24373]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.140


[[Page 24374]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.141

BILLING CODE 4910-59-C
    Each technology in the CAFE Model is assigned a learning schedule 
developed from the methodology explained previously. For example, the

[[Page 24375]]

following chart shows learning rates for several technologies 
applicable to midsize sedans, demonstrating that while the agencies 
estimate that such learning effects have already been almost entirely 
realized for engine turbocharging (a technology that has been in 
production for many years), the agencies estimate that significant 
opportunities to reduce the cost of the greatest levels of mass 
reduction (e.g., MR5) remain, and even greater opportunities remain to 
reduce the cost of batteries for HEVs, PHEVs, BEVs. In fact, for 
certain advanced technologies, the agencies determined that the results 
predicted by the standard learning curves progress ratio was not 
realistic, based on unusual market price and production relationships. 
For these technologies, the agencies developed specific learning 
estimates that may diverge from the 0.89 progress rate. As shown in 
Figure VI-14, these technologies include: Turbocharging and downsizing 
level 1 (TURBO1), variable turbo geometry electric (VTGE), aerodynamic 
drag reduction by 15 percent (AERO15), mass reduction level 5 (MR5), 20 
percent improvement in low-rolling resistance tire technology over the 
baseline, and battery integrated starter/generator (BISG).
[GRAPHIC] [TIFF OMITTED] TR30AP20.142

(5) Potential Future Approaches to Considering Cost Learning in the 
CAFE Model
    As discussed above, cost inputs to the CAFE model incorporate 
estimates of volume-based learning. As an alternative approach, the 
agencies have considered modifications to the CAFE model that would 
calculate degrees of volume-based learning dynamically, responding to 
the model's application of affected technologies. While it is intuitive 
that the degree of cost reduction achieved through experience producing 
a given technology should depend on the actual accumulated experience 
(i.e., volume) producing that technology, such dynamic implementation 
in the CAFE model is thus far infeasible. Insufficient data have been 
available regarding manufacturers' historical application of specific 
technology. Further, insofar as the agencies' estimates of underlying 
direct manufacturing costs already make some assumptions about volume 
and scale, insufficient information is currently available to determine 
how to dynamically adjust these underlying costs. It should be noted 
that if learning responds dynamically to volume, and volume responds 
dynamically to learning, an internally consistent model solution would 
likely require iteration of the CAFE model to seek a stable solution 
within the model's representation of multiyear planning. As discussed 
below, the CAFE model now supports iteration to balance vehicle

[[Page 24376]]

cost and fuel economy changes with corresponding changes in sales 
volumes, but, this iteration is not yet implemented in a manner that 
would necessarily support the balance of learning effects on a 
multiyear basis. The agencies invited comment on the issue, seeking 
data and methods that would provide the basis for a practicable 
approach to doing so. Having reviewed comments on cost learning 
effects, the agencies conclude it remains infeasible to calculate 
degrees of volume-based learning in a manner that responds dynamically 
to modeled technology application. The agencies will continue to 
examine this issue for future development.
e) Cost Accounting
    The CAFE model applied for the NPRM analysis used an incremental 
approach to specifying technology cost estimates, such that the cost 
for any given technology was specified as an incremental value, 
relative to the technology immediately preceding on the relevant 
technology pathway. For example, the cost of a 7-speed transmission was 
specified as an amount beyond the cost of a 6-speed transmission. This 
approach necessitated careful dynamic accounting for the progressive 
application of the technology as the model worked on a step-by-step 
basis to ``build'' a technology solution. As discussed in the 
corresponding model documentation, the model included complex logic to 
``back out'' some of these costs carefully when, for example, replacing 
a conventional powertrain with a hybrid-electric system.\661\
---------------------------------------------------------------------------

    \661\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting 
today's notice.
---------------------------------------------------------------------------

    To facilitate specification of detailed model inputs and review of 
detailed model outputs, today's CAFE model replaces incremental cost 
inputs with absolute cost inputs, such that the estimated cost of each 
technology is specified relative to a common reference point for the 
relevant technology pathway. For example, the cost of the above-
mentioned 7-speed transmission is specified relative to a 4-speed 
transmission, as is the cost of every other transmission technology. 
This change in the structure of cost inputs does not, by itself, change 
model results, but it does make the connection between these inputs and 
corresponding outputs more transparent. Model documentation 
accompanying today's analysis presents details of the updated structure 
for model cost inputs.
5. Other Inputs to the Agencies' Analysis
    CAFE Model input files described above defining the analysis fleet 
and the fuel-saving technologies to be included in the analysis span 
more than a million records, but deal with a relatively discrete range 
of subjects (e.g., what vehicles are in the fleet, what are the key 
characteristics of those vehicles, what fuel-saving technologies are 
expected to be available, and how might adding those technologies 
impact vehicles' fuel economy levels and costs). The CAFE Model makes 
use of a considerably wider range of other types of inputs, and most of 
these are contained in other model input files. The nature and function 
of many of these inputs remains unchanged relative to the model and 
input files applied for the analysis documented in the proposal that 
preceded today's notice. The CAFE Model documentation accompanying 
today's notice lists and describes all model inputs, and explains how 
inputs are used by the model. Many commenters addressed not only the 
model's function and design, but also specific inputs. Most input 
values are discussed either above (e.g., the preceding subsection 
addresses specific inputs regarding technology costs) or below, in 
subsections discussing specific economic, energy, safety, and 
environmental factors. The remainder of this subsection provides an 
overview of the scope of different model input files. The overview is 
organized based on CAFE Model file types, as in the model 
documentation.
a) Market Data File
    The ``Market Data'' file contains the detailed description--
discussed above--of the vehicle models and model configurations each 
manufacturer produces for sale in the U.S. The file also contains a 
range of other inputs that, though not specific to individual vehicle 
models, may be specific to individual manufacturers. The file contains 
a set of specific worksheets, as follows:
    ``Manufacturers'' worksheet: Lists specific manufacturers, 
indicates whether manufacturers are expected to prefer paying CAFE 
fines to applying technologies that would not be cost-effective, 
indicates what ``payback period'' defines buyers' willingness to pay 
for fuel economy improvements, enumerates CAFE and CO2 
credits banked from model years prior to those represented explicitly, 
and indicates how sales ``multipliers'' are to be applied when 
simulating compliance with CO2 standards.
    ``Credits and Adjustments'' worksheet: Enumerates estimates--
specific to each manufacturer and fleet--of expected CO2 and 
CAFE adjustments reflecting improved AC efficiency, reduced AC 
refrigerant leakage, improvements to ``off cycle'' efficiency, and 
production of flexible fuel vehicles (FFVs). The model applies AC 
refrigerant leakage adjustments only to CO2 levels, and 
applies FFV adjustments only to CAFE levels.
    ``Vehicles'' worksheet: Lists vehicle models and model 
configurations each manufacturer produces for sale in the U.S.; 
identifies shared vehicle platforms; indicates which engine and 
transmission is present in each vehicle model configuration; specifies 
each vehicle model configuration's fuel economy level, production 
volume, and average price; specifies several engineering 
characteristics (e.g., curb weight, footprint, and fuel tank volume); 
assigns each vehicle model configuration to a regulatory class, 
technology class, engine class, and safety class; specifies schedules 
on which specific vehicle models are expected to be redesigned and 
freshened; specifies how much U.S. labor is involved in producing each 
vehicle model/configuration; and indicates whether specific 
technologies are already present on specific vehicle model 
configurations, or, due to engineering or product planning 
considerations, should be skipped.
    ``Engines'' worksheet: Identifies specific engines used by each 
manufacturer and for each engine, lists a unique code (referenced by 
the engine code specified for each vehicle model configuration and 
identifies the fuel(s) with which the engine is compatible, the 
valvetrain design (e.g., DOHC), the engine's displacement, cylinder 
configuration and count, and the engine's aspiration type (e.g., 
naturally aspirated, turbocharged). The worksheet also indicates 
whether specific technologies are already present on specific engines, 
or, due to engineering or product planning considerations, should be 
skipped.
    ``Transmissions'' worksheet: Similar to the Engines worksheet, 
identifies specific transmissions used by each manufacturer and for 
each transmission, lists a unique code (referenced by the transmission 
code specified for each vehicle model configuration and identifies the 
type (e.g., automatic or CVT) and number of forward gears. Also 
indicates whether specific technologies are already present or, due to 
engineering or product planning considerations, should be skipped.

[[Page 24377]]

b) Technologies File
    The Technologies file identifies about six dozen technologies to be 
included in the analysis, indicates when and how widely each technology 
can be applied to specific types of vehicles, provides most of the 
inputs involved in estimating what costs will be incurred, and provides 
some of the inputs involved in estimating impacts on vehicle fuel 
consumption and weight. The file contains the following types of 
worksheets:
    ``Parameters'' worksheet: Not to be confused with the 
``Parameters'' file discussed below, this worksheet in the Technologies 
file indicates, for each technology class, the share of the vehicle's 
curb weight represented by the ``glider'' (the vehicle without the 
powertrain).
    ``Technologies'' worksheet: For each named technology, specifies 
the share of the entire fleet to which the technology may be 
additionally applied in each model year.
    Technology Class worksheets: In a separate worksheet for each of 
the 10 technology classes discussed above (and an additional 2--not 
used for this analysis--for heavy-duty pickup trucks and vans), 
identifies whether and how soon the technology is expected to be 
available for wide commercialization, specifies the percentage of miles 
a vehicle is expected to travel on a secondary fuel (if applicable, as 
for plug-in hybrid electric vehicles), indicates a vehicle's expected 
electric power and all-electric range (if applicable), specifies 
expected impacts on vehicle weight, specifies estimates of costs in 
each model year (and factors by which electric battery costs are 
expected to be reduced in each model year), specifies any estimates of 
maintenance and repair cost impacts, and specifies any estimates of 
consumers' willingness to pay for the technology.
    Engine Type worksheets: In a separate worksheet for each of 28 
initial engine types identified by cylinder count, number of cylinder 
banks, and configuration (DOHC, unless identified as OHV or SOHC), 
specifies estimates of costs in each model year, as well as any 
estimates of impacts on maintenance and repair costs.
c) Parameters File
    The ``Parameters'' file contains inputs spanning a range of 
considerations, such as economic and labor utilization impacts, vehicle 
fleet characteristics, fuel prices, scrappage and safety model 
coefficients, fuel properties, and emission rates. The file contains a 
set of specific worksheets, as follows:
    Economic Values worksheet: Specifies a variety of inputs, including 
social and consumer discount rates to be applied, the ``base year'' to 
which to discount social benefits and costs (i.e., the reference years 
for present value analysis), discount rates to be applied to the social 
cost of CO2 emissions, the elasticity of highway travel with 
respect to per-mile fuel costs (also referred to as the rebound 
effect), the gap between test (for certification) and on-road (aka real 
world) fuel economy, the fixed amount of time involved in each refuel 
event, the share of the tank refueled during an average refueling 
event, the value of travel time (in dollars per hour per vehicle), the 
estimated average number of miles between mid-trip EV recharging events 
(separately for 200 and 300-mile EVs), the rate (in miles of capacity 
per hour of charging) at which EV batteries are recharged during such 
events, the values (in dollars per vehicle-mile) of congestion and 
noise costs, costs of vehicle ownership and operation (e.g., sales 
tax), economic costs of oil imports, estimates of future macroeconomic 
measures (e.g., GDP), and rates of growth in overall highway travel 
(separately for low, reference, and high oil prices).
    Vehicle Age Data worksheet: Specifies nominal average survival 
rates and annual mileage accumulation for cars, vans and SUVs, and 
pickup trucks. These inputs are used only for displaying estimates of 
avoided fuel savings and CO2 emissions while the model is 
operating. Calculations reported in model output files reflect, among 
other things, application of the scrappage model.
    Fuel Prices worksheet: Separately for gasoline, E85, diesel, 
electricity, hydrogen, and CNG, specifies historical and estimated 
future fuel prices (and average rates of taxation). Includes values 
reflecting low, reference, and high estimates of oil prices.
    Scrappage Model Values worksheet: Specifies coefficients applied by 
the scrappage model, which the CAFE Model uses to estimate rates at 
which vehicles will be scrapped (removed from service) during the 
period covered by the analysis.
    Historic Fleet Data worksheet: For model years not simulated 
explicitly (here, model years through 2016), and separately for cars, 
vans and SUVs, and pickup trucks, specifies the initial size (i.e., 
number new vehicles produced for sale in the U.S.) of the fleet, the 
number still in service in the indicated calendar year (here, 2016), 
the relative shares of different fuel types, and the average fuel 
economy achieved by vehicles with different fuel types, and the 
averages of horsepower, curb weight, fuel capacity, and price (when 
new).
    Safety Values worksheet: Specifies coefficients used to estimate 
the extent to which changes in vehicle mass impact highway safety. Also 
specifies statistical value of highway fatalities, the share of 
incremental risk (of any additional driving) internalized by drivers, 
rates relating the cost of damages from non-fatal losses to the cost of 
fatalities, and rates relating the occurrence of non-fatal injuries to 
the occurrence of fatalities.
    Fatality Rates worksheet: Separately for each model year from 1975-
2050, and separately for each vehicle age (through 39 years) specifies 
the estimated nominal number of fatalities incurred per billion miles 
of travel by which to offset fatalities.
    Credit Trading Values worksheet: Specifies whether various 
provisions related to compliance credits are to be simulated (currently 
limited to credit carry-forward and transfers), and specifies the 
maximum number of years credits may be carried forward to future model 
years. Also specifies statutory (for CAFE only) limits on the quantity 
of credit that may be transferred between fleets, and specifies amounts 
of lifetime mileage accumulation to be assumed when adjusting the value 
of transferred credits. Also accommodates a setting indicating the 
maximum number of model years to consider when using expiring credits.
    Employment Values worksheet: Specifies the estimated average 
revenue OEMs and suppliers earn per employee, the retail price 
equivalent factor applied in developing technology costs, the average 
quantity of annual labor (in hours) per employee, a multiplier to apply 
to U.S. final assembly labor utilization in order to obtain estimated 
direct automotive manufacturing labor, and a multiplier to be applied 
to all labor hours.
    Fuel Properties worksheet: Separately for gasoline, E85, diesel, 
electricity, hydrogen, and CNG, specifies energy density, mass density, 
carbon content, and tailpipe SO2 emissions (grams per unit 
of energy).
    Fuel Import Assumptions worksheet: Separately for gasoline, E85, 
diesel, electricity, hydrogen, and CNG, specifies the extent to which 
(a) changes in fuel consumption lead to changes in net imports of 
finished fuel, (b) changes in fuel consumption lead to changes in 
domestic refining output, (c) changes in domestic refining output lead 
to changes in domestic crude oil production, and (d) changes in 
domestic refining output lead to changes in net imports of crude oil.

[[Page 24378]]

    Emissions Health Impacts worksheet: Separately for NOX, 
SO2 and PM2.5 emissions, separately for upstream 
and vehicular emissions, and for each of calendar years 2016, 2020, 
2025, and 2030, specifies estimates of various health impacts, such as 
premature deaths, acute bronchitis, and respiratory hospital 
admissions.
    Carbon Dioxide Emission Costs worksheet: For each calendar year 
through 2080, specifies low, average, and high estimates of the social 
cost of CO2 emissions, in dollars per metric ton. 
Accommodates analogous estimates for CH4 and N2O.
    Criteria Pollutant Emission Costs worksheet: Separately for 
NOX, SO2 and PM2.5 emissions, 
separately for upstream and vehicular emissions, and for each of 
calendar years 2016, 2020, 2025, and 2030, specifies social costs on a 
per-ton basis.
    Upstream Emissions (UE) worksheets: Separately for gasoline, E85, 
diesel, electricity, hydrogen, and CNG, and separately for calendar 
years 2017, 2020, 2025, 2030, 2035, 2040, 2045, and 2050, and 
separately for various upstream processes (e.g., petroleum refining), 
specifies emission factors (in grams per million BTU) for each included 
criteria pollutant (e.g., NOX) and toxic air contaminant 
(e.g., benzene).
    Tailpipe Emissions (TE) worksheets: Separately for gasoline and 
diesel, for each of model years 1975-2050, for each vehicle vintage 
through age 39, specifies vehicle tailpipe emission factors (in grams 
per mile) for CO, VOC, NOX, PM2.5, 
CH4, N2O, acetaldehyde, acrolein, benzene, 
butadiene, formaldehyde, and diesel PM10.
d) Scenarios File
    The CAFE Model represents each regulatory alternative as a discrete 
scenario, identifying the first-listed scenario as the baseline 
relative to which impacts are to be calculated. Each scenario is 
described in a worksheet in the Scenarios input file, with standards 
and related provisions specified separately for each regulatory class 
(passenger car or light truck) and each model year. Inputs specify the 
standards' functional forms and defining coefficients in each model 
year. Multiplicative factors and additive offsets are used to convert 
fuel economy targets to CO2 targets, the two being directly 
mathematically related by a linear transformation. Additional inputs 
specify minimum CAFE standards for domestic passenger car fleets, 
determine whether upstream emissions from electricity and hydrogen are 
to be included in CO2 compliance calculations, specify the 
governing rates for CAFE civil penalties, specify estimates of the 
value of CAFE and CO2 credits (for CAFE Model operating 
modes applying these values), specify how flexible fuel vehicles (FFVs) 
and PHEVs are to be accounted for in CAFE compliance calculations, 
specific caps on adjustments reflecting improvements to off-cycle and 
AC efficiency and emissions, specify any estimated amounts of average 
Federal tax credits earned by HEVs, PHEVs, BEVs, and FCVs. The 
worksheets also accommodate some other inputs, such those as involved 
in analyzing standards for heavy-duty pickups and vans, not used in 
today's analysis.
e) ``Run Time'' Settings
    In addition to inputs contained in the above-mentioned files, the 
CAFE Model makes use of some settings selected when operating the 
model. These include which standards (CAFE or CO2) are to be 
evaluated; what model years the analysis is to span; when technology 
application is to begin; what ``effective cost'' mode is to be used 
when selecting among technologies; whether use of compliance credits is 
to be simulated and, if so, until what model year; whether dynamic 
economic models are to be exercised and, if so, how many sales model 
iterations are to be undertaken and using what price elasticity; 
whether low, average, or high estimates are to be applied for fuel 
prices, the social cost of carbon, and fatality rates; by how much to 
scale benefits to consumers; and whether to report an implicit 
opportunity cost.
f) Simulation Inputs
    As mentioned above, the CAFE Model makes use of databases of 
estimates of fuel consumption impacts and, as applicable, battery costs 
for different combinations of fuel saving technologies. For today's 
analysis, the agencies developed these databases using a large set of 
full vehicle and accompanying battery cost model simulations developed 
by Argonne National Laboratory. To be used as files provided separately 
from the model and loaded every time the model is executed, these 
databases are prohibitively large, spanning more than a million records 
and more than half a gigabyte. To conserve space and speed model 
operation, the agencies have integrated the databases into the CAFE 
Model executable file. When the model is run, however, the databases 
are extracted and placed in an accessible location on the user's disk 
drive. The databases, each of which is in the form of a simple (if 
large) text file, are as follows:
    ``FE1_Adjustments.csv:'' This is the main database of fuel 
consumption estimates. Each record contains such estimates for a 
specific indexed (using a multidimensional ``key'') combination of 
technologies for each of the technology classes in the Market Data and 
Technologies files. Each estimate is specified as a percentage of the 
``base'' technology combination for the indicated technology class.
    ``FE2_Adjustments.csv:'' Specific to PHEVs, this is a database of 
fuel consumption estimates applicable to operation on electricity, 
specified in the same manner as those in the main database.
    ``Battery_Costs.csv:'' Specific to technology combinations 
involving vehicle electrification (including 12V stop-start systems), 
this is a database of estimates of corresponding base costs (before 
learning effects) for batteries in these systems.
g) On Road Fuel Economy and CO2 Emissions Gap
    Rather than rely on the compliance values of fuel economy for 
either historical vehicles or vehicles that go through the full 
compliance simulation, the model applies an ``on-road gap'' to 
represent the expected difference between fuel economy on the 
laboratory test cycle and fuel economy under real-world operation. In 
other words, all of the reported physical impacts analysis (including 
emissions impacts) are based on actual real world fuel consumption and 
emissions, not on values based on 2-cycle fuel economy ratings and 
CO2 emission rates, nor on regulatory incentives such as 
sales multipliers that treat a single vehicle as two vehicles, or that 
set aside emissions resulting from generation of electricity to power 
electric vehicles. This was a topic of interest in the recent peer 
review of the CAFE model. While the model currently allows the user to 
specify an on-road gap that varies by fuel type (gasoline, E85, diesel, 
electricity, hydrogen, and CNG), it does not vary over time, by vehicle 
age, or by technology combination. It is possible that the ``gap'' 
between laboratory fuel economy and real-world fuel economy has changed 
over time, that fuel economy changes as a vehicle ages, or that 
specific combinations of fuel-saving technologies have a larger 
discrepancy between laboratory and real-world fuel economy than others. 
For today's analysis, and considering data EPA collects from 
manufacturers regarding vehicles' fuel economy and CO2 as 
tested for both fuel economy and emissions compliance and for vehicle 
fuel economy and emissions labeling

[[Page 24379]]

(labeling making use of procedures spanning a wider range of real-world 
vehicle operating conditions), the agencies have determined that the 
future gap is, at this time, best estimated using the same values 
applied for the analysis documented in the NPRM. The agencies will 
continue to assess such test data and any other available data 
regarding real-world fuel economy and emissions and, as warranted, will 
revise methods and inputs representing the gap between laboratory and 
real-world fuel economy and CO2 emissions in future 
rulemakings. The sensitivity analysis summarized in the FRIA 
accompanying the final rule includes cases representing narrower and 
wider gaps.

C. The Model Applies Technologies Based on a Least-Cost Technology 
Pathway to Compliance, Given the Framework Above

    The CAFE model, discussed in detail above, is designed to simulate 
compliance with a given set of CAFE or tailpipe CO2 
emissions standards for each manufacturer that sells vehicles in the 
United States. For the final rule analysis, the model began with a 
representation of the MY 2017 vehicle model offerings for each 
manufacturer that included the specific engines and transmissions on 
each model variant, observed sales volumes, and all fuel economy 
improving technology that is already present on those vehicles. From 
there the model added technology, in response to the standards being 
considered, in a way that minimized the cost of compliance and 
reflected many real-world constraints faced by automobile 
manufacturers. The model addressed fleet year-by-year compliance, 
taking into consideration vehicle refresh and redesign schedules and 
shared platforms, engines, and transmissions among vehicles.
    The agencies evaluated a wide array of technologies manufacturers 
could use to improve the fuel economy of new vehicles, in both the 
immediate future and during the timeframe of this rulemaking, to meet 
the fuel economy and CO2 standards. The agencies evaluated 
costs for these technologies, and looked at how costs may change over 
time. The agencies also considered how fuel-saving technologies may be 
used on many types of vehicles (ranging from small cars to trucks) and 
how the technologies may perform in improving fuel economy and 
CO2 emissions in combination with other technologies. With 
cost and effectiveness estimates for technologies, the agencies 
forecast how manufacturers may respond to potential standards and can 
estimate the associated costs and benefits related to technology and 
equipment changes. This assists the assessment of technological 
feasibility and is a building block for the consideration of economic 
practicability of the standards.
    The agencies described in the NPRM that the characterization of 
current and anticipated fuel-saving technologies relied on portions of 
the analysis presented in the Draft TAR, in addition to new information 
that had been gathered and developed since conducting that analysis, 
and the significant, substantive input that was received during the 
Draft TAR comment period.\662\ The Draft TAR considered many 
technologies previously assessed in the 2012 final rule; \663\ in some 
cases, manufacturers have nearly universally adopted a technology in 
today's new vehicle fleet (for example, electric power steering), but 
in other cases, manufacturers only occasionally use a technology in 
today's new vehicle fleet (like turbocharged engines). For a few 
technologies considered in the 2012 rulemaking, manufacturers began 
implementing the technologies but have since largely pivoted to other 
technologies due to consumer acceptance issues (for instance, 
drivability and performance feel issues associated with some dual 
clutch transmissions without a torque converter) or limited commercial 
success.
---------------------------------------------------------------------------

    \662\ 83 FR 43021-22 (Aug. 24, 2018).
    \663\ 77 FR 62624 (Oct. 15, 2012).
---------------------------------------------------------------------------

    In some cases, EPA and NHTSA presented different analytical 
approaches in the Draft TAR. However, for the NPRM and final rule 
analysis, the agencies harmonized their analytical approach to use one 
set of effectiveness values (developed with one tool), one set of cost 
assumptions, and one set of assumptions about the limitations of some 
technologies. To develop these assumptions, the agencies evaluated many 
sources of data, in addition to many stakeholder comments received on 
the Draft TAR. The preferred approach was to harmonize on sources and 
methodologies that were data-driven and reproducible for independent 
verification, produced using tools utilized by OEMs, suppliers, and 
academic institutions, and using tools that could support both CAFE and 
CO2 analysis. As the agencies noted in the NPRM, a single 
set of assumptions also facilitated and focused public comment by 
reducing burden on stakeholders who sought to review all of the 
supporting documentation surrounding the analysis.
    The agencies also identified a preference to use values developed 
from careful review of commercialized technologies; however, in some 
cases for technologies that are new, and are not yet for sale in any 
vehicle, the analysis relied on information from other sources, 
including CBI and third-party research reports and publications. The 
agencies strived to keep the technology analysis as current as possible 
in light of the ongoing technology development and implementation in 
the automotive industry. Additional emerging technologies added for the 
final rule analysis are described in further detail, below.
    The agencies' process to develop effectiveness assumptions is 
described in detail in Section VI.B.3 Technology Effectiveness, and 
summarized here. The NPRM and final rule analysis modeled combinations 
of more than 50 fuel economy-improving technologies across 10 vehicle 
types (an increase from five vehicle types in NHTSA's Draft TAR 
analysis). Only 10 vehicle technology classes were used because large 
portions of the production volume in the analysis fleet have similar 
specifications, especially in highly competitive segments. For 
instance, many mid-sized sedans, small SUVs, and large SUVs coalesce 
around similar specifications, respectively. Baseline simulations have 
been aligned around these modal specifications. Parametrically 
combining these technologies generated more than 100,000 unique 
combinations per vehicle class. Multiplying the unique technology 
combinations by the 10 technology classes resulted in the simulation of 
more than one million individual full-vehicle system models. Modeling 
was also conducted to determine appropriate levels of engine downsizing 
required to maintain baseline vehicle performance when advanced mass 
reduction technology or advanced engine technology were applied. 
Performance neutrality is discussed in detail in VI.B.3.
    Some baseline vehicle assumptions used in the simulation modeling 
were updated since the Draft TAR based on public comments, and further 
assessment of the NPRM and final rule analysis fleets. The agencies 
updated assumptions about curb weight, as well as technology properties 
like baseline rolling resistance, aerodynamic drag coefficients, and 
frontal areas. Many of the assumptions are aligned with published 
research from the Department of Energy and other independent

[[Page 24380]]

sources.\664\ Additional transmission technologies and more levels of 
aerodynamic technologies than NHTSA presented in the Draft TAR analysis 
were also added for the analysis. Having additional technologies in the 
model allowed the agencies to assign baselines and estimate fuel-
savings opportunities with more precision.
---------------------------------------------------------------------------

    \664\ See, e.g., Islam, E., A. Moawad, N. Kim, and A. Rousseau, 
2018a, An Extensive Study on Vehicle Sizing, Energy Consumption and 
Cost of Advance Vehicle Technologies, Report No. ANL/ESD-17/17, 
Argonne National Laboratory, Lemont, Ill., Oct 2018. https://www.autonomie.net/pdfs/ANL_BaSce_FY17_Report_10042018.pdf. Last 
accessed March 18, 2020; Pannone, G. ``Technical Analysis of Vehicle 
Load Reduction Potential for Advanced Clean Cars,'' April 29, 2015. 
Available at https://www.arb.ca.gov/research/apr/past/13-313.pdf. 
Last accessed December 28, 2019.
---------------------------------------------------------------------------

    To develop technology cost assumptions, the agencies estimated 
present and future costs for fuel-saving technologies, taking into 
consideration the type of vehicle, or type of engine if technology 
costs vary by application. Since the 2012 final rule, many cost 
assessments, including tear down studies, were funded and completed, 
and presented as part of the Draft TAR analysis. These studies 
evaluated transmissions, engines, hybrid technologies, and mass 
reduction.\665\ The NPRM and final rule analyses use the 2016 Draft 
TAR's cost estimates for many technologies. In addition to those 
studies, the analysis also leveraged research reports from other 
organizations to assess costs.\666\ Consistent with past analyses, this 
analysis used BatPaC to provide estimates for future battery costs for 
hybrids, plug-in hybrids, and electric vehicles, taking into account 
the different battery design characteristics and taking into account 
the size of the battery for different applications.\667\ The agencies 
also updated technology costs for the NPRM to 2016 dollars, because, as 
in many cases, technology costs were estimated several years ago, and 
since then have further updated technology costs to 2018 dollars for 
the final rule.
---------------------------------------------------------------------------

    \665\ FEV prepared several cost analysis studies for EPA on 
subjects ranging from advanced 8-speed transmissions to belt 
alternator starter, or Start/Stop systems. NHTSA also contracted 
with Electricore and EDAG on teardown studies evaluating mass 
reduction. The 2015 NAS report on fuel economy technologies for 
light-duty vehicles also evaluated the agencies' technology costs 
developed based on these teardown studies, and the technology costs 
used in this proposal were updated accordingly.
    \666\ For example, the agencies relied on reports from the 
Department of Energy's Office of Energy Efficiency & Renewable 
Energy's Vehicle Technologies Office. More information on that 
office is available at https://www.energy.gov/eere/vehicles/vehicle-technologies-office. Other agency reports that were relied on for 
technology or other information are referenced throughout the NPRM 
and accompanying PRIA, and this final rule and the accompanying 
FRIA.
    \667\ For instance, battery electric vehicles with high levels 
of mass reduction may use a smaller battery than a comparable 
vehicle with less mass reduction technology and still deliver the 
same range on a charge. See, e.g., Ward, J. & Gohlke, D. & Nealer, 
Rachael. (2017). The Importance of Powertrain Downsizing in a 
Benefit-Cost Analysis of Vehicle Lightweighting. JOM. 69.
---------------------------------------------------------------------------

    Cost and effectiveness values were estimated for each technology 
included in the analysis. As mentioned above, more than 50 technologies 
were considered in the NPRM and final rule analyses, and the agencies 
evaluated many combinations of these technologies in many applications. 
In the NPRM, the agencies identified overarching potential issues in 
assessing technology effectiveness and cost, including:
     Baseline vehicle technology level assessed as too low, or 
too high. Compliance information was extensively reviewed and 
supplemented with available literature on the vehicle models considered 
in the analysis fleet. Manufacturers could also review the baseline 
technology assignments for their vehicles, and the analysis 
incorporates feedback received from manufacturers.
     Technology costs too low or too high. Tear down cost 
studies, CBI, literature, and the 2015 NAS study information were 
referenced to estimate technology costs. In cases where one technology 
appeared to exceed all other technologies on cost and effectiveness, 
information was acquired from additional sources to confirm or reject 
assumptions. Cost assumptions for emerging technologies were reassessed 
in cases where new information became available.
     Technology effectiveness too high or too low in 
combination with other vehicle technologies. Technology effectiveness 
was evaluated using the Autonomie full-vehicle simulation modeling, 
taking into account the impact of other technologies on the vehicle and 
the vehicle type. Inputs and modeling for the analysis took into 
account laboratory test data for production and some pre-production 
technologies, technical publications, manufacturer and supplier CBI, 
and simulation modeling of specific technologies. Evaluating recently 
introduced production products to inform the technology effectiveness 
models of emerging technologies was preferred; however, some 
technologies that are not yet in production were considered using CBI. 
Simulation modeling used carefully chosen baseline configurations to 
provide a consistent, reasonable reference point for the incremental 
effectiveness estimates.
     Vehicle performance not considered or applied in an 
infeasible manner. Performance criteria, including low speed 
acceleration (0-60 mph time), high speed acceleration (50-80 mph time), 
towing, and gradeability (six percent grade at 65 mph) were also 
considered. In the simulation modeling, resizing was applied to achieve 
the same performance level as the baseline for the least capable 
performance criteria but only with significant design changes. The 
analysis struck a balance by employing a frequency of engine downsizing 
that took product complexity and economies of scale into account.
     Availability of technologies for production application 
too soon or too late. A number of technologies were evaluated that are 
not yet in production. CBI was gathered on the maturity and timing of 
these technologies and the cadence at which manufacturers could adopt 
these technologies.
     Product complexity and design cadence constraints too low 
or too high. Product platforms, refresh and redesign cycles, shared 
engines, and shared transmissions were also considered in the analysis. 
Product complexity and the cadence of product launches were matched to 
historical values for each manufacturer.
     Customer acceptance under estimated or over estimated. 
Resale prices for hybrid vehicles, electric vehicles, and internal 
combustion engine vehicles were evaluated to assess consumer 
willingness to pay for those technologies. The analysis accounts for 
the differential in the cost for those technologies and the amount 
consumers have actually paid for those technologies. Separately, new 
dual-clutch transmissions and manual transmissions were applied to 
vehicles already equipped with these transmission architectures.
    The agencies sought comments on all assumptions for fuel economy 
technology costs, effectiveness, availability, and applicability to 
vehicles in the fleet.
    Several commenters compared the technology effectiveness and cost 
estimates from prior rulemaking actions to the NPRM, some commenting 
that the NPRM analysis represented a better balance of input from all 
stakeholders regarding the potential costs and benefits of future fuel 
economy improving technologies,\668\ and some commenting that the NPRM 
analysis represented a step back from the Draft TAR and EPA's Proposed 
Determination in terms of both the analysis itself and the resulting 
conclusions about the level of technology required to meet the

[[Page 24381]]

augural standards.\669\ Specifically, while some commenters stated that 
the Draft TAR and subsequent EPA midterm review documents had recently 
concluded that augural standards were achievable with very low levels 
of electrification based on currently available information on 
technology effectiveness and cost,\670\ other commenters reiterated 
that conventional gasoline powertrains alone were insufficient to 
achieve post-2021 model year targets.\671\
---------------------------------------------------------------------------

    \668\ See, e.g., NHTSA-2018-0067-11928.
    \669\ See, e.g., NHTSA-2018-0067-11873.
    \670\ See, e.g., NHTSA-2018-0067-11969.
    \671\ See, e.g., NHTSA-2018-0067-12150.
---------------------------------------------------------------------------

    Generally, the automotive industry supported the agencies' NPRM 
analysis over previous analyses. In addition to the automotive 
industry's support of the agencies' use of one modeling tool for 
analysis, discussed in Section IV, above, the industry also commented 
in support of specific technology effectiveness, cost, and adoption 
assumptions used in the updated analysis.
    The Alliance commented in support of the NPRM modeling approach, 
and referenced important technology-specific features of the modeling 
process, including ``The acknowledgement and application of real-world 
limitations on technology application including a limit on the number 
of engine displacements available to any one manufacturer, application 
of shared platforms, engines, and transmissions, and the reality that 
improvements and redesigns of components are not only extended across 
vehicles but sometimes constrained in implementation opportunity to 
common vehicle redesign cycles; recognition of the need for 
manufacturers to follow ``technology'' pathways that retain capital and 
implementation expertise, such as specializing in one type of engine or 
transmission instead of following an unconstrained optimization that 
would cause manufacturers to leap to unrelated technologies and show 
overly optimistic costs and benefits; the application of specific 
instead of generic technology descriptions that allow for the above-
mentioned real-world constraints; [and] the need to accommodate for 
intellectual property rights in that not all technologies will be 
available to all manufacturers.'' \672\
---------------------------------------------------------------------------

    \672\ NHTSA-2018-0067-12073, at 9.
---------------------------------------------------------------------------

    More specifically, the Alliance commented that the analysis 
appropriately restricted the application of some technologies, like the 
application of low rolling resistance tires on performance vehicles, 
and limited aerodynamic improvements for trucks and minivans.\673\ 
Similarly, the Alliance commented in support of the decision to exclude 
HCR2 technology from the analysis, citing previous comments stating 
that ``the inexplicably high benefits ascribed to this theoretical 
combination of technologies has not been validated by physical 
testing.''
---------------------------------------------------------------------------

    \673\ NHTSA-2018-0067-12073, at 134.
---------------------------------------------------------------------------

    Ford commented more broadly that ``[t]he previous analyses 
performed by the Agencies too often selected technology benefits from 
the high-end of the forecasted range, and cost from the lower-end, in 
part because deference was given to supplier or other third-party 
claims over manufacturers' estimates.'' \674\ Ford noted that, 
``[m]anufacturer estimates, while viewed as conservative by some, are 
informed by years of experience integrating new technologies into 
vehicle systems in a manner that avoids compromising other important 
attributes (NVH, utility, safety, etc.),'' continuing that ``[t]he need 
to preserve these attributes often limits the actualized benefit of a 
new technology, an effect insufficiently considered in projections from 
most non-OEM sources.'' Ford concluded, as mentioned above, that the 
NPRM analysis better balanced these considerations.
---------------------------------------------------------------------------

    \674\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    Toyota commented that the discrepancy between the automotive 
industry and prior regulatory assessments stemmed from ``agency 
modeling relying on overly optimistic assumptions about technology cost 
effectiveness and deployment rates.'' \675\ Toyota pointed to a prior 
analysis that projected compliance for Toyota's MY 2025 lineup using 
the ALPHA model as an example of how ``the agency's analysis failed to 
account for customer requirements (cost, power, weight-adding options, 
etc.) that erode optimal fuel economy, and normal business 
considerations that govern the pace of technology deployment.'' In 
contrast, Toyota stated that the ``[m]odeled technology cost, 
effectiveness, and compliance pathways in the proposed rulemaking rely 
on more recent data as well as more realistic assumptions about the 
level of technology already on the road today, the pace of technology 
deployment, and trade-offs between vehicle efficiency and customer 
requirements.''
---------------------------------------------------------------------------

    \675\ NHTSA-2018-0067-12150.
---------------------------------------------------------------------------

    Honda, in its feedback on the models used in the standard setting 
process, commented that ``the current version of the CAFE model is 
reasonably accurate in terms of technology efficiency, cost, and 
overall compliance considerations, and reflects a notable improvement 
over previous agency modeling efforts conducted over the past few 
years.'' \676\
---------------------------------------------------------------------------

    \676\ NHTSA-2018-0067-11818.
---------------------------------------------------------------------------

    FCA commented in recognition of the CAFE model improvements over 
the Draft TAR version, but noted they ``continue to believe that the 
cost and benefits used as inputs to the model are overly optimistic.'' 
\677\ FCA used its updated Jeep Wrangler Unlimited and Ram 1500 pickup 
models as examples of vehicles that ``provide real life examples of the 
costs and benefits that can be achieved with fuel and weight saving 
technology;'' however, ``after all of the real world concerns such as 
emissions, drivability, OBD, and fuels are considered, the benefits 
observed remain less than those derived by the Autonomie model and used 
as inputs to the Volpe model.''
---------------------------------------------------------------------------

    \677\ NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    Conversely, environmental groups, consumer groups, and some States 
and localities commented that the Draft TAR and subsequent EPA analyses 
were more representative of the current state of vehicle technologies. 
These groups all generally commented, in different terms, that the NPRM 
analysis technology effectiveness was understated and technology costs 
were overstated, and additional constraints the agencies placed on the 
analysis, like excluding technologies already in production or 
constraining technology pathways, also helped lead to that result.\678\
---------------------------------------------------------------------------

    \678\ NHTSA-2018-0067-11873; NHTSA-2018-0067-11984.
---------------------------------------------------------------------------

    ICCT commented that the agencies ``ignored their own rigorous 2015-
2017 technological assessment, and have adopted a series of invalid and 
unsupportable decisions which artificially constrain the availability 
and dramatically under-estimate levels of effectiveness of many 
different fuel economy improvement and GHG-reduction technologies and 
unreasonably increase modeled compliance costs.'' \679\ ICCT also 
commented that the agencies ignored, suppressed, dismissed, or 
restricted the use of work done to update technologies and technology 
cost and effectiveness assessments since the 2012 final rule for MYs 
2017-2025. ICCT stated that the ``invalid high cost result [of the 
modeled augural standards in 2025] was created by the agencies by 
making many dozens of unsupported changes in the technology 
effectiveness and availability inputs, the technology cost inputs, and 
the technology package constraints.''

[[Page 24382]]

ICCT stated that ``the agencies failed to capture the latest available 
information and, as a result, their assessment incorrectly and 
artificially overstates technology costs.''
---------------------------------------------------------------------------

    \679\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    CARB commented that the agencies did not present sufficient new 
evidence to change previous technical findings, specifically in regards 
to conventional vehicle technologies.\680\ CARB stated that instead of 
relying on new information, as had been asserted as justification for 
the proposal, the analysis was based on older data that did not reflect 
current technology. Accordingly, CARB pointed out that previous 
analysis by the agencies projected far less need for electrification 
than what was required in the proposal, stating that the underlying 
cause is a reduction in the assumed cumulative improvements for what 
advanced gasoline technology is able to achieve.
---------------------------------------------------------------------------

    \680\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    A coalition of States and Cities similarly commented that ``[t]he 
Agencies' conclusions regarding the technology necessary to meet the 
2025 standards and the cost of that technology run counter to the 
evidence before the agency, diverge from prior factual findings without 
explanation and without transparency as to the source of data relied 
on, and are unsupported by any reasoned analysis. Such analysis bears 
many hallmarks of an arbitrary and capricious action.'' \681\
---------------------------------------------------------------------------

    \681\ NHTSA-2018-0067-11735 (citing State Farm, 463 U.S. at 43; 
Fox Television, 556 U.S. at 515; Humane Soc. of U.S. v. Locke, 626 
F.3d 1040, 1049 (9th Cir. 2010)).
---------------------------------------------------------------------------

    Roush Industries, commenting on behalf of CARB, commented that 
``the 2018 PRIA projected average costs for technology implementation 
to achieve the existing standards to be significantly overstated and in 
conflict with the 2016 Draft TAR cost estimates generated by the 
Agencies only two years earlier.'' \682\ Roush commented that the Draft 
TAR analyses of cost and incremental fuel economy improvement necessary 
to achieve the augural standards was consistent with Roush's own 
estimates and other published data.
---------------------------------------------------------------------------

    \682\ NHTSA-2018-0067-11984.
---------------------------------------------------------------------------

    Similarly, H-D Systems (HDS), commenting on behalf of the 
California DOJ, commented that ``the estimates in the 2016 TAR on 
technology cost and effectiveness still represent the correct estimates 
based on the latest available data.'' \683\ HDS, in its analysis of the 
costs of technologies to meet different potential standards between the 
Draft TAR and the NPRM, noted that ``costs for most conventional (i.e., 
non-electric) drivetrain technologies were similar in both reports in 
that costs were within +5% of the average of the costs from the two 
reports. The only exception was the cost estimate for the High CR 
second generation Atkinson cycle or HCR2 engine which was estimated to 
be much more expensive. Due to differences in nomenclature, 
transmission technology costs could not be directly compared but were 
similar at the highest efficiency level. In contrast, cost of hybrid 
technology was estimated to be much higher in the PRIA and were 200 to 
250% higher for strong hybrids. Costs of drag reduction, rolling 
resistance reduction and auxiliary system technologies were also quite 
similar but the cost of mass reduction was substantially higher in the 
PRIA by a factor of 2 to 3. Costs of engine friction reduction appear 
not to be included in the cost computation for the PRIA although the 
technology appears to be integrated into some of the engine technology 
packages analyzed in the PRIA to estimate effectiveness.''
---------------------------------------------------------------------------

    \683\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    CFA commented that ``[t]he overarching discussion of technology 
developments that introduces the NHTSA analysis is fundamentally flawed 
and infects the entire proposal,'' taking issue with the NPRM statement 
that ``some options considered in the original order for the National 
Program ha[d] not worked out as EPA/NHTSA anticipated.'' \684\ CFA 
commented that the agencies failed to note that some technology options 
have performed better than anticipated, and ``the fact that some 
technologies have done better than expected is a basis for increasing 
the standards, not in the context of a mid-term review that was 
supposed to tweak the long-term program.''
---------------------------------------------------------------------------

    \684\ NHTSA-2018-0067-12005.
---------------------------------------------------------------------------

    NCAT commented that the ``inflation of projected technology costs 
does not appear to be attributable primarily to the projected cost of 
any given technology, but rather to modeling constraints on the 
application of such technologies to vehicles. Many of these constraints 
appear to be arbitrary and NHTSA's departure from prior analyses in 
these respects is not adequately supported.'' \685\
---------------------------------------------------------------------------

    \685\ NHTSA-2018-0067-11969.
---------------------------------------------------------------------------

    Environmental groups and States also commented that the agencies 
either should reincorporate all the Draft TAR or the EPA Proposed and 
Final Determination analyses' technologies, technology effectiveness 
values, and technology costs into the analysis, and/or compare the 
final rule analysis with those prior analyses to show how the updated 
assumptions changed the results from those prior analyses.
    For example, ICCT commented that ``[f]or the agencies to conduct a 
credible regulatory assessment they must remove all the technology 
availability constraints, re-incorporate and make available the full 
portfolio of technology options as was available in EPA's analysis for 
the original 2017 Final Determination, and include at least 15 g/mile 
CO2 for off-cycle credits by 2025, to credibly reflect the 
real-world technology developments in the auto industry.'' \686\ ICCT 
also stated that ``[t]he agencies need to identify each and every 
technology cost input used in their modeling, and provide a clear 
engineering and evidence based justification for why that cost differs 
from the costs employed in the extremely well documented and well 
justified Draft TAR and in EPA's 2016 TSD and 2017 Final Determination, 
taking into account the above discussion of significant new evidence 
developed since those prior estimates were made. Absent such disclosure 
and justification, the default assumption needs to be that the prior 
costs estimated based on the most recent data are more appropriate than 
the estimates used for the proposal.''
---------------------------------------------------------------------------

    \686\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    In addition, groups of commenters were equally split on the ability 
of technologies to meet different compliance targets. For example, the 
Alliance commented that ``the only technologies that have demonstrated 
the improvements necessary to meet the MY 2025 standards are strong 
hybrids, plug-in electric vehicles, and fuel cell electric vehicles. 
The Agencies' analysis for this Proposed Rule predict the need for 
significant growth in sales of electrified vehicles, a finding 
consistent with third-party analyses.'' \687\ In contrast, UCS 
commented that electrified powertrains ``are not especially relevant 
for the MY 2022-2025 regulations.'' \688\
---------------------------------------------------------------------------

    \687\ NHTSA-2018-0067-Alliance at 15.
    \688\ NHTSA-2018-0067-UCS at 23.
---------------------------------------------------------------------------

    The agencies are aware that the prior analyses concluded that 
compliance with the augural standards could largely be met through 
advances in gasoline vehicle technologies, and with only very low 
levels of strong hybrids and electric vehicles. As the agencies stated 
in the NPRM, consistent with both agencies' statutes, the proposal was 
entirely de novo, based on an entirely new analysis reflecting the best 
and most up-to-date information available to the agencies at the time 
of this rulemaking.\689\ As discussed in Section IV, Section VI.B, and 
further below, the NPRM and final rule analyses reflect updates to

[[Page 24383]]

technology effectiveness estimates, technology costs, and the 
methodology for applying technologies to vehicles that the agencies 
believed better represent the state of technology and the associated 
costs compared to prior analyses, that result in pathways to compliance 
that look both similar and different to those in prior analyses.
---------------------------------------------------------------------------

    \689\ 83 FR 42897.
---------------------------------------------------------------------------

    That said, several of the effectiveness and cost values used in the 
NPRM and final rule analysis were directly carried over from the 2012 
rule for MYs 2017-2025, Draft TAR, and EPA Midterm Evaluation 
analyses.\690\ Several others were carried over from the 2015 NAS 
report,\691\ which the agencies heavily relied upon in past analyses 
even if specific cost or effectiveness values were not used. Different 
technology effectiveness estimates, cost estimates, or adoption 
constraints were employed where the agencies had information, from 
technical reports, manufacturers, or other stakeholders, indicating 
that a technology could or could not be feasibly adopted in the 
rulemaking timeframe, or a technology could or could not be adopted in 
the way that the agencies had previously modeled it. Notably, most 
differences in pathways to compliance are attributable to only a few 
significant differences between this rulemaking analysis and prior 
rulemaking analyses.
---------------------------------------------------------------------------

    \690\ See, e.g., PRIA at 449, 451, 452, 453, 458.
    \691\ See, e.g., PRIA at 358-360.
---------------------------------------------------------------------------

    For example, as discussed in Section VI.B.3 Technology 
Effectiveness and Modeling and Section VI.C.1 Engine Paths, in the EPA 
Draft TAR and Proposed Determination analyses, effectiveness of HCR 
engine technologies and downsized turbocharged engine technologies were 
estimated using Tier 2 certification fuel. Tier 2 certified fuel has a 
higher octane rating compared to regular octane 
fuel.692 693 694 As summarized by EPA in the PD TSD, ``EPA's 
estimate of effectiveness for gasoline-fueled engines and engine 
technologies was based on Tier 2 Indolene fuel although protection for 
operation in-use on Tier 3 gasoline (87 AKI E10) was included in the 
analysis of engine technologies considered both within the Draft TAR 
and Proposed Determination. Additionally, in the technology assessment 
for this Proposed Determination, EPA has considered the required engine 
sizing and associated effectiveness adjustments when performance 
neutrality is maintained on 87AKI gasoline typical of real-world use.'' 
\695\
---------------------------------------------------------------------------

    \692\ Draft TAR at 5-228.
    \693\ Tier 2 fuel has an octane rating of 93. Typical regular 
grade fuel has an octane rating of 87 ((R+M)/2 octane.
    \694\ EPA Proposed Determination TSD at 2-209 to 2-212.
    \695\ EPA Proposed Determination TSD at 2-210.
---------------------------------------------------------------------------

    NHTSA's effectiveness analysis for the Draft TAR used some engine 
maps also developed using premium octane gasoline. However, at the time 
NHTSA stated the agency would ensure all future engine model 
development will be performed with regular grade octane gasoline.\696\ 
Commenters like Ford stated the effectiveness estimates for turbo 
downsized engine packages were too high, in part because of the use of 
high octane fuel. However they also commented in appreciation of 
NHTSA's acknowledgement that any subsequent analysis would be based on 
fuel at an appropriate octane level, as they stated the impact of the 
change needed to be reflected in future analyses.\697\
---------------------------------------------------------------------------

    \696\ Draft TAR at 5-504, 5-512.
    \697\ Ford Motor Company Response to the Draft TAR September 26, 
2016 NHTSA-2016-0068-0048, at 4.
---------------------------------------------------------------------------

    Engine specifications used to create the engine maps for the NPRM 
and the final rule analysis were developed using Tier 3 fuel to assure 
the engines were capable of operating on real world regular octane (87 
pump octane = (R+M/2)). The process was similar to what manufacturers 
must do to ensure engines have acceptable noise, vibration, harshness, 
drivability, performance, and will not fail prematurely when operated 
on regular octane fuel. This eliminated the need for any adjustments 
that were applied in the 2016 Draft TAR and PD TSD to account for Tier 
2 to Tier 3 fuel properties. This accounts for some of the 
effectiveness and cost differences for engine technologies between the 
Draft TAR/Proposed Determination and the NPRM/final rule. For more 
details, see Section VI.C.1 Engine Paths.
    The agencies believe ICCT's and other commenters' assertions that 
the engine maps should reflect Tier 2 fuel and not be updated for Tier 
3 fuel would ignore these important considerations, and would provide 
engine maps that could not achieve the fuel economy improvements unless 
operated on high octane fuel. Therefore, the agencies determined that 
engine maps developed for the Draft TAR and EPA Proposed Determination 
that were based on Tier 2 fuel should not be used for the NPRM and 
final rule analyses for these technical reasons.
    As another related example, the agencies described that prior 
analyses had relied heavily on the availability of the HCR2 (or ATK2) 
``future'' Atkinson Cycle engine as a cost-effective pathway to 
compliance for stringent alternatives, but many engine experts 
questioned its technical feasibility and near-term commercial 
practicability.\698\ The agencies explained that EPA staff began 
theoretical development of this conceptual engine with a best-in-class 
2.0L Atkinson cycle engine and then increased the efficiency of the 
engine map further, through the theoretical application of additional 
technologies in combination, including cylinder deactivation, engine 
friction reduction, and cooled exhaust gas recirculation. While the 
potential of such an engine is interesting, nevertheless the engine 
remains entirely speculative. No production HCR2/ATK2 engine, as 
outlined in the EPA SAE paper,\699\ has ever been commercially 
produced. Furthermore, the engine map has not been validated with 
hardware, bench data, or even on a prototype level (as no such engine 
exists to test to validate the engine map).
---------------------------------------------------------------------------

    \698\ 83 FR 43038.
    \699\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy 
Improvements from the Implementation of cEGR and CDA on an Atkinson 
Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017. Available at 
https://doi.org/10.4271/2017-01-1016.
---------------------------------------------------------------------------

    Vehicle manufacturers also commented on EPA's effectiveness 
assumptions and estimates of HCR2/ATK2 model's future penetration 
levels in the Draft TAR, stating ``[t]he effectiveness values for the 
`futured' ATK2 package--projected at 40% penetration in 2025MY and 
includes cooled exhaust gas recirculation (CEGR) and cylinder 
deactivation (DEAC)--are too high, primarily due to overtly-optimistic 
efficiencies in the base engine map, insufficient accounting of CEGR 
and DEAC integration losses, and no accounting of the impact of 91RON 
Tier 3 test fuel,'' and that ``44% fleet-wide penetration of ATK2 in 
2025MY is unrealistic given the limited number of powertrain refresh 
cycles available before 2025MY. In addition, it is unreasonable to 
assume that OEMs already heavily invested in different high-efficiency 
powertrain pathways (e.g., turbo-downsizing) would be able to commit 
the immense resources needed to reach these high ATK2 penetration 
levels in such a short time.'' \700\
---------------------------------------------------------------------------

    \700\ Ford Motor Company Response to the Draft TAR September 26, 
2016 NHTSA-2016-0068-0048, at 4.
---------------------------------------------------------------------------

    Accordingly, the agencies decided to not include HCR2 technology in 
the NPRM and final rule analysis. The engine model was not used because 
no observable physical demonstration of the speculative technology 
combination model has yet been created. Further,

[[Page 24384]]

many questions remain about the model's practicability as specified, 
especially in high load, low engine speed operating conditions. The 
HCR2 model combines multiple technologies to provide cumulative 
estimate of benefits without consideration the practical interaction of 
technologies. This approach runs contrary to the modeling approach 
attempted in the NPRM and final rule analysis. The approach the 
agencies tried to follow restricted models to adding discrete advanced 
technologies. This approach allowed an accounting of synergetic 
effects, identified incremental benefits, and increased the precision 
of cost estimates.
    As another example, further discussed in Section VI.B.1 Analysis 
Fleet, the agencies had traditionally taken different approaches to 
assigning baseline road load reduction technology assignments. For 
analyzing baseline levels of mass reduction in an analysis fleet, NHTSA 
had developed for the Draft TAR a regression model to summarize a 
vehicle's weight savings using a relative performance approach and 
accounting for vehicle content, using cost curves developed from 
teardown studies of a MY 2011 Honda Accord and MY 2014 Chevrolet 
Silverado pickup truck. EPA developed its own methodology that 
classified vehicles based on weight reductions from a MY 2008 vehicle, 
compared to the MY 2014 version of the same vehicle, using a cost curve 
from a tear-down study of a MY 2010 Toyota Venza. In the EPA's mass 
reduction technology costing approach, a cost reduction was applied 
when mass reduction 1 technology was applied to a system at mass 
reduction 0 technology level. NHTSA's approach, used in the NPRM and 
final rule analysis, set baseline mass reduction assignments so costs 
of implementing mass reduction technologies are fully applied as 
vehicle platforms move along the mass reduction technology path.
    The agencies also included additional advanced powertrain 
technologies and other vehicle-level technologies in the technology 
pathways between the Draft TAR and NPRM, and between the NPRM and final 
rule. However, manufacturers and suppliers have repeatedly told the 
agencies that there are diminishing returns to increasing the 
complexity of advanced gasoline engines, including in the amount of 
fuel efficiency benefit that they can provide. For example, Toyota 
commented, in response to the EPA SAE paper benchmarking the 2018 Camry 
with the 2.5L Atkinson-cycle engine and ``futuring'' midsize exemplar 
vehicles based on the generated engine map,\701\ that although EPA's 
addition of cylinder deactivation to the hypothetical 2025 exemplar 
vehicle is technically possible and would provide some fuel economy and 
CO2 benefit, the primary function of cylinder deactivation 
is to reduce engine pumping losses which the Atkinson cycle and EGR 
already accomplish on the 2018 Camry.\702\ Toyota concluded, ``The 
overlapping and redundant measures to reduce engine pumping losses 
would add costs with diminishing efficiency returns.'' Similarly, 
BorgWarner commented that they ``do not expect that variable 
compression ratio (VCR) or homogeneous charge compression ignition 
(HCCI) will see broad application in the short term, if ever. While 
each of these technologies can offer marginal efficiency gains at some 
engine speed-load conditions, the use of down-sized boosted engines 
with 8-10 speed transmissions makes it possible to run engines at near 
optimum conditions and effectively minimizes gains from VCR or HCCI. 
VCR mechanisms result in additional mass, cost and complexity, and true 
HCCI has yet to be demonstrated in a production vehicle. The agencies 
do not believe that OEMs will judge these technologies to be cost 
effective.'' \703\
---------------------------------------------------------------------------

    \701\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al., 
``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine 
with Cooled-EGR,'' SAE Technical Paper 2019-01-0249, 2019, 
doi:10.4271/2019-01-0249.
    \702\ NHTSA-2018-0067-12431, at 8.
    \703\ NHTSA-2018-0067-11895.
---------------------------------------------------------------------------

    So, while previous analyses may have shown pathways to compliance 
with increasingly complex advanced gasoline engines, the NPRM and final 
rule analyses more appropriately reflect that the most complex gasoline 
engine technologies will account for a smaller share of manufacturers' 
products during the rulemaking timeframe. However, despite this fact, 
the NPRM and final rule analysis include more advanced powertrain 
technologies than previous analyses, in part to account for important 
considerations like intellectual property and the fact that some 
manufacturers have already started down the path of incorporating a 
certain advanced engine technology in their product portfolio, and that 
abrupt switching to another advanced engine technology would result in 
unrealistic stranding of capital costs. In addition, greater precision 
in how cumulative technologies applied to engines, as estimated through 
the Autonomie effectiveness modeling, appropriately reflects the 
diminishing returns to efficiency benefits that those advanced engines 
can provide. Moreover, as identified by a wide range of commenters, 
battery costs are projected to fall in the rulemaking timeframe to a 
point where, in the compliance modeling, it becomes more cost effective 
to add electrification technologies to vehicles than to apply other 
advanced gasoline engine technologies.
    Finally, the agencies declined to incorporate some information and 
data for the NPRM or final rule central analysis for reasons discussed 
in the following sections. In general, the data produced by agencies or 
submitted by commenters failed to isolate effectiveness impacts of 
individual technologies (or in some cases a combination of two or 
several technologies). The data included effects from additional 
unaccounted and undocumented technologies. Because the effectiveness 
improvement measured or claimed resulted from more than just the 
reported sources, the actual effectiveness of the technology or 
technologies is obfuscated and easily under or over predicted. Using 
effectiveness values generated in this manner carries a high risk of 
double counting effectiveness and undercounting costs.
    In many cases, this problem exists where data or information is 
based on laboratory testing or on-road testing of production vehicles 
or components including engines and transmissions. Production vehicles 
and components usually include multiple technology improvements from 
one redesign to the next, and rarely incorporate just a single 
technology change. Furthermore, technology improvements on production 
vehicles in some cases cannot be readily observed, such as the level of 
mechanical friction in an engine, and isolation and identification of 
the improvement attributable to each technology would be impractical 
given the costs and time required to do so. That said, in some cases, 
where possible to do so, the agencies used the data or information from 
production vehicles to corroborate information from the Autonomie 
simulations. However, the agencies declined to apply that data or 
information directly in the analysis if the effectiveness improvement 
attributable to a particular technology could not be isolated.
    The agencies made these updates from prior analyses not, as some 
commenters have suggested, to ``artificially overstate technology 
costs,'' \704\ or to ``ignore the knowledge and expertise of the EPA 
engineering

[[Page 24385]]

and compliance staff,'' \705\ ``so that the model in many instances 
selects more expensive, less fuel efficient technology while excluding 
less expensive and more efficient alternatives,'' \706\ but because the 
updates reflected the agencies' reasonable assessment of the current 
state of vehicle technologies and their costs, and the state of future 
vehicle technologies and costs in the rulemaking timeframe.
---------------------------------------------------------------------------

    \704\ NHTSA-2018-0067-11741 at 7.
    \705\ NHTSA-2018-0067-11741 at I-23.
    \706\ NHTSA-2018-0067-12123.
---------------------------------------------------------------------------

    Separate from the decision to update assumptions used for the NPRM 
analysis from prior analyses, the agencies did refine some technology 
effectiveness and cost assumptions from the NPRM to this final rule 
analysis. In addition to being appropriate for technical reasons, this 
should address some commenters' overarching concerns about understated 
technology effectiveness and overstated technology costs. For example, 
several commenters noted that the costs of BISG/CISG systems were 
higher for small Cars/SUVs and medium cars than for medium SUVs and 
pickup trucks, which the Alliance and FCA described as ``implausible'' 
and ``misaligned with industry understanding,'' and which ICCT 
described as ``contrary to basic engineering logic, which holds that a 
system which would be smaller and have lower energy and power 
requirements would be less expensive, not more.'' \707\ The agencies 
agree, and have made changes to address this issue, as described in 
Section VI.C.3.a) Electrification.
---------------------------------------------------------------------------

    \707\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    After considering comments, the agencies also added several engine 
technologies and technology combinations for the final rule analysis. 
These included a basic high compression ratio Atkinson cycle engine, a 
variable compression ratio engine, a variable turbo geometry engine, 
and a variable turbo geometry with electric assist engine (VTGe). The 
NPRM discussed and provided engine maps for each of these technologies. 
The agencies also added new technology combinations including diesel 
engines with cylinder deactivation, turbocharged engines with advanced 
cylinder deactivation, diesel engines paired with manual transmissions, 
and diesel engines paired with 12-volt start-stop technology. 
Transmission revisions included updating the effectiveness of 6-speed 
automatic transmissions, applying updated shift logic for 10-speed 
automatic transmissions, and increasing the gear span for efficient 10-
speed automatic transmissions. Mass reduction technology was expanded 
to include up to 20 percent curb weight reduction, compared with up to 
10 percent for the NPRM. These changes, and the comments upon which 
they were based, are described in further detail in the following 
sections.
1. Engine Paths
    The internal combustion (IC) engine is a heat engine that converts 
chemical energy in a fuel into mechanical energy. Chemical energy of 
the fuel is first converted to thermal energy by means of combustion or 
oxidation with air inside the engine. This thermal energy raises the 
temperature and pressure of the gases within the engine, and the high-
pressure gas then expands against the internal mechanisms of the 
engine. This expansion is converted by the mechanical linkages of the 
engine to a rotating crankshaft, which is the output of the engine. The 
crankshaft, in turn, is connected to a transmission to transmit the 
rotating mechanical energy to the desired final use, particularly the 
propulsion of vehicles.
    IC engines can be categorized in a number of different ways 
depending upon which technologies are designed into the engine: By type 
of ignition (e.g., spark ignition or compression ignition), by engine 
cycle (e.g., Otto cycle or Atkinson cycle), by valve actuation (e.g., 
overhead valve (OHV), single overhead camshaft (SOHC), or dual overhead 
camshaft (DOHC)), by basic design (e.g., reciprocating or rotary), by 
configuration and number of cylinders (e.g., inline four-cylinder (I4) 
or V-shaped six-cylinder (V6)), by air intake (e.g., forced induction 
(turbo or super charging) or naturally aspirated), by method of fuel 
delivery (e.g., port injection or direction injection), by fuel type 
(e.g., gasoline or diesel), by application (e.g., passenger car or 
light truck),or by type of cooling (e.g., air-cooled or water-cooled). 
For each combination of technologies among the various categories, 
there is a theoretical maximum efficiency for all engines within that 
set. There are various metrics that can be used to compare engine 
efficiency, and the four metrics the agencies use or discuss in this 
preamble are:
     Brake specific fuel consumption (BSFC), which is the mass 
of fuel consumed per unit of work output (amount of fuel used to 
produce power);
     Brake thermal efficiency (BTE), which is the total fuel 
energy released per unit of work output (percentage of fuel used to 
produce power);
     Fuel consumption (gallons per mile), which looks at the 
gallons of fuel consumed per unit of work output (mile travelled); and
     Fuel economy (in MPG), which is the amount of work output 
(miles travelled) per unit (gallon) of fuel consumed.
    When comparing the efficiency of IC engines, it is important to 
identify the metric(s) used and the test cycle for the measurement 
because results vary widely when engines operate over different test 
cycles. Two-cycle fuel economy tests used to certify vehicles' 
compliance with the CAFE standards tend to overestimate the average 
fuel economy motorists will typically achieve during on-road 
operation.\708\ In the NPRM and for this final rule analysis, the 
agencies considered technology effectiveness for the 2-cycle test 
procedures and AC and off-cycle test procedures to evaluate how 
technologies could be applied for manufacturers to comply with 
standards. The agencies also considered real world operation beyond 
these test procedures when considering IC engine technologies in order 
to assure the technologies were configured and specified in a manner 
that could be used in real world vehicle applications.
---------------------------------------------------------------------------

    \708\ 77 FR 62988.
---------------------------------------------------------------------------

a) Fuel Octane
    As mentioned in other sections of the Preamble, the agencies go to 
great lengths to ensure engine technologies considered for potential 
compliance pathways are feasible for real-world implementation and 
effectiveness. An important facet of this evaluation are both the fuels 
that are used for efficiency testing and also the fuels that consumers 
may purchase in the marketplace.
    In the NPRM, the agencies included a general overview of fuel 
octane (stability) level, including levels currently available, and the 
potential impact of fuel octane on engines developed for the U.S. 
market.\709\ The agencies described that a typical, overarching goal of 
optimal spark-ignited engine design and operation is to maximize the 
greatest amount of energy from the fuel available, without manifesting 
detrimental impacts to the engine over expected operating conditions. 
Design factors, such as compression ratio, intake and exhaust value 
control specifications, and combustion chamber and piston 
characteristics, among others, are all impacted by the octane of the 
fuel consumers are anticipated to use.\710\
---------------------------------------------------------------------------

    \709\ PRIA at 253.
    \710\ In addition, PRIA Chapter 6 contains a brief discussion of 
fuel properties, octane levels used for engine simulation and in 
real-world testing, and how octane levels can impact performance 
under these test conditions.

---------------------------------------------------------------------------

[[Page 24386]]

    The agencies also discussed potential challenges associated with 
octane levels available currently, and how those octane levels may play 
a role in potential vehicle fuel efficiency improvements. Vehicle 
manufacturers typically develop their engines and engine control system 
calibrations based on the fuel available to consumers. In many cases, 
manufacturers may recommend a fuel grade for best performance and to 
prevent potential damage. In some cases, manufacturers may require a 
specific fuel grade for both best performance, to achieve advertised 
power ratings, and/or to prevent potential engine damage.
    Consumers, though, may or may not choose to follow the 
manufacturer's recommendation or requirement for a specific fuel grade 
for their vehicle. As such, vehicle manufacturers often choose to 
employ engine control strategies for scenarios where the consumer uses 
a lower than recommended, or required, fuel octane level, as a way to 
mitigate potential engine damage over the life of a vehicle. These 
strategies limit the extent to which some efficiency improving engine 
technologies can be implemented, such as increased compression ratio 
and intake system and combustion chamber designs that increase burn 
rates and rate of in-cylinder pressure rise. If the minimum octane 
level available in the market were higher (especially the current sub-
octane regular grade in the mountain states), vehicle manufacturers 
might not feel compelled to design vehicles sub-optimally to 
accommodate such blends.
    When knock (also referred to as detonation) is encountered during 
engine operation, at the most basic level, non-turbocharged engines can 
adjust the timing of the spark that ignites the fuel, as well as the 
amounts of fuel injected at each intake stroke (``fueling''). In 
turbocharged applications, knocking is typically controlled by 
adjusting boost levels along with spark timing and/or the amount of 
fuel injected. Past rulemakings discussed other techniques that may be 
employed to allow higher compression ratios, including optimizing spark 
timing, and adding of cooled exhaust gas recirculation (EGR). 
Regardless of the type of spark-ignition engine or technology employed, 
efforts to reduce or prevent knock with the lower-octane fuels that are 
available in the market result in the loss of potential power output, 
creating a ``knock-limited'' constraint on performance and efficiency.
    The agencies noted that despite limits imposed by available fuel 
grades, manufacturers continue to make progress in extracting more 
power and efficiency from spark-ignited engines. Production engines are 
safely operating with regular 87 AKI fuel with compression ratios and 
boost levels once viewed as only possible with premium fuel. According 
to the Department of Energy, the average gasoline octane level has 
remained fundamentally flat starting in the early 1980's and decreased 
slightly starting in the early 2000s. During this time, however, the 
average compression ratio for the U.S. fleet has increased from 8.4 to 
10.52, a more than 20 percent increase. As explained by the Department 
of Energy, ``[t]here is some concern that in the future, auto 
manufacturers will reach the limit of technological increases in 
compression ratios without further increases in the octane of the 
fuel.'' \711\ As such, manufacturers are still limited by the fuel 
grades available to consumers and the need to safeguard the durability 
of their products for all of the available fuels; thus, the potential 
improvement in the design of spark-ignition engines continues to be 
overshadowed by the fuel grades available to consumers.
---------------------------------------------------------------------------

    \711\ Fact of the Week, Fact #940: August 29, 2016 Diverging 
Trends of Engine Compression Ratio and Gasoline Octane Rating, U.S. 
Department of Energy, https://www.energy.gov/eere/vehicles/fact-940-august-29-2016-diverging-trends-engine-compression-ratio-and-gasoline-octane (last visited Mar. 21, 2018).
---------------------------------------------------------------------------

    EPA and NHTSA also described ongoing research and positions from 
automakers and advocacy groups on fuel octane levels, including 
comments received during past agency rulemakings and on the 2016 Draft 
TAR regarding the potential for increasing octane levels in the U.S. 
market. The agencies described arguments for adjusting to octane 
levels, including making today's premium grade the base grade of fuel 
available, which could enable low cost design changes to improve fuel 
economy and reduce tailpipe CO2 emissions. Challenges 
associated with this approach include the increased cost to consumers 
who drive vehicles designed for current regular octane grade fuel, who 
would not benefit from the use of the higher cost higher-octane fuel. 
The costs of such a transition to higher-octane fuel would be high and 
persist well into the future, since unless current regular octane fuel 
were unavailable in the North American market, manufacturers would be 
effectively unable to redesign their engines to operate on higher-
octane fuel. In addition, the full benefits of such a transition would 
not be realized until vehicles with such redesigned engines were 
produced for a sufficient number of model years largely to replace the 
current on-road vehicle fleet. The transition to net positive benefits 
would take many years.
    The agencies also described input received from renewable fuel 
industry stakeholders and from the automotive industry supporting high-
octane gasoline fuel blends to enable fuel economy and CO2 
improving technologies such as higher compression ratio engines. 
Stakeholders suggested that mid-level (e.g., E30) high-octane ethanol 
blends should be considered and that EPA should consider requiring that 
mid-level blends be made available at service stations. Stakeholders 
supporting higher-octane blends suggested that higher-octane gasoline 
could provide auto manufacturers with more flexibility to meet more 
stringent standards by enabling opportunities for use of lower tailpipe 
CO2 emitting technologies (e.g., higher compression ratio 
engines, improved turbocharging, optimized engine combustion).
    The agencies sought additional comment in the NPRM on various 
aspects of current fuel octane levels and how fuel octane could play a 
role in the future. More specifically, the agencies sought comment on 
how increasing fuel octane levels could have an impact on product 
offerings and engine technologies, as well as what improvements to fuel 
economy and tailpipe CO2 emissions could result from higher-
octane fuels. The agencies sought comment on an ideal octane level for 
mass-market consumption, and whether there were downsides with 
increasing the available octane levels and, potentially, eliminating 
lower-octane fuel blends. EPA also requested comment on whether and how 
EPA could require the production and use of higher-octane gasoline 
consistent with Title II of the Clean Air Act.
    The agencies received numerous, wide-ranging comments in response 
to the NPRM discussion, and some direct responses to the agencies' 
requests for comments. The commenters included fuel producers, 
individual vehicle manufactures, environmental groups, vehicle 
suppliers, fuel advocacy groups, and agricultural organizations, among 
others. Commenters provided a broad range of comments ranging from 
explication of the many challenges to increasing available octane 
levels, to claims of the substantial efficiency

[[Page 24387]]

increases that could be easily obtained by requiring higher-octane 
levels.
    Several ethanol industry stakeholders commented in support of 
requiring higher-octane fuels using mid-level ethanol blends. The High-
Octane, Low Carbon (HOLC) Alliance commented that it believes ``NHTSA 
and EPA have a critical opportunity to cost-effectively ensure progress 
in fuel efficiency and CO2 emissions standards. Scientific 
experts agree that high-octane, low-carbon fuel can yield greater fuel 
economy and emissions benefits when paired with internal combustion 
engines (ICEs). But, to realize such benefits, automobile manufacturers 
require approval sooner rather than later to such fuels. Alternatively, 
automobile manufacturers will be limited in their ability to maximize 
the environmental performance of their vehicles until non-liquid fuel 
engines become more readily available. In finalizing the Proposed Rule, 
the HOLC Alliance strongly urges EPA and NHTSA to establish a pathway 
forward toward incentivizing the production and adoption of higher-
octane, lower carbon fuels. By doing so, EPA and NHTSA can continue to 
incrementally increase CO2 and fuel economy standards, 
respectively.'' \712\
---------------------------------------------------------------------------

    \712\ HOLC Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
4196.
---------------------------------------------------------------------------

    Renewable Fuels Associations (RFA) commented that ``it strongly 
believes vehicles and fuels must be considered together as integrated 
systems. As EPA has recognized in the past, a `systems approach enables 
emission reductions that are both technologically feasible and cost 
effective beyond what would be possible looking at vehicle and fuel 
standards in isolation.' Because ethanol-based high-octane low-carbon 
fuel blends would enable cost-effective gains in fuel economy and 
carbon dioxide reductions, the agencies should take steps to support 
[high-octane low-carbon] fuels in the final SAFE rule.'' \713\
---------------------------------------------------------------------------

    \713\ RFA, Detailed Comments, EPA-HQ-OAR-2018-0283-4409.
---------------------------------------------------------------------------

    RFA cited several studies indicating benefits are available from 
raising the floor of fuel octane levels currently available, and, 
particularly, ``[t]he results from the studies reviewed generally 
support a main conclusion that splash blending ethanol is a highly 
effective means of raising the octane rating of gasoline and enabling 
low-cost efficiencies and reduced emissions in modern spark-ignition 
engines.'' \714\ In addition, National Corn Growers Association stated 
that, ``[w]ithout a change in fuel, automakers are reaching the limits 
on the efficiency gains that can be achieved with technology changes.'' 
\715\
---------------------------------------------------------------------------

    \714\ RFA, Detailed Comments, EPA-HQ-OAR-2018-0283-4409.
    \715\ National Corn Growers Association, https://www.ncga.com/file/1621/NCGA%20Comments20Docket%20No.%20EPA-HQ-OAR-2018-0283%20and%20NHTSA-2018-0067.pdf.
---------------------------------------------------------------------------

    The National Corn Growers Association, in conjunction with 
associated corn growing and agricultural groups, pointedly stated the 
EPA should, ``[s]et a minimum fuel octane level of 98 RON and phase out 
low octane fuels as new optimized vehicles enter the market in MY 
2023,'' and concluded that approving a ``midlevel ethanol blend vehicle 
certification fuel would enable automakers to expedite design and 
testing of optimized vehicles for use with this new fuel.'' \716\
---------------------------------------------------------------------------

    \716\ National Corn Growers Association, https://www.ncga.com/file/1621/NCGA%20Comments%20Docket%20No.%20EPA-HQ-OAR-2018-0283%20and%20NHTSA-2018-0067.pdf.
---------------------------------------------------------------------------

    The 25x25 Alliance commented that ``to meet the dual goals of 
greater fuel efficiency and reduced GHG emissions, the utilization of 
higher compression spark ignition internal combustion engines will be 
essential. Increasing engine compression improves thermal efficiency. 
However, as compression increases, higher-octane fuels will be needed 
to prevent engine knock. Automakers and advocacy groups have expressed 
support for increases to fuel octane levels for the US market. Ethanol 
with its octane rating of 113 offers engine knock resistance at a lower 
cost than any other octane booster in gasoline. In addition, ethanol's 
lower direct and life-cycle GHG emissions as compared to gasoline are 
well documented. For this reason, a fuel produced from a mixture of 
ethanol and gasoline and used in conjunction with advanced high 
compression engines presents itself as a technology pathway capable of 
complying with new CAFE/GHG standards.'' They continue, ``HOLC 
supporters recognize numerous barriers and other associated regulatory 
hurdles must be resolved before HOLC ethanol fuels are adopted at large 
scale. . . 25x25 believes it is imperative that the vehicle and fuel be 
treated as a comprehensive system. To date CAFE/GHG standards have 
largely focused on vehicle engine technology. Advanced engine vehicles 
perform best in concert with fuels of suitable properties and 
composition to optimally enable and power them.'' \717\
---------------------------------------------------------------------------

    \717\ 25x25 Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
4210.
---------------------------------------------------------------------------

    The American Coalition for Ethanol (ACE) commented that ``high-
octane blends comprised of 25 to 30 percent ethanol would help bring 
down the cost for consumers compared to the premium-priced octane level 
advocated by oil refiners. Ethanol has a blending octane rating of 
nearly 113 and trades at a steep discount to gasoline. In many 
wholesale markets today, ethanol costs at least 60 cents per gallon 
less than gasoline. Ethanol delivers the highest octane at the lowest 
cost, allowing automakers to benefit by continuing to develop high-
compression engine technologies and other product offerings to achieve 
efficiency improvements and reduced emissions. The ideal way to 
transition from today's legacy fleet to new vehicles with advanced 
engine technologies designed to run optimally on a high-octane fuel is 
to utilize FFVs as bridge vehicles that can provide immediate demand 
for mid-level ethanol blends.'' \718\
---------------------------------------------------------------------------

    \718\ ACE, Detailed Comments, EPA-HQ-OAR-2018-0283-4033.
---------------------------------------------------------------------------

    Growth Energy commented that with a mid-level ethanol blend, 
automakers not only get higher-octane that they can use to optimize 
engines and gain further fuel efficiency, they will also see a fuel 
that has demonstrably lower carbon dioxide emissions.\719\ The Illinois 
Corn Growers' Association et al., commented that ``NHTSA and EPA must 
adapt the existing regulatory structure to reflect the specific 
characteristics of mid-level blend fuels. Working together, the ethanol 
industry, automakers, EPA and NHTSA can bring about, during the period 
covered by the SAFE program, a new generation of high efficiency 
internal combustion engines optimized to take advantage of this new 
fuel's unique properties.'' \720\
---------------------------------------------------------------------------

    \719\ Growth Energy, Detailed Comments, EPA-HQ-OAR-2010-0799- 
9540-A2.
    \720\ Comment removed because it contains copyrighted data, 
Illinois Corn Growers Association, et al., https://www.regulations.gov/document?D=EPA-HQ-OAR-2018-0283-4198.
---------------------------------------------------------------------------

    Ethanol industry commenters provided comment on several EPA actions 
they believe would be necessary to support higher-octane mid-level fuel 
blends:
     Set a minimum fuel octane level and phase out low-octane 
fuels as new optimized vehicles enter the market;
     Approve a high-octane, mid-level ethanol blend vehicle 
certification fuel;
     Correct the fuel economy formula by updating the R-Factor 
to be at or nearly ``1'' to reflect documented operation of modern 
engine technology;
     Extend a RVP waiver of 1 psi to all gasoline containing at 
least 10 percent ethanol;
     Adopt the Argonne National Laboratory GREET model to 
determine updated lifecycle carbon emissions for ethanol;

[[Page 24388]]

     Establish meaningful credits to automakers to incentivize 
transition to higher-octane fuel vehicles and continue to support flex-
fuel vehicles; and
     Provide equal treatment to vehicle technologies that 
reduce carbon emissions.
    The Clean Fuels Development Coalition, et al. suggested that, ``the 
`ideal octane level' to optimize LDV performance, fuel efficiency, and 
reduce harmful emissions and consumer costs is 98-100 RON produced with 
E30+ `clean octane.' '' \721\ Concurrently, the HOLC Alliance and ACE, 
among others, also supported that 98 to 100 RON would be ideal octane 
levels for the nation.\722\
---------------------------------------------------------------------------

    \721\ Clean Fuels Development Coalition, et al., Detailed 
Comments, NHTSA-2018-0067-11988.
    \722\ HOLC Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
4196; ACE, Detailed Comments, EPA-HQ-OAR-2018-0283-4033.
---------------------------------------------------------------------------

    BorgWarner, a supplier to major automobile manufacturers, commented 
that ``[f]uel octane is a limiting factor in the selection of 
compression ratio for all spark-ignition engines and the amount of 
boost for turbocharged engines. Higher-octane is particularly effective 
for using higher compression ratios with boosted engines,'' and stated 
that ``[t]here is substantial merit to raising the minimum octane 
required because current fuel pricing penalizes consumers for using 
higher-octane fuel. A base octane of 95 RON would be consistent with 
Europe. This would allow consistent development of engines for the 
broader US-EU market. Prior to the introduction of ethanol into 
gasoline, the base blend for regular fuel was typically 92 RON. 
Addition of 10% ethanol to this base blend gave 95 RON regular, so the 
base blend would be reformulated to retain the 92 RON at a lower cost. 
Returning to the previous base blend would be cost effective to the 
consumer.'' \723\
---------------------------------------------------------------------------

    \723\ BorgWarner, Detailed Comments, EPA-HQ-OAR-2018-0283-4174.
---------------------------------------------------------------------------

    Auto manufacturers also provided comment on the topic of higher-
octane fuels. The Alliance of Automobile Manufacturers (the Auto 
Alliance) commented that it ``has long advocated for the availability 
of cost-effective, higher-octane fuel. The Alliance also believes the 
Agencies should require a transition to a higher minimum-octane 
gasoline (minimum 95-98 RON). There are several ways to produce higher-
octane grade gasoline, such as expanding the ethanol availability, but 
the Alliance does not promote any sole or particular pathway.'' \724\ 
The Alliance reiterated its position regarding fuel octane levels 
where, ``[t]he Alliance has long supported two goals regarding the 
octane (anti-knock) properties of gasoline: (1) The availability of 
cost effective higher-octane fuels, greater than 95 Research Octane 
Number (RON) and (2) the immediate elimination of subgrade fuel less 
than 87 anti-knock index (AKI).'' The Alliance also noted that ``[t]he 
higher-octane fuel that is available today is sold as a premium grade. 
To support future engine technologies, the approach taken with today's 
premium fuel option would not be expected to provide an attractive 
value proposition to the customer; therefore, a new higher minimum-
octane gasoline, 95-98 RON, is needed to achieve anticipated 
performance.''
---------------------------------------------------------------------------

    \724\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Ford Motor Company agreed with the Auto Alliance's collective 
comments on fuel octane level and added specific support to raising 
minimum octane levels, stating that ``Ford concurs with those comments 
and supports increasing the marketplace octane rating in the U.S. to a 
minimum of 95 Research Octane Number (RON).'' Ford also generally 
supported the agencies' fuel octane discussion in terms of impacts to 
vehicle performance, where ``[h]igher octane gasoline enables 
opportunities for the use of key energy-efficient technologies, 
including: Higher compression ratio engines, lighter and smaller 
engines, improved turbocharging, optimized engine combustion phasing/
timing, and low temperature combustion strategies. All of these 
technologies paired with higher-octane gasoline permit smaller engines 
to meet the demands of the consumer while at the same time providing 
higher overall efficiencies.'' \725\
---------------------------------------------------------------------------

    \725\ Ford, Detailed Comments, EPA-HQ-OAR-2018-0283-5691.
---------------------------------------------------------------------------

    Volkswagen commented ``[t]here may be several potential ways to 
achieve a high-octane fuel that may be more costly to the vehicle than 
others. Achieving an E10 high-octane fuel may mean a different hardware 
set than on E20 or E30 high-octane fuel. Elimination of sub-grades of 
market fuel (less than 87AKI) quickly is very important. If current 87 
AKI and 85 AKI fuels remain in the market for backward compatibility 
(such as if an E30 were chosen as the high-octane fuel of the future), 
a robust method at the fuel dispensing station and incorporated into 
the fueling station equipment to prevent mis-fueling is necessary. 
However, an E10 high-octane pathway might have far fewer compatibility 
problems and might bring extra fuel economy to the drivers of those 
current vehicles.'' \726\
---------------------------------------------------------------------------

    \726\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    The agencies also received comments from the petroleum industry 
regarding higher-octane fuels. API commented that ``[g]iven the 
multiple engine technology pathways available to the automakers for 
achieving future fuel economy and CO2 emissions targets, the 
challenge of determining future market fuel gasoline octane number 
needs is complex and not yet settled. API believes that the octane 
number issue should be part of a comprehensive transport policy that 
addresses both vehicles and fuels as a system. API and its members are 
engaged in collaborations with the automakers and other stakeholders to 
better understand future fuel requirements for emerging powertrain 
technologies.'' API also commented ``the future for gasoline octane 
number will be driven by the stringency of regulations that set future 
fuel economy and CO2 requirements, the collective responses 
of the automakers to those regulations, consumer preferences regarding 
vehicles and fuels, and fuel supply economics. EPA's authority to 
regulate gasoline octane number is doubtful. Therefore, EPA should not 
attempt to regulate gasoline octane number at this time.'' \727\
---------------------------------------------------------------------------

    \727\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
---------------------------------------------------------------------------

    In terms of challenges associated with potential high-octane fuel 
deployment, the American Fuel & Petrochemical Manufacturers (AFPM) 
commented that, ``[a]side from a lack of legal authority, EPA faces 
numerous technical, logistical, and legal challenges and uncertainties 
in requiring the use of higher-octane fuels. Any such requirement would 
need a separate rulemaking dedicated to such a purpose with an 
extensive technical record in support, including test data on vehicles 
designed for the higher-octane fuel and on the existing fleet with and 
without higher-octane.'' \728\
---------------------------------------------------------------------------

    \728\ AFPM, Detailed Comments, EPA-HQ-OAR-2018-0283-5698.
---------------------------------------------------------------------------

    AFPM also commented that it does not support the potential 
regulatory requirement for the production or use of higher octane 
gasoline as a compliance option. AFPM commented that EPA lacks the 
authority to require the use of higher octane fuels under CAA Sec.  
211(c)(1)(A). AFPM further commented ``[t]he only vehicles legally 
permitted to use more than 15 percent ethanol blends are flex-fuel 
vehicles, which are currently certified to utilize both E10 and E85. 
Without an alternative certification for an auto

[[Page 24389]]

manufacturer to build an E30 certified vehicle, which would require 
extensive testing and certification procedures as well as sufficient 
market availability of the certification fuel, it would be 
inappropriate for the Administration to consider such vehicles as a 
viable option in the 2022-2026 compliance period.''
    Gasoline retailers also commented regarding higher-octane fuels. 
NACS and SIGMA commented that they support examining the use of such 
fuels as a potential path towards future emissions reductions and that 
it will be important that the agencies appropriately consider and 
address a variety of related issues, including:
    1. How to allow and handle the expanded sales of higher-octane 
fuels, which may include fuels that currently face barriers to sale, 
such as E15;
    2. Streamlining the registration and regulation of higher-level 
blends of ethanol;
    3. Addressing misfueling liability concerns of retailers;
    4. Streamlining federal labeling requirements and ensuring federal 
preemption of state requirements; and
    5. Addressing any other regulatory and legislative challenges 
associated with the use of higher-octane fuels.\729\
---------------------------------------------------------------------------

    \729\ Joint submission on behalf of NACS and SIGMA, Detailed 
Comments, EPA-HQ-OAR-2018-0283-5824.
---------------------------------------------------------------------------

    NATSO commented that ``the Agencies should under no circumstances 
consider `requiring that mid-level [ethanol] blends be made available 
at service stations' '' and went on to say that ``retailers would need 
to be assured that they will not be held responsible for customers that 
misfuel . . . Federal dispenser labeling requirements would have to be 
streamlined and state requirements would have to be preempted. . . Auto 
manufacturers would have to warrant all new higher-octane vehicles up 
to at least E15 depending upon vehicles' capabilities, and would have 
to affirmatively state which cars in the existing fleet can run on E15 
and ensure that the cars are warrantied or retroactively warrantied as 
such.'' \730\
---------------------------------------------------------------------------

    \730\ NATSO, Detailed Comment, EPA-HQ-OAR-2018-0283-5484.
---------------------------------------------------------------------------

    UCS commented that ``[a]n orderly transition to high-octane fuel 
would take several years to complete. It will take time for the 
necessary regulations to be finalized, for vehicles optimized for high-
octane gasoline to come to market and to build out the fuel 
distribution infrastructure to make this fuel broadly available. And 
even once high-octane gasoline is in use, it will take more time for 
automakers to phase-in new models optimized for high-octane fuel and to 
fully replace the legacy E10 fleet. Another factor to consider is that 
the rising share of high-octane gasoline will be buffered by falling 
sales of gasoline, given increasing fuel efficiency, such that the 
overall demand for ethanol will change more slowly. The agencies' 
expectation is that high-octane gasoline will not significantly enter 
commerce before 2026, and subsequently will only gradually gain market 
share through 2040. There is no realistic prospect of completing this 
process before 2025 or 2026, the timeframe of this rulemaking. The 
appropriate context for this discussion within vehicle rules is the 
next round of fuel economy and emission standards. Even then, an 
expeditious rulemaking process will be required to achieve adequate 
regulatory clarity to facilitate rapid adoption post-2026.'' UCS also 
commented ``[we] strongly oppose granting fuel economy credits based on 
the technical potential of vehicles to operate on high-octane fuel 
before there is clear evidence that high-octane fuel is in use and the 
potential fuel economy benefits are being realized on the road.'' \731\
---------------------------------------------------------------------------

    \731\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    The agencies have reviewed the submissions received in response to 
their solicitation of comments concerning fuel octane levels and 
recognize the potential that higher-octane fuels, coupled with advanced 
engine technologies, can provide for improvements to fuel economy and 
tailpipe CO2 emissions. The agencies agree with commenters 
that establishing a higher minimum octane for gasoline is a complex 
undertaking that would require consideration of a wide array of 
difficult issues. In light of the complexity of the constellation of 
issues, the fact that EPA did not propose new octane requirements, and 
that EPA's authority to set fuel requirements resides in CAA section 
211(c)(1), the agencies recognize that the present rulemaking is not 
the appropriate vehicle to set octane levels. If EPA pursues future 
rulemaking action on this topic, it would consider these comments in 
that context and in consideration of the appropriate statutory 
provisions. The agencies note that the current vehicle certification 
process provides a path to certify a vehicle requiring the use of high-
octane fuel, which allows the impact of such fuels to be captured over 
the required certification test cycles for CO2 emissions and 
fuel economy.
    EPA also is declining to adopt new incentives for flex-fueled 
vehicles (FFVs) (vehicles designed to operate on gasoline or E85 or a 
mixture), as some commenters suggested. FFV incentives were not 
identified by EPA in its request for comments in the proposed rule and 
are outside the scope of this rulemaking.
    The analyses conducted for this rulemaking assumed the use of Tier 
3 fuels, where applicable, which are considered directly 
representative, or a reasonable proxy for, fuels available for 
consumers to purchase. As explained in the previous paragraph, agency 
actions related to test fuels, consumer available fuels, or flexible-
fuel incentives are out of scope of this rulemaking. However, to the 
extent that the agencies consider any additional rulemaking actions 
related to fuel octane requirements and/or availability, the agencies 
note that further analysis to set CAFE and CO2 standards 
would also reflect any potential, related impacts of those potential 
changes.
b) Engine Maps
    Engine paths include numerous engine technologies that 
manufacturers can use to improve fuel economy and reduce CO2 
emissions. Some engine technologies can be incorporated into existing 
engine design architectures with minor or moderate changes to the 
engine, but many engine technologies require an entirely new engine 
architecture or a major refresh. For this final rule analysis, twenty-
three unique engine technologies are available for adoption, and are 
evaluated uniquely across the ten separate vehicle types (technology 
classes).
    For the NPRM and final rule analysis, the impact of engine 
technologies on fuel consumption, torque, and other metrics was 
characterized using GT-POWER(copyright) modeling conducted by IAV 
Automotive Engineering, Inc. (IAV). IAV is one of the world's leading 
automotive industry engineering service partners and has extensive 
experience in testing and modeling engines and combustion. GT-POWER is 
a commercially available engine modeling tool with detailed cylinder 
and combustion modeling capabilities.\732\ GT-POWER is used to simulate 
engine behavior and provides data on engine metrics, including power, 
torque, airflow, volumetric efficiency, fuel consumption, turbocharger 
performance, and other parameters. The primary outputs of IAV's use of 
GT-POWER for this

[[Page 24390]]

analysis are the development of engine maps that provide operating 
characteristics of engines equipped with specific technologies.
---------------------------------------------------------------------------

    \732\ More information regarding GT Power Modeling is available 
at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
---------------------------------------------------------------------------

    When an engine is running, at any given point in time, the 
operation can be characterized by the engine's crankshaft rotational 
speed (typically in revolutions per minute, or RPM) and engine output 
(torque) level. Engines can operate at a range of engine speed and 
torque levels. Engine maps provide a visual representation of various 
engine performance characteristics at each engine speed and torque 
combination across the operating range of the engine. A common example 
of a performance characteristic is BSFC.\733\ Other characteristics 
include engine emissions, engine efficiency, and engine power.
---------------------------------------------------------------------------

    \733\ The amount of fuel needed to achieve a specific power, or 
how efficiently an engine uses fuel to produce work.
---------------------------------------------------------------------------

    Engine maps have the appearance of topographical maps, typically 
with engine speed on the horizontal axis and engine torque on the 
vertical axis. A third engine characteristic, BSFC, is displayed as 
contours, defining the operating regions for that BSFC with each 
contour showing all operating points at a specified BSFC value. Once 
created, the data they contain is referenced for engine fuel 
consumption at a given engine speed and torque operating point.
    For the NPRM and final rule analysis, the agencies relied on IAV to 
develop engine maps representing each of the engine technologies. IAV 
used benchmark production engine test data, component test data, and 
manufacturers and suppliers' technical publications to develop a one-
dimensional GT-POWER engine model for the baseline engine technology 
configuration. Technologies were incrementally added to the baseline 
model to assess their impact on fuel consumption. The following is a 
representative example of how IAV created the engine maps used in this 
analysis.
    First, IAV defined the characteristics of Eng01 (a base VVT engine) 
and optimized it for all the combustion parameters while minimizing 
fuel consumption and maintaining performance. The result of this was a 
fuel map as a function of BMEP and engine RPM. IAV then took the same 
Eng01 and adopted characteristics of SGDI technology to the base 
engine. The new engine (Eng18, VVT and SGDI) was then optimized for all 
combustion parameters while minimizing fuel consumption and maintaining 
performance. The result was an engine fuel map for Eng18, as a function 
of BMEP and engine speed. The engine map is directly comparable to the 
engine map for Eng01 and the difference in those engine maps 
specifically identifies the effectiveness impact of VVT and SGDI 
technologies. This process was repeated for all of the IAV engine maps 
that used Eng01 (VVT) as the baseline engine. This methodology ensured 
the engine maps represent the maximum improvement in BSFC for each 
engine configuration change, while considering real world design 
constraints.
    IAV used its global engine database that includes benchmarking 
data, engine test data, single cylinder test data, prior modeling 
studies, and technical publications and information presented at 
conferences to populate the assumptions and inputs used for engine map 
modeling, and to validate the ultimate results.\734\ Argonne used the 
engine maps resulting from this analysis as inputs for the Autonomie 
full vehicle modeling and simulation.
---------------------------------------------------------------------------

    \734\ Friedrich, I., Pucher, H., and Offer, T., ``Automatic 
Model Calibration for Engine-Process Simulation with Heat-Release 
Prediction,'' SAE Technical Paper 2006-01-0655, 2006, https://doi.org/10.4271/2006-01-0655. Rezaei, R., Eckert, P., Seebode, J., 
and Behnk, K., ``Zero-Dimensional Modeling of Combustion and Heat 
Release Rate in DI Diesel Engines,'' SAE Int. J. Engines 5(3):874-
885, 2012, https://doi.org/10.4271/2012-01-1065. Multistage 
Supercharging for Downsizing with Reduced Compression Ratio (2015). 
MTZ Rene Berndt, Rene Pohlke, Christopher Severin and Matthias 
Diezemann IAV GmbH. Symbiosis of Energy Recovery and Downsizing 
(2014). September 2014 MTZ Publication Heiko Neukirchner, Torsten 
Semper, Daniel Luederitz and Oliver Dingel IAV GmbH.
---------------------------------------------------------------------------

    As described in the NPRM and PRIA, the agencies developed engine 
maps for technologies that are in production today or that are expected 
to be available in the rulemaking timeframe. The agencies recognize 
that engines with the same combination of technologies produced by 
different manufacturers will have differences in BSFC and other 
performance measures, due to differences in the design of engine 
hardware (e.g., intake runners and head ports, valves, combustion 
chambers, piston profile, compression ratios, exhaust runners and 
ports, turbochargers, etc.), control software, and emission 
calibration. Therefore, the engine maps are intended to represent the 
levels of performance that can be achieved on average across the 
industry in the rulemaking timeframe.
    Accordingly, the agencies noted that it was expected that the 
engine maps developed for this analysis will differ from engine maps 
for manufacturers' specific engines. For a given engine configuration, 
some production engines may be less efficient and some may be more 
efficient than the engine maps presented in the analysis. However, the 
agencies intended and expected that the incremental changes in 
performance modeled for this analysis, due to changes in technologies 
or technology combinations, will be similar to the incremental changes 
in performance observed in manufacturers' engines for the same changes 
in technologies or technology combinations. Most importantly, using a 
single engine model as a reference provides a common base for 
comparison of all incremental changes resulting from technology 
changes, and anchors incremental technology effectiveness values to a 
common reference. The effectiveness values from the internal simulation 
results were validated against detailed engine maps produced from 
engine benchmarking programs, as well as published information from 
industry and academia, ensuring reasonable representation of simulated 
engine technologies.\735\
---------------------------------------------------------------------------

    \735\ Bottcher, L., Grigoriadis, P. ``ANL--BSFC map prediction 
Engines 22-26.'' IAV (April 30, 2019). 20190430_ANL_Eng 22-26 
Updated_Docket.pdf.
---------------------------------------------------------------------------

    As discussed in the NPRM, the agencies updated the list of engine 
technologies, before and after the Draft TAR, based on stakeholder 
comments and consultations with CARB, Argonne, and IAV. The technology 
list was built on the technologies that were considered in the 2012 
final rule, and included technologies that are being implemented or 
that are under development and feasible for production in the 
rulemaking timeframe. The agencies noted that some advanced engines 
were included in the simulation that were, and often still are, not yet 
in production, and the engine maps for those engines were either based 
on CBI or theoretical data. The agencies also stated in the NPRM that 
the final rule analysis may include updated engine maps for existing 
modeled engines, or entirely new maps added to the analysis if either 
action could improve the quality of the fleet-wide analysis.
    While there are a large number of possible combinations of engine 
technologies, the agencies categorized the IAV engine maps used in the 
NPRM full vehicle simulations into six categories. The categories were 
based on engine architecture and include: Dual overhead camshaft (DOHC) 
engines, single overhead camshaft (SOHC) engines, turbocharged engines, 
hybrid Atkinson cycle engines,\736\ non-hybrid

[[Page 24391]]

Atkinson mode engines, and diesel engines. Another unique technology 
that was available for adoption for the NPRM analysis was the advanced 
cylinder deactivation (ADEAC) for the SOHC and DOHC engines, however 
this technology was modeled using a fixed effectiveness value rather 
than an engine map, because the agencies did not have sufficient data 
to be used as input to the engine map or full vehicle simulation 
modeling. In addition, the agencies provided potential engine maps and 
additional specifications for several other technologies that could be 
considered for the final rule analysis. These included a basic high 
compression ratio Atkinson mode engine, a Miller cycle engine, and an 
engine with an electric assist.
---------------------------------------------------------------------------

    \736\ These types of Atkinson cycle engines are mainly for 
hybrid applications like Toyota Prius or Ford C-Max.
---------------------------------------------------------------------------

    The full list of engine maps used in the NPRM is presented in Table 
VI-39 below.
BILLING CODE 4910-59-P

[[Page 24392]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.143

BILLING CODE 4910-59-C
    The full list of engine maps used in this final rule analysis is 
presented in Table VI-40.

[[Page 24393]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.144


[[Page 24394]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.145


[[Page 24395]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.146

BILLING CODE 4910-59-C
    Comments on engine maps varied, with industry commenters generally 
supporting the maps used in the NPRM analysis and CARB and 
environmental advocate commenters generally objecting to the maps. The 
Alliance argued that previously-modeled fuel efficiency improvements 
for downsized, turbocharged engine technologies were ``highly 
optimistic,'' and stated that the updated engine maps used for the NPRM 
analysis were an improvement.
    ICCT argued that the IAV engine maps used for the NPRM analysis 
were out of date, and better engine maps benchmarked by EPA staff were 
available and should have been used instead.\737\ UCS similarly stated 
that Argonne work used for previous CAFE technical documents had relied 
on outdated engine maps, and that the new IAV engine maps used in this 
rulemaking were developed for a different purpose and had not been 
benchmarked against the latest engines either on the road or in 
development.\738\ ICCT questioned whether the agencies had validated 
engines 13 and 14 with physical testing and/or simulation modeling to 
the level of quality of EPA's simulation modeling.\739\ ICCT further 
asserted that EPA's benchmarked engine maps had been ``knowingly 
disregarded'' for the NPRM analysis, and stated that the NPRM analysis 
was therefore arbitrary.\740\ ICCT commented that the agencies must 
conduct and disclose a systematic investigation and comparison of 
engine benchmarking, engine modeling, and transmission modeling 
completed by EPA, Ricardo, and Argonne for model year 2014-2018 
vehicles. ICCT recommended that the agencies rely on engine maps used 
for past EPA ALPHA modeling while the agencies conduct such an 
investigation.
---------------------------------------------------------------------------

    \737\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-49.
    \738\ Union of Concerned Scientists, Technical Appendix, Docket 
No. NHTSA-2018-0067-12039, at 4.
    \739\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-46.
    \740\ ICCT, Docket No. NHTSA-2018-0067-11741, at I-49.
---------------------------------------------------------------------------

    The agencies believe it is most important for engine map data to 
provide accurate BSFC information for known technologies and technology 
levels. The agencies disagree with statements that IAV engine maps are 
outdated. The majority of the engine maps were developed specifically 
to support the midterm review and encompass engine technologies that 
are present in the analysis fleet and technologies that could be 
applied in the rulemaking timeframe. In many cases those engine 
technologies are mainstream today and will continue to be during the 
rulemaking timeframe. For example, the engines on some MY 2017 vehicles 
in the analysis fleet have technologies that were initially introduced 
ten, or more, years ago. Having engine maps representative of those 
technologies is important for the analysis. The most basic engine 
technology levels also provide a useful baseline for the incremental 
improvements for other engine technologies. The timeframe for the 
testing or modeling is unimportant, because time by itself doesn't 
impact engine map data. A given engine or model will produce the same 
BSFC map regardless of when testing or modeling is conducted. 
Simplistic discounting of engine maps based on temporal considerations 
alone could result in discarding useful technical information. Also, 
narrow use of temporal considerations would also result in the 
discarding of several engine maps from Ricardo that were used for the 
EPA Draft TAR and Proposed Determination analyses.\741\ Therefore, with 
the engine maps used representing current technologies regardless of 
development date, the agencies do not agree with commenter assertions.
---------------------------------------------------------------------------

    \741\ Ricardo, Inc. ``Computer Simulation of Light-Duty Vehicle 
Technologies for Greenhouse Gas Emission Reduction in the 2020-2025 
Timeframe.'' Ricardo (December 2011). https://nepis.epa.gov/Exe/ZyPDF.cgi/P100D57R.PDF?Dockey=P100D57R.PDF. Last accessed Jan 14, 
2020.
---------------------------------------------------------------------------

    The same commenters also appear to misunderstand how the agencies' 
effectiveness data, including engine maps, were used in the NPRM 
analysis (and in past rulemakings). The analysis never applies absolute 
BSFC levels from the engine maps to any vehicle model or configuration 
for the rulemaking analysis. The absolute fuel economy values from the 
full vehicle Autonomie simulations are used only to determine 
incremental effectiveness for switching from one technology to another 
technology. The incremental effectiveness is applied to the absolute 
fuel economy of vehicles in the analysis fleet, which are based on CAFE 
compliance data. For subsequent technology changes, incremental 
effectiveness is applied to the absolute fuel economy level of the 
previous technology configuration. Therefore, for a technically sound 
analysis, it is most important that the differences in BSFC among the 
engine maps be accurate, and not the absolute values of the individual 
engine maps. However, achieving this can be challenging.
    A technically sound approach is to use a single or very small 
number of baseline engine configurations with well-defined BSFC maps, 
and then, in a very systematic and controlled process, add specific 
well-defined technologies and create a BSFC map for each unique

[[Page 24396]]

technology combination. This could theoretically be done through engine 
or vehicle testing, but testing would need to be conducted on a single 
engine, and each configuration would require physical parts and 
associated engine calibrations to assess the impact of each technology 
configuration, which is impractical for the rulemaking analysis because 
of the extensive design, prototype part fabrication, development, and 
laboratory resources that are required to evaluate each unique 
configuration. Modeling is an approach used by industry to assess an 
array of technologies with more limited testing. Modeling offers the 
opportunity to isolate the effects of individual technologies by using 
a single or small number of baseline engine configurations and 
incrementally adding technologies to those baseline configurations. 
This provides a consistent reference point for the BSFC maps for each 
technology and for combinations of technologies which enables the 
differences in effectiveness among technologies to be carefully 
identified and quantified. The agencies selected this approach for the 
NPRM and final rule. Engine maps were created by IAV using this 
technically sound and rigorous methodology. Both absolute engine maps 
and the incremental differences in engine maps were presented in the 
PRIA.
    Using a mix of engine maps from engine modeling and from 
benchmarking data provides no common reference for measuring impacts of 
adding specific technological improvements. In addition, as discussed 
in further detail in Section VI.C.1.e), manufacturers often implement 
multiple fuel-saving technologies simultaneously when redesigning a 
vehicle and it is not possible to isolate the effect of individual 
technologies by using laboratory measurements of a single production 
engine or vehicle with a combination of technologies. Because so many 
vehicle and engine changes are involved, it is not possible to 
attribute effectiveness improvements accurately for benchmarked engines 
to specific technology changes. This leads to overcounting or 
undercounting technology effectiveness.
    Further, while two or more different manufacturers may produce 
engines with the same high level technologies (such as a DOHC engine 
with VVT and SGDI), each manufacturer's engine will have unique 
component designs that cause its version of the engine to have a unique 
engine map. For example, engines with the same high level technologies 
have unique intake manifold and exhaust manifold runners, cylinder head 
ports and combustion chamber geometry that impact charge motion, 
combustion and efficiency, as well as unique valve control, compression 
ratios, engine friction, cooling systems, and fuel injector spray 
characteristics, among other factors. The agencies developed and used a 
single engine map to represent each technology and each combination of 
engine technologies.
    Therefore, it should not be expected that any of the agencies' 
engine maps would necessarily align with a specific manufacturer's 
engine, unless of course the engine map was developed from that 
specific engine. The agencies do not agree that comparing an engine map 
used for the rulemaking analysis to a single specific benchmarked 
engine has technical relevance, beyond serving as a general 
corroboration for the engine map. When a vehicle is benchmarked, the 
resulting data is dictated by the unique combination of technologies 
and design constraints for the whole vehicle system. For these reasons, 
the agencies do not agree with ICCT that Eng13 and Eng14 should be 
validated by conducting full vehicle modeling and comparing the results 
with a single benchmarked vehicle. The engine maps used in this 
analysis are precisely controlled for specific incremental technology 
adoption and not for comparisons of absolute performance of a specific 
vehicle's engine.
    Differences are also explained by the NPRM and final rule analyses 
using large-scale full vehicle Autonomie simulations to estimate 
effectiveness instead of rough LPM approximations based on limited 
ALPHA simulation work.\742\ These issues are discussed in more detail 
in Section VI.B.3.
---------------------------------------------------------------------------

    \742\ 2016 EPA Proposed Determination TSD at p.2-276 to 2-279
---------------------------------------------------------------------------

    Accordingly, the agencies declined directly to use the Ricardo and 
other EPA engine maps created from engine benchmarking as inputs for 
this rulemaking because, among other reasons discussed below, they did 
not afford the opportunity to evaluate the effectiveness improvements 
for specific, individual technologies. For example, the 2018 Toyota 
Camry 2.5L engine that EPA benchmarked had a broad array of observable 
technologies, and several more that were not observable.\743\ However, 
there was no baseline from which to isolate or compare any of the 
individual technology improvements. For example, Toyota commented on 
this benchmarking, stating:
---------------------------------------------------------------------------

    \743\ EPA Test Data. 2018 Toyota Camry 2.5L A25A-FKS Engine Tier 
3 Fuel. Available at https://www.epa.gov/sites/production/files/2019-04/2018-toyota-2.5l-a25a-fks-engine-tier3-fuel-test-data-package-dated-04-08-19.zip. Last accessed Nov. 20, 2019.

    Past Toyota comments on Atkinson-cycle benefits have addressed 
only those derived from variable valve timing (VVT) with late intake 
valve closing (LIVC) that enables a 13:1 compression ratio. The 
total 18.6 percent improvement of the 2018 Camry 2.5L over the 
previous generation also includes benefits from cEGR and internal 
engine design changes such as to the block, cylinder head, pistons, 
valvetrain, as well as drivetrain and body/chassis 
enhancements.\744\
---------------------------------------------------------------------------

    \744\ NHTSA-2018-0067-12431. Supplemental Comments--Toyota Motor 
North America, at p. 1-2.

    Toyota's comments emphasize that the efficiency improvements in 
this engine were driven by several additional technological 
improvements, and not merely the cEGR, Atkinson cycle engine and higher 
compression ratio design that was assumed for the EPA Draft TAR and 
Proposed Determination analyses.\745\
---------------------------------------------------------------------------

    \745\ EPA PD TSD at 2-229.
---------------------------------------------------------------------------

    The agencies do agree component, engine, and vehicle test data are 
very important for validating systems models, such as Autonomie, and 
for validating model inputs, such as engine maps. Accordingly, the 
agencies did fully consider engine maps used in prior rulemakings, 
along with a broad array of other data as part of the process for 
evaluating the IAV engine maps used for the NPRM and the final rule 
analysis simulation work. Engine maps from Ricardo, EPA benchmarking, 
NHTSA-sponsored benchmarking,\746\ information from technical papers 
and conferences,\747\ extensive data and

[[Page 24397]]

expertise from the Argonne AMTL vehicle testing group and Energy 
modeling group, \748\ and the 2015 NAS report,\749\ were all sources 
used to confirm that incremental technology effectiveness estimates 
were appropriate. The engine maps developed by IAV provided reliable 
and reasonable estimates for the incremental impacts of engine 
technologies. The use of this approach explains some of the 
effectiveness differences between the NPRM and final rule analyses, and 
the EPA Draft TAR and Proposed Determination analyses.
---------------------------------------------------------------------------

    \746\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford F-
150 3.5 V6 EcoBoost with a 10 speed transmission.'' DOT HS 812 520.
    \747\ Maruyama, F., Kojima, M., and Kanda, T., ``Development of 
New CVT for Compact Car,'' SAE Technical Paper 2015-01-1091, 2015, 
doi:10.4271/2015-01-1091. Shelby, M., Leone, T., Byrd, K., and Wong, 
F., ``Fuel Economy Potential of Variable Compression Ratio for Light 
Duty Vehicles,'' SAE Int. J. Engines 10(3):2017, doi:10.4271/2017-
01-0639. Eisazadeh-Far, K. and Younkins, M., ``Fuel Economy Gains 
through Dynamic-Skip-Fire in Spark Ignition Engines,'' SAE Technical 
Paper 2016-01-0672, 2016, doi:10.4271/2016-01-0672. Wade, R., 
Murphy, S., Cross, P., and Hansen, C., ``A Variable Displacement 
Supercharger Performance Evaluation,'' SAE Technical Paper 2017-01-
0640, 2017, doi:10.4271/2017-01-0640. Hakariya, M., Toda, T., and 
Sakai, M., ``The New Toyota Inline 4-Cylinder 2.5L Gasoline 
Engine,'' SAE Technical Paper 2017-01-1021, 2017, doi:10.4271/2017-
01-1021. Ogino, K., Yakabe, Y., and Chujo, K., ``Development of the 
New V6 3.5L Gasoline Direct Injection Engine,'' SAE Technical Paper 
2017-01-1022, 2017, doi:10.4271/2017-01-1022. Shibata, M., Kawamata, 
M., Komatsu, H., Maeyama, K. et al., ``New 1.0L I3 Turbocharged 
Gasoline Direct Injection Engine,'' SAE Technical Paper 2017-01-
1029, 2017, doi:10.4271/2017-01-1029. Conway, G., Robertson, D., 
Chadwell, C., McDonald, J. et al., ``Evaluation of Emerging 
Technologies on a 1.6 L Turbocharged GDI Engine,'' SAE Technical 
Paper 2018-01-1423, 2018, doi:10.4271/2018-01-1423.
    \748\ ANL Energy Group. https://www.anl.gov/es; ANL AMTL group. 
https://www.anl.gov/es/advanced-mobility-technology-laboratory.
    \749\ National Research Council. 2015. Cost, Effectiveness, and 
Deployment of Fuel Economy Technologies for Light-Duty Vehicles. 
Washington, DC--The National Academies Press, at pp. 294-305. 
https://doi.org/10.17226/21744.
---------------------------------------------------------------------------

    In considering ICCT's comment about using IAV engine maps or EPA's 
engine maps, as an exercise, the agencies compared two IAV engine maps 
to the EPA's benchmarked Toyota 2.5L naturally aspirated engine and 
Honda's 1.5L turbocharged downsized engine.750 751 The IAV 
engines were modeled and simulated in a midsize non-performance vehicle 
with an automatic transmission and the same road load technologies, 
MR0, ROLL0 and AERO0, to isolate for the benefits associated with the 
specific engine maps.\752\ Eng 12, a 1.6L, 4 cylinder, turbocharged, 
SGDI, DOHC, dual cam VVT, VVL engine was selected as the closest engine 
configuration to the Honda 1.5L. Eng 22b, a 2.5L, 4 cylinder, VVT 
Atkinson cycle engine, was selected as the closest engine configuration 
to the Toyota 2.5L. As discussed before, both the Toyota 2.5L naturally 
aspirated engine and Honda's 1.5L engine have incorporated a number of 
fuel saving technologies including improved accessories and engine 
friction reduction. In order to assure an ``apples-to-apples'' 
comparison, both IACC and EFR technologies were applied to the IAV 
engine maps. IACC technology provides an additional 3.6% incremental 
improvement and EFR provides an additional 1.4% incremental improvement 
beyond the IAV engine maps for midsize non-performance vehicles.\753\
---------------------------------------------------------------------------

    \750\ Toyota 2.5L TNGA Prototype Engine From 2016 SAE Paper--
ALPHA Map Package. Version 2017-12. Ann Arbor, MI: US EPA National 
Vehicle and Fuel Emissions Laboratory, National Center for Advanced 
Technology, 2017.
    \751\ Honda 1.5L Turbo Prototype Engine From 2016 SAE Paper--
ALPHA Map Package. Version 2017-12. Ann Arbor, MI: US EPA National 
Vehicle and Fuel Emissions Laboratory, National Center for Advanced 
Technology, 2017.
    \752\ See ANL--All Assumptions_Summary_FRM_06172019_FINAL and 
ANL--Summary of Main Component Performance 
Assumptions_FRM_06172019_FINAL for midsize class characteristics.
    \753\ The NPRM and this final rule analysis allowed the adoption 
of IACC technologies in the CAFE model that provided an additional 
3.6% incremental improvement for the midsize car vehicle class. As 
discussed in [Section VI.C Other Technologies], these benefits are 
not shown in the IAV engine simulated results, so they were added 
manually for this comparison.
---------------------------------------------------------------------------

    The comparison shows effectiveness of the IAV engine maps and 
effectiveness values for the final rule analysis are in line with the 
Honda 1.5L and the Toyota 2.5L benchmarked engines. Figure VI-15 below 
shows the effectiveness improvements for the EPA benchmarked engines 
and the corresponding IAV engine maps incremental to a baseline 
vehicle. Accordingly, the agencies believe that the methodology used in 
this analysis, and the engine maps and incremental effectiveness values 
used, are in line with benchmarking data and are reasonable for the 
rulemaking analysis. The agencies believe the approach used in this 
rulemaking analysis appropriately allows the agencies to account for a 
wide array of engine technologies that could be adopted during the 
rulemaking timeframe. Declining to use manufacturer-specific engines 
allows the agencies to ensure that all effectiveness and cost 
improvements due to the incremental addition of fuel economy improving 
technologies are appropriately accounted for.

[[Page 24398]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.147

    Next, Roush Industries (``Roush''), writing on behalf of the 
California Air Resources Board, commented that the NPRM-modeled engines 
vary in cylinder size, which would significantly alter combustion, heat 
transfer, knock tolerance, and other important operating 
parameters.\754\ Roush stated that a more accurate simulation, which 
would improve incremental fuel economy improvement, should maintain a 
consistent cylinder displacement (500cc) and vary the number of 
cylinders or expected fuel consumption maps.\755\
---------------------------------------------------------------------------

    \754\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 12.
    \755\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 12.
---------------------------------------------------------------------------

    The agencies believe that holding cylinder volume constant is the 
appropriate approach to research seeking to identify the impacts of 
technological changes on BSFC, torque, power, and other 
characteristics, when holding cylinder volume constant. However, as 
explained in Section VI.B.3.a)(2) Maintaining Vehicle Attributes and 
Section VI.B.3.a)(6) Performance Neutrality, CAFE and CO2 
rulemaking analyses attempt to maintain vehicle attributes, including 
performance, and hold all of the attributes constant when showing 
pathways that improve fuel economy. Therefore, the agencies' analyses 
require engine maps that attempt to hold performance constant--not 
necessarily cylinder size. Since certain fuel economy improving 
technologies would increase performance if cylinder size is held 
constant, such as when adding turbocharging technology, the agencies 
appropriately include changes in displacement and cylinder volume for 
technologies that have a significant impact on engine torque and power, 
such as turbocharging. For a number of fuel economy improving 
technologies that had smaller impacts on engine torque and power, the 
engine maps were created with cylinder volume held constant. Table VI-
39 identifies the engine displacement information for each of the 
engine maps. For example, the same engine displacement (2.0 L) and 
cylinder displacement (500 cc) was used for creating engine maps for 
naturally aspirated engines Eng01, Eng02, Eng03, Eng04, Eng05a, Eng5b, 
Eng06a, Eng07a, and Eng08a, whereas engine displacement (1.6 L) and 
cylinder displacement (400 cc) is used for creating the engine map for 
turbocharged engine Eng12 in order to maintain performance. The 
agencies have concluded that the approach used for the NPRM and the 
final rule analysis is the most technically sound approach given the 
data needs and assessments required for CAFE and CO2 
rulemaking.
    Roush also commented as follows:

[S]everal of the base engine maps used in the 2018 PRIA analysis 
exhibit maximum thermal efficiency (lowest fuel consumption) at 
2000-3000 rpm and at maximum load, which is unrealistic for normal 
passenger vehicle engines. Such maps will over predict fuel economy 
for extremely down-sized applications (very small engine in a heavy 
vehicle). This is because there is no fuel economy penalty for 
running the engine at a high loads point where, in reality, BSFC is 
high due to retarding spark timing to prevent knocking and fuel 
enrichment to reduce exhaust temperatures to protect exhaust valves 
and turbocharger components.\756\
---------------------------------------------------------------------------

    \756\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 11.

    For example, Roush stated that Eng12 is predicted to have its 
highest efficiency at very high load and high engine speeds with no 
degradation in brake specific fuel consumption (BSFC) at engine speeds 
between 2,000 rpm and 4,500 rpm all the way up to peak load, which is 
unrealistic because

[[Page 24399]]

turbocharged engines at high loads require retarded spark timing to 
prevent knock and fuel enrichment to prevent overheating of the 
turbocharger and related components.\757\ Roush stated that these 
factors would increase fuel consumption and reduce efficiency under 
real-world conditions.\758\ Roush also stated that another effect of 
the Eng12 fuel consumption curve would be to predict unreasonably good 
fuel consumption at very high power levels for downsized turbocharged 
engines. Roush stated this could bias technology pathways in over-
predicting fuel economy benefits for small engines installed in heavier 
vehicles, causing an overly optimistic predicted performance of the 
vehicle with regard to drivability, acceleration, and fuel consumption, 
which would create unrealistic real-world pathways to compliance.\759\
---------------------------------------------------------------------------

    \757\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 18.
    \758\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 19.
    \759\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 23.
---------------------------------------------------------------------------

    As discussed in the Argonne model documentation for the final rule 
analysis, the simulations used to determine incremental effectiveness 
for the NPRM and final rule analyses were conducted using 2-cycle test 
procedures, because they are the test procedures used for CAFE and 
CO2 compliance.\760\ Therefore, the engines maps are 
intended to represent BSFC accurately under those test conditions and 
do not need to capture BSFC under every operating condition. During 2-
cycle test conditions, engines do not operate for extended periods at 
the speed and high load conditions noted by Roush. A few vehicle and 
engine combinations may operate at those speed and load points only 
briefly during the 2-cycle CAFE and CO2 tests. Engines are 
capable of operating for short periods of time under higher exhaust 
temperature conditions and manufacturers commonly delay fuel enrichment 
until it is needed to protect engine components (in particular exhaust 
valves and exhaust manifolds) from excessive temperatures that can 
impact engine durability. Fuel enrichment can be delayed because it 
takes a period of time at higher temperature for components to heat up 
and reach a temperature that would impact durability. Because these 
high speed and load conditions occur for a relatively short time during 
the CAFE and CO2 test cycles, and then return to lower speed 
and/or load conditions with lower exhaust temperature, engines operate 
for the entire CAFE and CO2 test cycles without triggering 
fuel enrichment. The fuel enrichment delay also enables vehicles to 
comply with criteria emission regulations and improves real world fuel 
economy. Therefore, the engine maps used for the NPRM and final rule 
analysis fully represent how engines operate during CAFE and 
CO2 test cycles, and properly do not include fuel enrichment 
at all 2-cycle operating conditions. Also, a trained knock model was 
used to develop the engine maps, and the spark timing reflects 
appropriate levels for engine operation during the delay in fuel 
enrichment.
---------------------------------------------------------------------------

    \760\ A Detailed Vehicle Simulation Process To Support CAFE and 
CO2 Standards for the MY 2021-2026 Final Rule Analysis.
---------------------------------------------------------------------------

    Next, regarding developing the NPRM engine maps to account for Tier 
3 test fuel, the Alliance and Ford stated that the engine maps using 
Tier 3 test fuel represented an improvement over prior analyses. The 
Alliance stated that previous EPA modeling had incorrectly used Tier 2 
premium octane fuel to predict the benefits of engine technologies, 
which overstated fuel economy gains that would be achievable when using 
regular-grade octane Tier 3 fuel. Ford provided similar comments, and 
also noted that regular grade octane fuel will be required for 
compliance after the 2020 model year.\761\
---------------------------------------------------------------------------

    \761\ Ford Motors, Attachment, Docket No. EPA-HQ-OAR-2018-0283-
5691, at 7.
---------------------------------------------------------------------------

    In contrast, ICCT and UCS both commented that the agencies had 
incorrectly updated the IAV engine maps developed with Tier 2 test fuel 
to account for Tier 3 fuel.\762\ ICCT stated that the update reduced 
the effectiveness of the turbo technologies and suggested that the fuel 
update adjustment should not have been done at all, stating 
manufacturers that label vehicles as ``premium fuel recommended'' are 
required to show no emissions changes over all test cycles when using 
premium octane fuel and therefore reducing effectiveness for fuel 
differences, as the agencies did with the IAV engine maps, is 
unrealistic and inappropriate.
---------------------------------------------------------------------------

    \762\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-82; Union of Concerned 
Scientists, Technical Appendix, Docket No. NHTSA-2018-0067-12039, at 
p. 15.
---------------------------------------------------------------------------

    UCS also commented more specifically on the impact of the 
adjustment from Tier 2 to Tier 3 fuel related to the knock threshold 
for advanced engines, noting that manufacturers consider different 
approaches to different fuels, and not all of those approaches 
necessitate reductions in efficiency, as the agencies' assumption 
suggests. UCS stated that charge cooling can reduce knock in direct 
injection engines, resulting in an ``effective octane'' difference of a 
six point increase for E10, thus potentially compensating for the 
difference in octane between Tier 2 (E0 93 AKI) and Tier 3 (E10 87 AKI) 
fuels. UCS argued that excluding this consideration led the agencies to 
restrict advanced engines like HCR2 and reduce the effectiveness of 
turbocharged engines with CEGR. UCS suggested that there would be a 
reduction in the costs between the baseline and proposed standards if 
the analysis allowed the application of HCR2 engines and corrected the 
effectiveness of turbocharged CEGR engines.
    Both ICCT and UCS also stated that the adjustment ignored a 2018 
EPA study showing that, while fuel consumption increases with the 
switch from Tier 2 to Tier 3 test fuel, emissions are reduced, meaning 
that the agencies' adjustment is wrong ``for some technologies because 
[CO2]-per-mile emissions can be lower with the switch to 
higher octane ethanol blends.'' UCS also stated that the adjustment 
factor applied is wrong for two reasons, first because converting 
solely with energy density would assume a 3.7 percent increase in fuel 
consumption compared to the observed 2.7 percent increase, and second 
because the adjustment goes in the wrong direction when applied to 
CO2 emissions, which show a reduction of 1.4 percent on the 
test cycle. UCS stated that the Autonomie model accordingly overstates 
CO2 emissions on Tier 3 fuel by 4.2 percent. UCS argued that 
the adjustment to account for Tier 3 test fuel therefore double counts 
any penalty in fuel economy and ignores CO2 tailpipe 
reductions, which would result in an improvement on the test cycle. 
Because the CAFE test procedure already has an adjustment in place to 
correct for fuel properties relative to 1975 test fuel, but carbon-
related exhaust emissions do not, UCS stated that the fuel adjustment 
could lead to drastically conservative fuel economy and CO2 
curves.
    ICCT stated that the agencies could fix this issue by relying on 
EPA's engine maps, where EPA had accounted for cost and effectiveness 
of technology used to protect operation on regular octane fuel by 
increasing costs and reducing effectiveness.
    Some of these comments can be addressed with a simple 
clarification: The NPRM contained text that was

[[Page 24400]]

inconsistent regarding how the analysis accounted for the engine maps 
(which were based on Tier 3 fuel). The separate model documentation 
correctly described that, for the NPRM analysis, the agencies developed 
fuel maps for Tier 3 fuel and did not adjust the final Autonomie 
outputs.\763\ The NPRM text, however, incorrectly stated that ``(a)n 
adjustment factor was applied to the Autonomie simulation results to 
adjust them to reflect Tier 2 certification fuel. Argonne adjusted the 
vehicle fuel economy results to present certification fuel by using the 
ratio of the lower heating values to the rest and certification 
fuels.'' In fact, no adjustments were made to the NPRM Autonomie 
simulation outputs, as the modeled engine maps were appropriately 
modeled using Tier 3 fuel.
---------------------------------------------------------------------------

    \763\ NHTSA-2018-0067-0007 at 177-178 and 191.
---------------------------------------------------------------------------

    As discussed in detail in VI.C.1.a) Fuel Octane, engine 
specifications used to create the engine maps for the NPRM and the 
final rule were developed using Tier 3 fuel. Tier 3 fuel was used to 
ensure the engines were capable of operating on real world regular 
octane (87 pump octane = (R+M/2)). This capability is in line with what 
manufacturers must do to ensure engines have acceptable noise, 
vibration, harshness, drivability and performance levels, and will not 
fail prematurely when operated on regular octane fuel. If the agencies 
developed engine maps based on Tier 2 fuel alone, the engine maps would 
reflect the engines' ability to have higher compression ratios and to 
operate with greater levels of spark advance than could be implemented 
by manufacturers, who must take into account operation on regular 
octane fuels used by a majority of U.S. consumers.\764\ Not considering 
regular octane fuel operation by manufacturers would lead to engine 
durability, and engine noise, vibration, harshness, and drivability 
issues. Manufacturers have told the agencies that even for vehicles 
designed to operate on high octane fuel, the engines and controls must 
be designed to operate on every fuel octane level available in the U.S. 
to avoid these issues.\765\ Thus, developing engine maps based on Tier 
2 fuel alone would incorrectly overstate the BSFC improvements 
achievable in the real world.
---------------------------------------------------------------------------

    \764\ Tamm, D.C., Devenish, G.N. Finelt, D.N. Kalt, L.K. 
``Analysis of Gasoline Octane Costs'' Baiker and O'brien, Inc. 
Prepared for EIA. October 18, 2018. https://www.eia.gov/analysis/octanestudy/pdf/phase1.pdf at 11-13.
    \765\ Ford Motor Company. NHTSA-2016-0068-0048 at 3. Auto 
Alliance comments for 2016 draft TAR. Attachment 7 Limitations of 
Ricardo Fuel Economy Analysis of Downsizing. NHTSA-2016-0068-0070.
---------------------------------------------------------------------------

    Based on these comments and considerations, the agencies determined 
the engine maps developed for the NPRM appropriately account for fuel 
octane, and better approximate BSFC achieved by the majority of engines 
used in the U.S. vehicle fleet. The agencies believe ICCT's and other 
commenters' assertions that the engine maps should reflect Tier 2 fuel 
and not be updated for Tier 3 fuel would ignore these important 
considerations, and would provide engine maps that could not be 
achieved by engines in the real world. The agencies determined that 
engine maps developed for the Draft TAR and EPA Proposed Determination 
that were based on Tier 2 fuel should not be used for the NPRM and 
final rule analyses for these reasons.
    EPA is addressing the impact of Tier 3 fuel on fuel economy and 
CO2 emissions compliance test results as part of a separate 
rulemaking. The separate rulemaking may establish an adjustment to 
account for the impacts of the change in test fuel. Those impacts are 
beyond the scope of this rulemaking. The analysis for this rule uses 
fuel economy and CO2 emissions of the vehicles in the MY 
2017 analysis fleet as the reference for absolute fuel economy and 
CO2 emissions. The analysis starts with absolute compliance 
data from MY 2017 and adopts technologies incrementally to determine 
future compliance. Because MY 2017 absolute compliance values are based 
on Tier 2 fuel, and standards are based on the use of Tier 2 fuel, 
there is no need to make any adjustments for the differences in energy 
content and carbon content of Tier 2 and Tier 3 fuel.\766\
---------------------------------------------------------------------------

    \766\ During the 1980s, the U.S. Environmental Protection Agency 
(EPA) incorporated the R factor into fuel economy calculations in 
order to address concerns about the impacts of test fuel property 
variations on corporate average fuel economy (CAFE) compliance, 
which is determined using the Federal Test Procedure (FTP) and 
Highway Fuel Economy Test (HFET) cycles. The R factor is defined as 
the ratio of the percent change in fuel economy to the percent 
change in volumetric heating value for tests conducted using two 
differing fuels.
---------------------------------------------------------------------------

    The agencies considered ICCT's statement that manufacturers that 
label vehicles as ``premium fuel recommended'' are required to show no 
emissions changes over all test cycles when using regular octane fuel, 
and therefore reducing effectiveness for fuel differences as the 
agencies did with the IAV engine maps is unrealistic and inappropriate. 
The agencies believe these conclusions are technically incorrect. The 
existence of an EPA compliance regulation does not impact the laws of 
nature, which govern issues associated with the impact of fuel octane 
on the ability to improve engine BSFC and on engine durability, noise, 
vibration, harshness, and drivability. It is widely recognized and 
accepted that higher octane fuels allow engines to be designed with 
higher compression ratios, faster combustion rates, and more optimal 
spark advance, which improve BSFC. Section VI.C.1.a) discusses comments 
advocating for increasing the minimum fuel octane specification to 
enable these improvements. The engine maps developed by IAV and used 
for the Draft TAR and NPRM were consistent with these trends and showed 
that BSFC is better with Tier 2 (higher octane) fuel than Tier 3 (lower 
octane) fuel.\767\ ICCT did not provide any data supporting the concept 
that there is no shift in BSFC, fuel economy, or CO2 
emissions when engines are optimized with different octane fuels, or 
between Tier 2 and Tier 3 fuel. It is appropriate to note that the EPA 
regulation does provide a tolerance which in practice allows a small 
level of shift in emissions.\768\
---------------------------------------------------------------------------

    \767\ See BSFC difference between engines modeled with Tier 3 
fuel versus high octane fuel by IAV in PRIA 6.3.2.2.20.9 at 288 to 
PRIA 6.3.2.20.11 at 292.
    \768\ 40 CFR 1066.210 (b) Accuracy and Precision.
---------------------------------------------------------------------------

    Regarding comments that certain combinations of technologies can 
enable BSFC improvements while controlling spark knock, the agencies in 
fact considered a very broad array of engine technology combinations 
for the analysis, including several added technologies as discussed 
further below. The agencies believe the rigorous methodology used to 
develop the engine maps resulted in engine maps representing the 
maximum improvement in BSFC for each engine configuration, while also 
addressing real world constraints. Engine maps for the new technologies 
were presented in PRIA Chapter 6.3.2.2.16.4. The PRIA also discussed 
that IAV maps were developed considering a very comprehensive list of 
combustion operating parameters as part of the IAV GT-Power engine 
modeling. IAV's GT-Power engine modeling included sub-models to account 
for heat release through a predictive combustion model, knock 
characteristic through a kinetic fit knock model, physics-based heat 
flow model physics based friction model, and IAV's proprietary 
Optimization Tool Box.\769\ These independent models were

[[Page 24401]]

run concurrently to make sure engine design requirements were met for 
each engine configuration that was modeled.
---------------------------------------------------------------------------

    \769\ IAV's Optimization Tool Box is a module of IAV Engine. IAV 
Engine, as the basic platform for designing engine mechanics, 
provides a large number of tools that have proven their worth across 
the globe in several decades of automotive development work at IAV. 
The modules help designers, computation engineers and simulation 
specialists in designing mechanical engine components--for example, 
in laying out valvetrains and timing gears as well as crankshafts.
---------------------------------------------------------------------------

    Finally, in response to the agencies' request for comment on 
including the additional engine maps presented in the NPRM as potential 
technological pathways, several commenters stated that the agencies 
should include those technologies, in addition to other emerging engine 
technologies.\770\ After considering these comments, the agencies added 
several engine technologies and technology combinations to the final 
rule analysis. The additions included a basic high compression ratio 
Atkinson mode engine (HCR0), a variable compression ratio engine (VCR), 
a variable turbo geometry engine (VTG), and a variable turbo geometry 
with electric assist engine (VTGe). The agencies also added advanced 
cylinder deactivation technology (TURBOAD) to Eng12 (TURBOD) in the 
Autonomie modeling for the final rule analysis. Like with ADEAC, the 
agencies did not have IAV engine maps for TURBOAD, so the agencies took 
the effectiveness values as predicted by full vehicle simulations of a 
TURBOD and added 1.5 percent or 3 percent respectively for I-4 engines 
and V-6 or V-8 engines, as explained in more detail further below. The 
agencies also included more iterations of existing technologies, like 
diesel engines with cylinder deactivation, diesel engines paired with 
manual transmissions, and diesel engines paired with 12-volt start stop 
technology, in addition to more combinations of hybrid technologies 
that are discussed further in Section VI.C.3, below.
---------------------------------------------------------------------------

    \770\ ICCT Docket # NHTSA-2018-0067-11741 at I-19--I-22; CARB 
Docket # NHTSA-2018-0067-11873 at 107-108.
---------------------------------------------------------------------------

    The following sections list and describe the comprehensive set of 
engine technologies and combinations of engine technologies that have 
been included in the analysis. The agencies also discuss the additional 
engine technologies added for the final rule, and reasons for excluding 
a small number of technologies proffered by commenters. The agencies 
believe the wide array of engine technologies included in the final 
rule analysis and the methodology used to develop the engine maps to 
measure the effectiveness of those technologies reasonably represents 
the scope of technologies that should be considered during the 
rulemaking timeframe.
c) Engine Modeling in the CAFE Model
(1) Basic Engines
    The NPRM described that there are a number of engine technologies 
that manufacturers can use to improve fuel economy and CO2 
emissions. Some engine technologies can be incorporated into existing 
engines with minor or moderate changes to the engines, but many engine 
technologies require an entirely new engine architecture. The terms 
``basic engine technologies'' and ``advanced engine technologies'' are 
used only to define how the CAFE model applies a specific engine 
technology and handles incremental costs and effectiveness 
improvements. ``Basic engine technologies'' refer to technologies that, 
in many cases, can be adapted to an existing engine with minor or 
moderate changes to the engine, compared to ``advanced engine 
technologies'' that generally require significant changes or an 
entirely new engine architecture.
    In the CAFE model, basic engine technologies may be applied in 
combination with other basic engine technologies; advanced engine 
technologies (defined by an engine map) stand alone as an exclusive 
engine technology. The words ``basic'' and ``advanced'' are not meant 
to confer any information about the level of sophistication of the 
technology. Also, many advanced engine technology definitions include 
some basic engine technologies, but these basic technologies are 
already accounted for in the costs and effectiveness values of the 
advance engine. The ``basic engine technologies'' need not be (and are 
not) applied in addition to the ``advanced engine technologies'' in the 
CAFE model.
(a) DOHC
    In the NPRM analysis, the agencies characterized dual overhead cam 
(DOHC) engine technology as ``basic.'' DOHC engine configurations have 
two camshafts per cylinder head, one operating the intake valves and 
one operating the exhaust valves. Four basic engine technologies--
variable valve timing (VVT), variable valve lift (VVL), stoichiometric 
gasoline direction injection (SGDI), and basic cylinder deactivation 
(DEAC)--were considered for DOHC engines. Implementing these 
technologies involves changes to the cylinder head of the engine, but 
the engine block, crankshaft, pistons, and connecting rods require few, 
if any, changes.
    Variable valve timing (VVT) is a family of valve-train designs that 
dynamically adjusts the timing of the intake valves, exhaust valves, or 
both, in relation to piston position. VVT can reduce pumping losses, 
provide increased engine torque and horsepower over a broad engine 
operating range, and allow unique operating modes, such as Atkinson 
cycle operation, to further enhance efficiency. VVT is nearly 
universally used in the MY 2017 fleet.\771\ In the NPRM analysis, the 
VVT technology modeled by IAV was based on dual (independent) cam 
phasing. This was a more advanced VVT technology that allowed 
controlling of valve overlap, which can be used to control internal EGR 
to minimize fuel consumption at low engine loads.\772\ VVT enables 
control of many aspects of air flow, exhaust scavenging, and combustion 
relative to fixed valve timing engines. Engine parameters such as 
volumetric efficiency, effective compression ratio, and internal 
exhaust gas recirculation (iEGR) can all be enabled and accurately 
controlled by a VVT system.
---------------------------------------------------------------------------

    \771\ 98.1 percent of MY2017 vehicles are equipped with VVT. EPA 
Report. The 2018 EPA Automotive Trends Report. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at 
Table 4.1 Production Share by Engine technology.
    \772\ 2015 NAS at p. 32.
---------------------------------------------------------------------------

    Variable valve lift (VVL) dynamically adjusts the distance a valve 
travels from the valve seat optimizing airflow over a broad range of 
engine operating conditions. The technology can increase effectiveness 
by reducing pumping losses and may improve efficiency by affecting in-
cylinder charge (fuel and air mixture), motion, and combustion. VVL is 
less common in the 2017 fleet than VVT. Some manufacturers have 
implemented a limited, discrete approach to VVL where just two valve 
lift profiles are available versus a full-range, continuously variable 
implementation.
    Stoichiometric gasoline direct injection (SGDI) sprays fuel at high 
pressure directly into the combustion chamber, which provides cooling 
of the in-cylinder charge via in-cylinder fuel vaporization to improve 
spark knock tolerance and enable an increase in compression ratio and/
or more optimal spark timing for improved efficiency. SGDI appears in 
about half of basic engines produced in MY 2017, and the technology is 
used in many advanced engines as well.\773\
---------------------------------------------------------------------------

    \773\ 49.7 percent of MY2017 vehicles are equipped with SGDI. 
EPA Report. The 2018 EPA Automotive Trends Report. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at 
Table 4.1 Production Share by Engine technology.
---------------------------------------------------------------------------

    Basic cylinder deactivation (DEAC) disables intake and exhaust 
valves and

[[Page 24402]]

turns off fuel injection for the deactivated cylinders during light-
load operation. The engine runs temporarily as though it were a smaller 
engine, which reduces pumping losses and improves efficiency. In the MY 
2017 fleet, manufacturers used DEAC on V6, V8, V10, and V12 engines in 
OHV, SOHC, and DOHC engine configurations. With some engine 
configurations in some operating conditions, DEAC creates noise-
vibration-and-harshness (NVH) challenges. NVH challenges are 
significant for V6 and I4 DEAC configurations, and limit the operating 
range where DEAC can operate. For I4 engine configurations with smaller 
displacements, there are fewer operating conditions where engine load 
is low enough to use DEAC, which limits effectiveness. No manufacturers 
produced I4 DEAC engines in MY 2017. Typically, the smaller the engine 
displacement, the less opportunity DEAC provides to improve fuel 
consumption.
    The agencies provided engine fuel maps for each of the eight DOHC 
engines (Eng01, Eng02, Eng03, Eng04, Eng18, Eng19, Eng20, and Eng21) 
used for the NPRM analysis. Each of these engines incrementally added 
technology to Eng01, a basic VVT engine, while holding all other 
factors constant like ambient temperature, ambient pressure, and fuel 
type.
    For the NPRM analysis, the agencies estimated the effectiveness of 
DEAC using full vehicle modeling and simulation. In the NPRM PRIA 
6.2.1.2, the agencies discussed how Autonomie uses a specific control 
logic for cylinder deactivation for naturally aspirated engines that 
takes into consideration for noise, vibration, and harshness.\774\ For 
the final rule analysis, the agencies took steps to use full vehicle 
modeling and simulation to apply DEAC to both naturally aspirated and 
turbocharged engines. The same control logic was applied to the 
turbocharged engine cylinder deactivation (TURBOD) for the final rule 
analysis.
---------------------------------------------------------------------------

    \774\ NHTSA-2018-0067-1972. ``Preliminary Regulatory Impact 
Analysis (PRIA) The Safer Affordable Fuel-Efficient (SAFE) Vehicles 
Rule for Model Year 2021-2026 Passenger Cars and Light Trucks,'' at 
191.
---------------------------------------------------------------------------

    The agencies used the same assumptions for advanced cylinder 
deactivation (ADEAC) in the final rule analysis. In the NPRM the 
agencies stated engine maps were not available at the time of the 
analysis, and said that ADEAC was estimated to improve a basic engine 
with VVL, VVT, SGDO, and DEAC by three percent (for 4 cylinder engines) 
and six percent (for engines with more than 4 cylinders).\775\ The new 
technology combination for turbocharged advanced cylinder deactivation 
(TURBOAD) uses a similar approach for determining effectiveness. The 
agencies have applied a one-and-a-half percent effectiveness 
improvement estimate for 4-cylinder or smaller engines and a three 
percent effectiveness estimate for 6-cylinder or larger engines 
relative to TURBOD.
---------------------------------------------------------------------------

    \775\ 83 FR 430039 (Aug. 24, 2018).
---------------------------------------------------------------------------

    For the final rule analysis the basic engine path for DOHCs are 
shown in Figure VI-16 and the high-level engine specifications are 
shown in Table VI-41. The baseline basic DOHC engine, Eng01, was the 
starting point and other engine technologies were incrementally adopted 
to determine effectiveness. Adoption of DEAC technology for 
turbocharged engines is discussed in Section VI.C.1.e)(2). Similarly, 
ADEAC technology is discussed in Section VI.C.1.e)(4).
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.148


[[Page 24403]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.149

BILLING CODE 4910-59-C
(b) SOHC
    Similar to DOHC engines, SOHC engines were characterized as 
``Basic'' engine technologies in the NPRM analysis. They are 
characterized by having a single camshaft in the cylinder head 
operating both the intake and exhaust valves. Four basic engine 
technologies, VVT, VVL, SGDI, and DEAC were considered for SOHC 
engines. Implementing these technologies involves changes to the 
cylinder head of the engine, but the engine block, crankshaft, pistons, 
and connecting rods require few, if any, changes.
    The agencies provided engine fuel maps for each of these types of 
SOHC engines and requested comments. Engine maps 5b, 6a, 7a, and 8a 
were modeled SOHC engines. The SOHC engine models used engine 5a, which 
was based on Eng01 as a reference, by removing one camshaft. Eng5a was 
included for the Draft TAR, but not included for the NPRM analysis due 
to high BSFC from higher friction that was inherited from the DOHC 
engine design. A level 0.1 bar of friction reduction over the entire 
operating range for engine maps 5b, 6a, 7a, and 8a was applied to 
represent improvements over existing engine designs. The addition of 
friction reduction to these engines was a result of consideration of 
deliberative interagency comments received during the Draft TAR review 
process noting higher fuel consumption on the baseline SOHC engine 5a 
relative to other modern SOHC engines.
    Meszler on behalf of NRDC commented that ``[a]lthough variable 
valve timing (VVT) technology is identified as an available refresh 
technology, the NPRM CAFE model (unlike the version used for the 2016 
TAR analysis) actually assumes that all baseline vehicles include VVT 
technology. As a result, the approximately 9 percent of model year 2016 
sales that do not actually include VVT are not credited with any 
efficiency benefit for adoption of the technology . . . . '' \776\
---------------------------------------------------------------------------

    \776\ Meszler, at 32.
---------------------------------------------------------------------------

    We agree with this comment, and for the final rule analysis updated 
the CAFE model to add a non-VVT level engine in the 2017 analysis fleet 
and to allow those vehicles to adopt VVT technologies at a refresh or 
redesign. However, the agencies did not have engine maps for the non-
VVT engines, so the agencies applied a fixed-value effectiveness 
estimate from similar VVT engine maps to represent the effectiveness 
for non-VVT engines. The agencies used the effectiveness of a similar 
configuration technology package of another engine to represent non-VVT 
engines. Non-VVT SOHC engines may add any combination of VVL with SGDI 
and DEAC. The agencies believe that the estimated effectiveness used 
for VVT engines was appropriate because the effectiveness offset is in 
line with 2015 NAS estimates for VVT engines with respect to VVL 
engines.777 778
---------------------------------------------------------------------------

    \777\ Baseline effectiveness references for SOHC;VVT; SGDI; 
AT5;CONV;ROLL0;MR0;AERO0, SOHC;VVT; DEAC; AT5;CONV;ROLL0;MR0;AERO0, 
SOHC;VVT;VVL; DEAC; AT5;CONV;ROLL0;MR0;AERO0, and SOHC;VVT; 
SGDI;DEAC; AT5;CONV;ROLL0;MR0;AERO0 were used to represent SOHC;VVL; 
SGDI; AT5;CONV;ROLL0;MR0;AERO0, SOHC;VVL;DEAC; 
AT5;CONV;ROLL0;MR0;AERO0, and SOHC;VVL; SGDI;DEAC; 
AT5;CONV;ROLL0;MR0;AERO0 baseline combinations. These combinations 
represented only 2% of the models and 3.1% sales by volume in the MY 
2017 baseline fleet.
    \778\ 2015 NAS Table 2.7 and Table 2.8 at 32-33.
---------------------------------------------------------------------------

    The basic engine path for SOHC engines used in this final rule is 
shown in Figure VI-17 and the specifications are shown in Table VI-42. 
Note, that Eng5a is only a reference used to build the rest of the SOHC 
engines.
BILLING CODE 4910-59-P

[[Page 24404]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.150

[GRAPHIC] [TIFF OMITTED] TR30AP20.151

BILLING CODE 4910-59-C
(2) Turbocharged Downsized Engines
    Engine maps 12, 13, and 14 modeled turbocharged downsized engines. 
Turbocharged downsized engines are characterized by technology that can 
create greater-than-atmospheric pressure in the engine intake manifold 
when higher output is needed. The raised pressure results in an 
increased volume of airflow into the cylinder supporting combustion, 
increasing the specific power of the engine. An increased specific 
power means the engine can generate more power per unit of volume, 
which allows engine volume to be reduced while maintaining performance, 
thereby increasing fuel efficiency. IAV Eng12 was the base engine for 
all simulated turbocharged engines and was validated using engine 
dynamometer test data.\779\
---------------------------------------------------------------------------

    \779\ Bottcher, L. Grigoriads, P. ``ANL--BSFC map prediction 
Engines 22-26'' April, 30, 2019. IAV_20190430_Eng 22-26 
Updated_Docket.pdf.
---------------------------------------------------------------------------

    One notable change that the agencies made for the NPRM analysis 
based on stakeholder comments to the Draft TAR was to update the turbo 
family engine maps to assume operation on regular octane fuel (Tier 3, 
or 87 AKI), instead of premium fuel (Tier 2, or 93 AKI), to assure the 
maps accounted for real world constraints that impact durability and 
drivability, and noise, vibration, and harshness. Using regular octane 
fuel is consistent with the fuel octane that manufacturers specify be 
used in the majority of vehicles (manufacturers generally only specify 
premium fuel is required for higher performance models, although that 
is not always the case), and enables the modeling to account for 
important design and calibration issues associated with regular octane 
fuel. The agencies noted in the NPRM that using the updated engine maps 
addressed over-estimation of potential fuel economy improvements and 
ensured that the analysis reflected real-world constraints faced by 
manufacturers to assure engine durability and acceptable drivability. 
Importantly, assuming no change in fuel octane required to

[[Page 24405]]

operate a vehicle ensures that the agencies are modeling technology 
pathways that can improve fuel economy while maintaining vehicle 
performance, capability, and other attributes.
    Compared with the NHTSA analysis in the Draft TAR, the turbocharged 
and downsized engine maps adjusted at high torque and low speed 
operation, and at high speed operation to account for knock limitations 
when using regular octane fuel. The knock model used to develop the 
turbocharged engines was trained on production and development engines 
tested at IAV to quantify the effects of different octane fuels.\780\ 
Below the knock threshold, there is no change to the fuel consumption 
maps. The agencies noted that with the fuel octane change there are 
generally two major effects in the regions where the engine is knock-
limited: First, spark timing is retarded causing a reduction in 
combustion efficiency and hence an increase in BSFC, and second, an 
increase in combustion and exhaust temperatures requiring fuel 
enrichment to cool those temperatures for engine component protection 
and resulting in increased BSFC.781 782
---------------------------------------------------------------------------

    \780\ Knock models are based on Gamma Technology's kinetic fit 
model per the technical paper titled, ``A combustion model for IC 
engine combustion simulations with multi-component fuels,'' by 
YoungChul Ra, Rolf D. Reitz--Engine Research Center, University of 
Wisconsin-Madison.
    \781\ Fuel enrichment is extra fuel is injected at the intake 
manifold port or directly into the cylinder. Fuel vaporization and 
the fuel's thermal mass reduces combustion and exhaust temperatures. 
Changes to the air/fuel ratio also impact combustion speed which 
impacts the knock limit.
    \782\ Singh, E. and Dibble, R., ``Effectiveness of Fuel 
Enrichment on Knock Suppression in a Gasoline Spark-Ignited 
Engine,'' SAE Technical Paper 2018-01-1665, 2018, https://doi.org/10.4271/2018-01-1665.
---------------------------------------------------------------------------

    The agencies also noted that for Eng14, the turbocharged downsized 
engine with cooled exhaust gas recirculation (cEGR), cEGR was added at 
the higher speeds where further reduction in combustion temperature was 
required. The higher specific heat capacity of cEGR reduced the need 
for fuel enrichment by lowering combustion temperatures and limiting 
the amount of spark retardation necessary to manage spark knock. With 
increasing load, cEGR is also used to lower combustion temperatures to 
reduce NOx emissions. The agencies explained that because IAV's models 
are not trained for emissions, cEGR was only considered for areas that 
are knock-limited and/or to reduce combustion temperatures. Because 
cEGR has the impact of slowing down burn rates, the amount of cEGR that 
could be utilized was balanced to maintain efficient combustion. 
Combustion stability was also evaluated to assure cEGR rates did not 
cause excessive cycle-to-cycle combustion variations, which adversely 
impact drivability.\783\
---------------------------------------------------------------------------

    \783\ Heywood. B. J, Internal Combustion Engine Fundamentals, at 
413-37, McGraw-Hill (1988).
---------------------------------------------------------------------------

    Some commenters criticized these downsized turbocharged IAV maps, 
referencing deliberative EPA comments docketed pursuant to the Clean 
Air Act procedural requirements at 42 U.S.C. 7607, which stated that 
the assumptions for Eng12's fuel octane, heating value, and carbon 
content were not representative of certification fuel and did not 
appear to be consistently used for the various engine maps, concluding 
that the resultant engine maps were not representative of 
CO2 performance of turbocharged engines over the 
certification cycle. ICCT stated it appeared these concerns had not 
been addressed for the NPRM, and that ``this problem essentially 
affect[ed] all engines on the turbocharged engine pathway.'' \784\
---------------------------------------------------------------------------

    \784\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-46.
---------------------------------------------------------------------------

    The agencies disagree with ICCT's comments relating both to whether 
fuel specifications were used consistently and whether the fuel 
specifications for fuel octane, heating value and carbon content were 
representative of the same fuel. First, the EPA deliberative comments 
were resolved in the deliberative process through the clarification 
that a single fuel specification was used to develop all of the engines 
and engine maps. Therefore, the engine maps are internally consistent. 
The fuel specification was presented in the NPRM section PRIA Chapter 
6.3.2.2.17. Second, the agencies considered future fuel and emissions 
standards by using regular octane fuel for this analysis. The 
assumptions for the fuel used in this analysis align with the EPA's 
Tier 3 standards that went into effect January 1, 2017.\785\ For the 
reasons discussed further above, the agencies believe it is important 
to use Tier 3 fuel for engine maps used for rulemaking analysis.
---------------------------------------------------------------------------

    \785\ Final Rule for Control of Air Pollution from Motor 
Vehicles: Tier 3 Motor Vehicle Emission and Fuel Standards. https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-motor-vehicles-tier-3. Last accessed September 
26, 2019. Docket EPA-HQ-OAR-2011-0135.
---------------------------------------------------------------------------

    Roush claimed that the turbocharged engine maps used in the 
analysis were responsible for an overly-conservative estimate of 
underlying combustion engine efficiencies, arguing that many production 
engines available today use the same technology packages identified in 
the PRIA but with significantly higher efficiencies.\786\ Roush noted 
that the base turbocharged engine map used in the PRIA, Eng12, is 
assumed to have variable valve lift (VVL), but with a turbocharged 
engine the benefit of VVL over dual variable valve timing (VVT) is 
limited.\787\ Roush argued that almost all vehicle manufacturers use 
lower-cost dual VVT systems in their turbocharged engines, and that the 
agencies' base turbocharged engine assumption is unrealistic with a 
correspondingly high cost.\788\
---------------------------------------------------------------------------

    \786\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 16.
    \787\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 17.
    \788\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 17.
---------------------------------------------------------------------------

    Roush contrasted its critique of Eng12 with an EPA ALPHA run of a 
2016 Honda Civic 1.5L turbocharged engine (L15B7) with continuously 
variable intake and exhaust camshaft phasing (CVVT), which is less 
expensive than the CVVL, arguing that it showed greater efficiency over 
more of the engine map at a lower cost than Eng12. Roush further argued 
that since the L15B7 engine is the first generation of the new Honda 
turbocharged engine, ``even further fuel consumption improvement is 
highly likely in the period through MY2025.'' \789\
---------------------------------------------------------------------------

    \789\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 18.
---------------------------------------------------------------------------

    As the agencies explained further above, from a technical 
perspective there is no reason why the 2016 Honda Civic 1.5 L Turbo 
should have an engine map that is the same as Eng12, Eng13, or Eng14. 
The turbocharged engine technologies represented by Eng12, Eng13 and 
Eng14 are not representative of any specific engine from any one 
manufacturer. Honda's 1.5L turbocharged engine incorporates a unique 
combination of technologies including electric wastegate, sodium-filled 
exhaust valves, light weight internal components, friction reduction 
technologies, 2-stage oil pump, low viscosity oil (0W-20), and a unique 
exhaust system.\790\
---------------------------------------------------------------------------

    \790\ Honda Press Release. ``2016 Honda Civic Sedan Press Kit--
Powertrain'' October 18, 2015. https://hondanews.com/en-US/releases/2016-honda-civic-sedan-press-kit-overview?page=178. Last accessed 
Feb. 12, 2020.
---------------------------------------------------------------------------

    While there are an enormous number of different technology 
combinations that manufacturers could apply on their

[[Page 24406]]

engines, the agencies' analysis must select a reasonable number of 
configurations--in fact, the agencies analyze thousands of unique make/
model/powertrain combinations and apply them to over one hundred 
thousand unique technology combinations for each of ten classes for 
this rulemaking. See Section VI.B.3.a)(6) and Section VI.B.3 for more 
details. For turbocharged engines, the agencies selected eight 
combinations which represent a wide range of technologies, combinations 
of technologies, and effectiveness improvements for the rulemaking 
analysis, as listed in Table VI-40. Three of the combinations were 
added based on commenter's recommendations. While it is possible to 
identify other combinations, such as the unique technologies Honda 
chose for its 1.5L Turbo engine, agencies do not believe it would be 
appropriate to select all of the technologies on one specific 
manufacturer's engine for the rulemaking analysis. Doing so would, 
appropriately, raise questions about the availability of proprietary 
designs and controls to other manufacturers, among other 
considerations.
    The agencies also believe that the engine maps for Eng12, Eng13 and 
Eng14 show reasonable differences in BSFC maps that characterize the 
impact of each of these technology combinations, and differences 
relative to naturally aspirated engines. As discussed further above, 
incremental differences in BSFC are used for the rulemaking analysis. 
Roush's comments center on the comparison of absolute effectiveness 
values for a specific production vehicle, and do not address 
incremental effectiveness among a range of technologies, nor the 
appropriate baseline reference for the Honda 1.5L Turbo for technology 
content and for effectiveness. The ALPHA simulation for the 2016 Honda 
Civic 1.5L turbocharged engine provides absolute test data and has no 
baseline for assessing incremental effectiveness. Because there is no 
baseline, there is no basis for identifying which specific technologies 
have changed, nor any basis for determining the incremental 
effectiveness of each individual technology.
    Regarding Roush's comment that that further fuel consumption 
improvement for the Honda L15B7 is highly likely in the period through 
MY 2025, Roush provided no information or data on what specific 
technologies would further improve the fuel consumption of that engine. 
With no defined new technology to consider, there is no basis for 
estimating the costs, nor for estimating the effectiveness of Roush's 
assertion. Without further information, the agencies can only point to 
the additional engine technologies considered for this final rule, 
discussed further below.
    ICCT also stated that IAV's handling of cooled EGR (cEGR) in the 
engine maps was inappropriate, as IAV analyzed cEGR as a knock-
abatement technology instead of a fuel efficiency technology. ICCT 
stated that this is reason that the NPRM analysis showed no benefit to 
cEGR, and if the agencies had used EPA's properly modeled cEGR 
effectiveness based on validated data, the effectiveness of cEGR would 
have been more realistic.
    Similarly, Roush commented that cEGR application in the modeled 
turbocharged engines is excluded in engine operating modes that highly 
influence vehicle fuel economy. Roush contrasted Eng13, a turbocharged 
engine with VVT, direct injection, and cEGR, with the Mazda 2.5L 
SkyActiv Turbo engine available in the 2016 Mazda CX-9, which also 
employs cEGR.
    The agencies believe Eng14 was created and modeled using a sound 
technical methodology, using constraints that the industry uses to 
ensure the engines would meet durability and customer acceptability 
criteria. IAV turbocharged engines adopted VVT and VVL to maximize 
volumetric efficiency and improve the combustion process. Engines with 
VVT control intake and exhaust valve timing to recycle burned exhaust 
gas into the combustion chamber. The recycling of exhaust gases using 
VVT is commonly called internal EGR. Cooled EGR (cEGR) is a second 
method for diluting the incoming air that takes exhaust gases, passes 
them through a cooler to reduce their temperature, and then mixes them 
with incoming air in the intake manifold. Diluting the incoming air 
with inert exhaust gas reduces pumping losses, thereby improving BSFC. 
The dilution also reduces combustion rates, temperatures, and 
pressures, which mitigates spark knock and reduces the need for fuel 
enrichment at higher loads to control exhaust temperature for component 
durability (typically, exhaust valves and exhaust manifold). Not only 
does this exhaust gas displace some incoming air, but it also heats the 
incoming air and lowers its density. Both interactions lower the 
volumetric efficiency of the engine.\791\ Cooled EGR is a more 
effective way of reducing combustion temperature in higher load and 
higher speed engines like turbocharged engines.
---------------------------------------------------------------------------

    \791\ Volumetric efficiency (VE) in internal combustion engine 
engineering is defined as the ratio of the mass density of the air-
fuel mixture drawn into the cylinder at atmospheric pressure (during 
the intake stroke) to the mass density of the same volume of air in 
the intake manifold. Ideally, you want this to be high as possible 
to maximize thermal efficiency during the power stroke (combustion 
phase).
---------------------------------------------------------------------------

    As mentioned above, IAV developed engine specifications, including 
the rate of internal EGR and cEGR, using variation in combustion 
criteria used by industry to ensure the engines would meet durability 
and customer acceptability criteria. In addition to reducing pumping 
losses, EGR slows the combustion rate and causes combustion to be less 
consistent cycle-to-cycle as the concentration increases. Industry and 
researchers use a measurement known as coefficient of variation of 
indicated mean effective pressure (COV of IMEP) to evaluate combustion 
stability. Industry commonly recognizes values greater than 3.0 percent 
as unacceptable because above those levels, the combustion instability 
creates a noticeable and objectionable drivability problem for vehicle 
occupants, referred to as ``surge.'' Surge is perceived as the vehicle 
accelerating and decelerating erratically, instead of running smoothly. 
IAV set EGR rates at each of the engine operating conditions at the 
highest level that did not exceed 3.0 percent COV of IMEP. Therefore, 
the IAV engine maps did maximize efficiency within real-world 
constraints, similar to how manufacturers develop their engines. At the 
lower speed and load conditions of the 2-cycle tests, the COV of IMEP 
threshold was reached using internal EGR alone, so additional cEGR was 
not applied. At higher load conditions, such as the US06 cycle, cEGR 
was applied.
    ICCT's statement that the engine maps were only developed 
considering knock-abatement is inaccurate. In the PRIA Chapter 
6.3.2.2.11, the agencies discussed the application of internal EGR in 
combination with cEGR for Eng14. VVT technology, with which Eng14 is 
equipped, maximizes EGR usage first in areas where the engine primarily 
operates, such as low load and low speed area like city cycle and 
highway cycle tests used in CAFE compliance testing. Cooled EGR is 
applied at higher speed and higher load conditions, such as the US06 
test cycle.
    Using EPA's modeled cEGR would have resulted in infeasible engine 
maps because they were developed assuming the exclusive use of high 
octane Tier 2 fuel, and using a COV of IMEP threshold of 5 percent, 
which is beyond the level that is deemed acceptable to consumers in the 
real world.\792\ The use of these

[[Page 24407]]

criteria results in engine maps with BSFC levels that cannot be 
achieved by manufacturers that must ensure their engines are durable 
and are acceptable to customers with fuels that are used and available. 
The reference engine for EPA's cEGR concept was a 2010 Ricardo 
prototype V6 engine that used 98 RON fuel (93AKI or premium fuel) to 
determine effectiveness.\793\ The problems associated with using high 
octane Tier 2 to develop engine maps are discussed in detail in Section 
VI.C.1.a). The issues associated with excessive cEGR rates and COV of 
IMEP, are discussed immediately above. In addition, the cEGR engine 
maps that EPA used were never evaluated with regular octane Tier 3 fuel 
to assess the further degradation in BSFC and COV of IMEP that would 
occur where spark advance would need to be decreased to address spark 
knock, as decreasing spark advance directionally makes both BSFC and 
COV of IMEP worse.\794\ Also, because some models are still under 
development, ALPHA effectiveness estimates in the Draft TAR and derived 
for the Proposed Determination do not provide the best available basis 
for assessing effectiveness impacts.\795\ Therefore, the assumptions 
used for the EPA Draft TAR and Proposed Determination engine maps 
overstate feasible improvements and therefore do not provide meaningful 
comparisons to the engine maps used for the NPRM and final rule 
analyses.
---------------------------------------------------------------------------

    \792\ EPA Proposed Determination TSD at 2-295.
    \793\ 2016 EPA Technical Support Document at p. 2-312 in section 
2.3.4.1.9 Table 2.69. EPA-420-R-16-021, November 2016. Available at 
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf.
    \794\ 2016 EPA Technical Support Document at p. 2-312 in section 
2.3.4.1.9. EPA-420-R-16-021, November 2016. Available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf.
    \795\ Advanced Light-Duty Powertrain and Hybrid Analysis (ALPHA) 
Tool. Available at https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha#v1.0. Version 2.2. Incomplete Models in 
ALPHA2.2_TechWalkExamples\Ford Tech Walk\publish_Escape_AWD_matrix.
---------------------------------------------------------------------------

    Finally, with regards to Roush's comparison of Eng13 to the 2016 
Mazda SkyActiv-G 2.5L Turbo, the agencies believe these engines use 
technologies that are sufficiently different so as to render a 
comparison not useful, even for a very rough validation of Eng13. Most 
fundamentally, as discussed in PRIA Chapter 6.3.2.2.11 and 6.3.2.2.13, 
the Mazda 2.5L Turbo is a Miller cycle engine, whereas Eng13 is an Otto 
cycle engine. Also, the Mazda 2.5L Turbo has cEGR, whereas Eng13 does 
not.\796\ On a more detailed level, as described in PRIA Chapter 
6.3.2.2.20.10, Eng13 has a BSFC of 238 g/kwh, whereas Roush refers to 
an engine having a BSFC of 250 g/kwh.\797\ The agencies therefore 
believe comparing the 2016 Mazda SkyActiv-G 2.5L Turbo to Eng13 is not 
a useful or relevant comparison. In the PRIA, the agencies included an 
engine map for a Miller cycle engine and requested comments on whether 
it should be included in the final rule analysis. Based on the 
comments, as discussed further below, the agencies added a Miller cycle 
engine to the final rule analysis.
---------------------------------------------------------------------------

    \796\ NHTSA Benchmarking, ``Laboratory Testing of a 2016 Mazda 
CX9 2.5 I4 with a 6 Speed Transmission.'' DOT HS 812 519.
    \797\ NHTSA-2018-0067-11984 at p. 20 of 37 Figure 8.
---------------------------------------------------------------------------

(3) Non-HEV Atkinson Mode Engines
    Manufacturers use a variety of designs and technologies to obtain 
an engine's highest thermal efficiency while maintaining drivability 
and performance. While the Otto cycle has historically been used by the 
vast majority of gasoline based engines, one way to improve thermal 
efficiency is by using alternative combustion cycles. One such 
alternative combustion cycle that can be used in place of the Otto 
cycle to achieve a higher maximum thermal efficiency is the Atkinson 
cycle. Atkinson cycle operation is achieved by modifying the Otto cycle 
engines' crank and valvetrain mechanics to maintain compression ratio 
while increasing expansion ratio.798 799 800 Specifically, 
in Otto cycle operation, the exhaust valve is opened near the end of 
the power stroke, allowing exhaust gases out of the cylinder. The 
pressure in the cylinder is still about three to five atmospheres.\801\ 
Currently, there are two common approaches to achieving Atkinson Cycle 
operation: Either the exhaust valve timing or the intake valve timing 
are modified. In the first instance, the exhaust valve is not opened 
until enough expansion has occurred for the cylinder pressure to be 
equivalent to atmospheric pressure. The energy that typically is lost 
when the exhaust valve opens in Otto cycle is captured in the Atkinson 
cycle, leading to higher thermal efficiency. Modifying the intake valve 
timing, the most common way to achieve Atkinson cycle operation, 
involves allowing the intake valve to stay open during some portion of 
compression stroke. As a result, some of the fresh charge is driven 
back into the intake manifold by the raising piston so the cylinder is 
never completely filled with air, allowing optimized capture of 
combustion-created pressure.
---------------------------------------------------------------------------

    \798\ Otto cycle is a four-stroke cycle that has four piston 
movements over two engine revolutions for each cycle. First stroke: 
Intake or induction; seconds stroke: Compression; third stroke: 
Expansion or power stroke; and finally, fourth stroke: Exhaust.
    \799\ Compression ratio is the ratio of the maximum to minimum 
volume in the cylinder of an internal combustion engine.
    \800\ Expansion ratio is the ratio of maximum to minimum volume 
in the cylinder of an IC engine when the valves are closed (i.e., 
the piston is traveling from top to bottom to produce work).
    \801\ Pulkrabek. W.W. ``Engineering Fundamentals of the Internal 
Combustion Engine.'' 2nd edition. Pearson Prentice Hall, at p. 118.
---------------------------------------------------------------------------

    While Atkinson cycle engines have higher theoretical thermal 
efficiency compared to Otto cycle engines, the Atkinson cycle engine 
delivers that higher efficiency at the cost of power density.\802\ The 
reduced power density is because of lower operation pressures in the 
cylinder than in a typical Otto cycle engine. Accordingly, Atkinson 
cycle engines have been ideal for hybrid vehicles because their 
electric motor can make up for lost power density.
---------------------------------------------------------------------------

    \802\ Power density is the engine power per unit of displacement 
(= [Engine Power]/[Engine Displacement]).
---------------------------------------------------------------------------

    As vehicle technologies have become more sophisticated, 
descriptions of Atkinson cycle engines and Atkinson mode engine 
technologies have been used interchangeably, and often incorrectly, in 
association with high compression ratio (HCR) engines by the agencies 
and stakeholders. Although they both achieve an overall higher thermal 
efficiency than Otto cycle-only engines, they differ in execution 
depending on engine load. For the following discussion, Atkinson 
technologies considered in the analysis can be categorized into three 
groups: (1) Atkinson engines, (2) Atkinson-mode engines, and (3) 
Atkinson-enabled engines, which are variable valve timing engines with 
late intake closing that enables the Atkinson cycle mode. As discussed 
earlier, because power density is traded for efficiency, there is a 
limit to where Atkinson technology can be applied. While any vehicle 
could, theoretically, adopt an Atkinson-mode engine or an engine that 
enables operating in Atkinson cycle mode, the difference in vehicle 
application (high-performance versus standard-performance vehicles, 
towing requirements, trucks) leads to different effectiveness levels. 
The range of effectiveness appeared to create confusion among 
stakeholders regarding how the technology is applied to vehicles for 
compliance modeling and simulation.
    Atkinson engines are engines that operate full-time in the Atkinson 
cycle. As mentioned above, the most common method of operation used by 
Atkinson engines currently is late intake closing.

[[Page 24408]]

This approach allows backflow from the combustion chamber into the 
intake manifold, reducing the dynamic compression ratio, but providing 
a higher expansion ratio. This improves thermal efficiency but reduces 
power density. As a result of limited engine operation, these engines 
tend to have lower specific power.\803\ The lower specific power tends 
to relegate these engines to hybrid vehicles applications, as coupling 
the engines to electric motors can compensate for the lower specific 
power. The Toyota Prius is an example of a vehicle that uses an 
Atkinson engine. Typically, vehicles that use an Atkinson cycle engine 
incorporate various fuel-efficient technologies like aerodynamic 
improvements, advanced continuously variable transmissions, mass 
reduction, and many other technologies to minimize engine load and 
attain high thermal efficiency.\804\ The 2017 Toyota Prius achieved a 
peak thermal efficiency of 40 percent.\805\
---------------------------------------------------------------------------

    \803\ Specific power is the maximum power produced per 
displacement typically in units of hp/L or kw/l.
    \804\ Toyota. ``Under the Hood of the All-new Toyota Prius.'' 
Oct. 13, 2015. Available at https://global.toyota/en/detail/9827044. 
Last accessed Nov. 22, 2019.
    \805\ Matsuo, S., Ikeda, E., Ito, Y., and Nishiura, H., ``The 
New Toyota Inline 4 Cylinder 1.8L ESTEC 2ZR-FXE Gasoline Engine for 
Hybrid Car,'' SAE Technical Paper 2016-01-0684, 2016, https://doi.org/10.4271/2016-01-0684.
---------------------------------------------------------------------------

    Atkinson-mode engines are engines that use both the Otto cycle and 
Atkinson cycle during operation, switching between the modes of 
operation based on engine loads. During high loads the engine will 
operate in the power-dense Otto cycle mode, while at low loads the 
engine will operate in the higher-efficiency Atkinson cycle mode. The 
magnitude of efficiency improvement experienced by a vehicle using this 
technology is directly related to how much of the vehicle's operation 
time is spent in Atkinson mode. This means vehicles that typically 
operate at a high load, like a truck towing a trailer, will spend more 
time in the Otto mode and less time in the Atkinson cycle mode, and 
will achieve a lower overall efficiency improvement over a traditional 
Atkinson engine that operates full-time in the Atkinson cycle. As a 
result, manufacturers will try to use this type of engine in 
conjunction with other technologies that reduce engine load, which 
allows the engine to operate more frequently in Atkinson cycle mode. 
For example, manufacturers could reduce parasitic losses by 
incorporating more efficient accessory technologies, or reducing 
overall vehicle mass and aerodynamic drag. These technologies are 
enablers for Atkinson-mode engines. When these types of technologies 
are adopted, it reduces the parasitic losses and, in turn, reduces the 
time the engine is in high load region. An example of an Atkinson-mode 
engine is the MY 2017 Mazda 3.
    The last type of Atkinson-type engine, the Atkinson-enabled engine, 
can be characterized by primarily running the Otto cycle, but can 
achieve Atkinson-mode using variable valve timing (VVT) technology. 
Some engines use changes in VVT on the intake side to enable Atkinson 
cycle operation in low load, low speed operation, like city driving. 
These types of engines are typically used in applications that 
generally require higher specific power such that it would be 
infeasible to use Atkinson-mode engines or Atkinson engines. These 
vehicles tend to have higher load demands due to towing requirements, 
payload requirements, greater aerodynamic drag from larger frontal 
areas, greater tire rolling resistance from larger tires and higher 
driveline losses from four-wheel drive or all-wheel drive (e.g., SUVs 
and pickup trucks). These higher load demands tend to push these 
engines more frequently to the less efficient region of the engine map 
and limit the amount of Atkinson operation. An example of the Atkinson-
enabled engine is the Toyota MY 2017 Tacoma 3.5L 6-cylinder engine.
    EPA developed two engine maps representing non-hybrid Atkinson 
engines to support the 2016 Draft TAR, Proposed Determination, and 
first Final Determination.\806\ Referred to as ATK and ATK2, the 
engines represented a current non-hybrid Atkinson cycle engine based on 
the 2.0L 2014 Mazda SkyActiv-G (ATK) engine, and a future Atkinson 
engine concept based on the Mazda engines, but adding cooled EGR, 
cylinder deactivation, and an increased compression ratio (14:1) 
developed for full vehicle modeling and simulation (ATK2). For the 2016 
Draft TAR, the agencies adopted EPA's high compression ratio (HCR) 
engine maps as Eng24 and Eng25, which corresponded to HCR1 and HCR2 in 
the CAFE modeling.
---------------------------------------------------------------------------

    \806\ 2016 LD Draft Technical Assessment Report (TAR), Vehicle 
Greenhouse Gas Emission Standards and Corporate Average Fuel Economy 
Standards for Model Years 2022-2025; at p. 5-282. Proposed 
Determination on the Appropriateness of the Model Year 2022-2025 
Light-Duty Vehicle Greenhouse Gas Emissions Standards under the 
Midterm Evaluation; pp. 22 & A-7. Final Determination on the 
Appropriateness of the Model Year 2022-2025 Light-Duty Vehicle 
Greenhouse Gas Emissions Standards under the Midterm Evaluation, 
Response to Comments; pp. 29 & 52.
---------------------------------------------------------------------------

    The Alliance had provided significant comments on the 2016 Draft 
TAR regarding the engine maps for HCR engines.\807\ The Alliance 
detailed concerns regarding the feasibility and effectiveness of Eng24 
(HCR1) and Eng25 (HCR2). Many of the comments on the 2016 Draft TAR 
noted that the modeling projected an implausible rapid fleet 
penetration for these technologies, and overestimated effectiveness. 
Commenters stated the overestimation was due largely to modeling with 
use of high-octane fuel and the addition of other technologies like 
cEGR and cylinder deactivation (DEAC) using theoretical assumptions 
that exceed the bounds of operation of components. In contrast, other 
commenters had stated that EPA's work on the future Atkinson concept 
``has shown this pathway to be a promising alternative way to match the 
levels of improvement from a 27-bar BMEP turbocharged engine,'' and 
that ``it is prudent to assume that the robust body of evidence EPA is 
putting together based on benchmarking and modeling data is a 
reasonable assessment of the technology's potential.'' \808\
---------------------------------------------------------------------------

    \807\ Alliance of Automobile Manufacturers, Alliance of 
Automobile Manufacturers Comments on Draft Technical Assessment 
Report: Midterm Evaluation of Light-Duty Greenhouse Gas Emission 
Standards and Corporate Average Fuel Economy Standards for Model 
Years 2022-2025 (EPA-420-D-16-900, July 2016), at 45 (Sept. 26, 
2016), Docket ID EPA-HQ-OAR-2015-0827-4089 and NHTSA-2016-0068-0072.
    \808\ Union of Concerned Scientists Comments Concerning the 
Draft Technical Assessment Report for the Mid-term Evaluation of 
Model Year 2022-2025 Light-duty Vehicle Greenhouse Gas Emissions and 
Fuel Economy Standards, at 10-11.
---------------------------------------------------------------------------

    For the NPRM analysis, the agencies included EPA's engine maps. The 
agencies allowed HCR1 to be applied only for a few manufacturers that 
indicated they would pursue this technology pathway versus alternative 
pathways, such as downsized turbocharged engines. The agencies were 
also careful to maintain vehicle performance and utility attributes 
when considering the application of Atkinson-type technologies. Current 
Atkinson capable engines have incorporated other technologies to reduce 
load in order to maximize time in Atkinson operation and to offset the 
power loss partially. This includes improved accessories, addition of 
friction reduction technologies, and other technologies that reduce 
engine load. Although modern improvements to engines have allowed 
Atkinson operation to occur more often (because of lower engine loads) 
for passenger cars, larger vehicles capable of carrying more cargo and 
occupants, and towing larger and heavier trailers, have more limited 
potential Atkinson operation. Those

[[Page 24409]]

adoption features are discussed further in Section VI.C.1.e) Adoption 
Features, below.
    As stated in the NPRM, the agencies excluded the HCR2 concept 
engine from the central analysis for several reasons. First, the 
concept was not subjected to validation to assess its technical 
feasibility. The concept was only modeled with high octane Tier 2 fuel. 
The HCR2's capability to operate on regular octane Tier 3 fuel was 
assessed using non-cycle specific operation, necessitating adjustments 
to the final results to account for Tier 3 fuel properties from Tier 2 
operation, instead of simply operating the engine on Tier 3 to generate 
effectiveness estimates.\809\ As discussed further above and in Section 
VI.C.1.a), fuel octane affects engine durability, performance, 
drivability, and noise, vibration and harshness. Assumptions about 
compression ratio, EGR rates, and use of cylinder deactivation were not 
adequately validated. PRIA Chapter 6.3.2.2.20.18 discussed many 
questions about HCR2 technology's practicability as specified, 
especially in high load, low engine speed operating conditions. There 
also has been no observable physical demonstration of the technology 
assumptions. Many manufacturer engine experts questioned its technical 
feasibility and commercial practicability during the model years 
covered by the rulemaking. Stakeholders like the Alliance had 
previously asked for the engine to be removed from the rulemaking 
analyses until the performance could be validated with engine 
hardware.\810\ For these reasons, the agencies considered the HCR2 
engine too speculative to include in the NPRM central analysis. 
However, the agencies did provide a sensitivity analysis that included 
the HCR2 engine.
---------------------------------------------------------------------------

    \809\ EPA PD TSD at 2-210.
    \810\ NHTSA-2016-0068-0070 at 45.
---------------------------------------------------------------------------

    Comments on HCR1 and HCR2 varied, with commenters split on issues 
like whether HCR2 was speculative or real, whether there was technology 
in the fleet that could adequately be represented by HCR2, and the 
effectiveness of HCR2 in the analysis.
    The Alliance commented in support of the decision to exclude HCR2 
from the analysis, citing previous comments to the Draft TAR and 
proposed determination ``detailing concerns of feasibility and 
effectiveness of the non-hybrid Atkinson engine technology packages, 
including cooled exhaust gas recirculation (``CEGR'') and cylinder 
deactivation.'' \811\ Specifically, the Alliance's comments ``noted 
that the modeling projected an implausibly rapid fleet penetration of 
this complex engine technology and overestimated its effectiveness, due 
largely to modeling with high-octane fuel and the theoretical addition 
of CEGR plus cylinder deactivation.'' The Alliance concluded that ``the 
inexplicably high benefits ascribed to this theoretical combination of 
technologies has not been validated by physical testing.'' Ford 
commented that previous assessments had ``over-estimated both the 
effectiveness and near-term penetration of advanced Atkinson technology 
powertrains,'' stating that ``[t]he effectiveness of the `futured' 
Atkinson package (HCR2) that includes cooled exhaust gas recirculation 
(CEGR) and cylinder deactivation (DEAC) is excessively high, primarily 
due to overly-optimistic efficiencies in the base engine map, 
insufficient accounting of CEGR and DEAC integration losses, and no 
accounting of the impact of 91RON Tier 3 test fuel. Given the 
speculative and optimistic modeling of this technology combination, 
Ford supports limiting the use of HCR2 technology to reference only, as 
described in the Proposed Rule.'' \812\ Separately, in support of its 
overarching comments that the NPRM modeling better reflected reality 
over prior regulatory assessments, Toyota commented that the 
effectiveness estimates for Atkinson cycle engine technology in the 
NPRM may still have been overstated.\813\
---------------------------------------------------------------------------

    \811\ NHTSA-2018-0067-12073.
    \812\ NHTSA-2018-0067-11928.
    \813\ NHTSA-2018-0067-12150.
---------------------------------------------------------------------------

    In contrast, CARB, ICCT, Meszler Engineering Services, UCS, and 
other stakeholders commented in different respects, with the broad 
themes being: (1) That the change in approach towards HCR engines from 
the Draft TAR and Proposed Determination to the NPRM was not justified, 
was inadequately justified, or was based on justification from the 
industry and not the agencies' own independent judgment; (2) that HCR2 
as defined by EPA does exist and therefore should be used in the 
analysis; and (3) that even if HCR2 technology does not exist exactly 
as EPA defined it, other technologies in the fleet provide the same 
level of efficiency improvement as HCR2 and therefore it should be used 
in the analysis. Many of these commenters stated that if HCR2 had been 
allowed in the compliance analysis, as shown in the NPRM sensitivity 
analysis allowing HCR2 to be applied, compliance costs would have been 
reduced dramatically, ``on par with NHTSA and EPA estimates in the 
TAR.'' 814 815
---------------------------------------------------------------------------

    \814\ NHTSA-2018-0067-11741.
    \815\ NRDC, Attachement2_CAFE Model Tech Issues.pdf. Docket No. 
NHTSA-2018-0067-11723, at 7-13. ICCT, Full Comments Summary. Docket 
No. NHTSA-2018-0067-117411, at I-2.
---------------------------------------------------------------------------

    Specifically, ICCT, CARB, and UCS took issue with the agencies' 
description of HCR2 technology as speculative, stating that description 
contrasted with how EPA described the technology in prior documents. 
ICCT commented that ``in the Draft TAR and Final Determination, EPA 
observed the real-world advances toward production vehicles using HCR2 
technology, and determined that that technology could be adopted by 
automakers during the compliance period.'' \816\ ICCT stated that in 
the NPRM, ``without rational explanation, the agencies now describe 
this technology as `speculative' and have omitted the technology from 
their primary compliance scenarios altogether.'' CARB similarly 
commented that ``[t]he fact that the Agencies, especially EPA, make [a 
statement that HCR2 is entirely speculative] is genuinely impossible to 
credit.'' \817\ In support, all three commenters referenced EPA's 
hardware testing of a European Mazda engine,\818\ with ICCT stating 
that HCR2 was dismissed as entirely speculative ``despite the careful 
benchmarking of improved HCR engines by EPA,'' while CARB and UCS 
similarly cited this hardware testing to rebut the Alliance's assertion 
that the effectiveness values for HCR2 was ``seriously overestimated.''
---------------------------------------------------------------------------

    \816\ NHTSA-2018-0067-11741.
    \817\ NHTSA-2018-0067-11873.
    \818\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy 
Improvements from the Implementation of cEGR and CDA on an Atkinson 
Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
2017-01-1016.
---------------------------------------------------------------------------

    ICCT also took issue with the NPRM statements that ``many engine 
experts questioned [HCR2's] technical feasibility and near-term 
commercial practicability,'' \819\ and that ``[s]takeholders asked for 
the engine to be removed from compliance simulations until the 
performance could be validated with engine hardware,'' with references 
to comments from Fiat-Chrysler (stating ``Remove ATK2 from OMEGA model 
until the performance is validated'' and ``ATK2--High Compression 
engines coupled with Cylinder Deactivation and Cooled EGR are unlikely 
to deliver modeled results, meet customer needs, or be ready for 
commercial application.''),\820\ and comments from the Alliance of 
Automobile Manufacturers, stating that

[[Page 24410]]

``[There] is no current example of combined Atkinson, plus cooled EGR, 
plus cylinder deactivation technology in the present fleet to verify 
EPA's modeled benefits and . . . EPA could not provide physical test 
results replicating its modeled benefits of these combined 
technologies.'' \821\ ICCT stated that the agencies did not identify 
any such comments or evidence from engine experts, or agency analysis 
of them. ICCT stated that ``it is clear that NHTSA is deferring to 
stakeholders, and that EPA has been forced to defer to NHTSA.''
---------------------------------------------------------------------------

    \819\ 83 FR 43038.
    \820\ Id. (citing NHTSA-2016-0068-0082).
    \821\ Id. (citing EPA-HQ-OAR-2015-0827-6156).
---------------------------------------------------------------------------

    ICCT also cited interagency review documents where EPA stated 
``[t]here are Atkinson engine vehicles on the road today (2018 [Toyota] 
Camry and Corolla with cooled EGR and the 2019 Mazda CX5 and Mazda6 
with cylinder deac) that use high geometric compression ratio Atkinson 
cycle technology that is improved from the first generation, MY2012 
vintage ``HCR1'' technology. While it is true that no production 
vehicle has both cooled EGR and cylinder deac, as the EPA ``HCR2'' 
engine did, nonetheless, these existing engines demonstrate better 
efficiency than estimated by EPA. Therefore, it would be appropriate to 
continue to use EPA's cooled EGR + deac engine map to represent 
``HCR2'' engines.'' \822\
---------------------------------------------------------------------------

    \822\ NHTSA-2018-0067-11741, Attachment3_ICCT 15page summary and 
full comments appendix, at I-10 (citing Docket Entry: E.O. 12866 
Review Materials for The Safer Affordable Fuel-Efficient (SAFE) 
Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light 
Trucks NPRM, Docket ID EPA-HQ-OAR-2018-0283-0453 (hereinafter 
``EO12866 Review Materials''), File: 
``EO_12866_Review_EPA_comments_on_the_NPRM_sent_to_OMB,_June_29,_2018
'' at 82, https://www.regulations.gov/document?D=EPA-HQ-OAR-2018-0283-0453).
---------------------------------------------------------------------------

    More specifically regarding the technical specifications of the 
HCR2 engine, ICCT and others stated that EPA had already addressed 
concerns brought by the Alliance \823\ on (1) the base engine fuel 
consumption maps used as the foundation of the HCR2 engine map; \824\ 
(2) practical limitations for cEGR to limit engine knock; \825\ (3) the 
reliance on the availability of cylinder deactivation at unrealistic 
speed and load operating points; (4) the impact of 91 RON market and 
certification test fuels; and (5) the ability to implement HCR2 
technology in existing vehicle architectures.\826\
---------------------------------------------------------------------------

    \823\ EPA-HQ-OAR-2015-0827-4089; EPA-HQ-OAR-2015-0827-6156.
    \824\ NHTSA-2018-0067-11741 (``EPA showed how its ``difference'' 
engine maps validly represented performance of the ATK2 [HCR2] 
packages including on different fuels (pp. 301-02); and that the 
difference maps submitted in the industry comment ``provided no 
information to compare vintage or application of the actual engine 
or engines tested, and did not state whether or not testing was 
conducted,'' lacking any information on ``test and/or analytical 
methods, assumptions, fuel properties, environment test conditions, 
how the engine was controlled or how control was modeled, the number 
of data points gathered to generate the AAM `difference map' to 
assure that identical testing and a sufficient fit of data was 
performed'' (p. 301). In addition, EPA showed that concerns about 
knock due to use of cooled exhaust gas recirculation had been 
considered and resolved by ignition improvements (p. 302).'').
    \825\ NHTSA-2018-0067-12039 (``The agencies appear to have 
relied upon the differences between anti-knock properties of Tier 2 
and Tier 3 fuels, mistakenly focusing solely on octane while 
ignoring ethanol content. . . . this fails to acknowledge the anti-
knock benefit of charge cooling related to ethanol, which more than 
compensates for the change in octane. HCR2 therefore should not be 
omitted out of concerns around knock.'').
    \826\ NHTSA-2018-0067-11741. ICCT stated that EPA had previously 
concluded that existing engine architectures were ``well adapted for 
[HCR] technology, and well adapted for the emerging next level HCR2 
package of technologies, since the foundational technologies of 
gasoline direct injection, increased valve phasing authority, higher 
compression ratios, and cooled exhaust gas recirculation are already 
in widespread use.'' ICCT also commented that ``EPA correctly 
observed that there was sufficient lead time to adopt the HCR2 
technology before MY2022 and that it could be incorporated without 
requiring major vehicle redesigns.''
---------------------------------------------------------------------------

    CARB, UCS, and ICCT all stated, in different terms, that even if 
HCR2 technology does not exist exactly as EPA defined it, other 
technologies that exist in the fleet provide the same level of 
efficiency improvement as HCR2, specifically referencing the MY 2018 
Toyota Camry engine and various Mazda engines, and claiming that HCR2 
should therefore be used in the analysis. Specifically, CARB stated 
that these engines ``are already achieving similar efficiency as the 
modeled HCR2 package even though they don't have the full complement of 
technologies (i.e., CEGR and DEAC) used in the HCR2 package.'' \827\ 
CARB stated that these engines' ``existence as production engines today 
certainly speaks to the feasibility of this technology for modeling 
that goes out to 2030MY.'' \828\ Similarly, UCS stated that while the 
2018 Toyota Camry engine ``does not have all of the features of the 
HCR2 package constructed by EPA, it achieves similar levels of 
performance, thus rendering the agencies' rationale for excluding HCR2 
moot--this is a production vehicle using Tier 3 fuel which achieves 
performance equivalent to HCR2.'' \829\ Similarly, ICCT cited their own 
analysis of the 2018 Toyota Camry for the propositions that the package 
of technologies on the Camry exceeds the efficiency gains projected by 
EPA's OMEGA model, meaning that EPA's projections for the HCR2 engine 
might understate its effectiveness, and the early problems with low-end 
torque losses associated with Atkinson cycle engines have been 
completely solved.\830\ ICCT stated that ``[t]his evaluation of a real 
world vehicle that comes close to meeting all of the elements of an 
HCR2 engine makes it clear that HCR2 engines are far from a speculative 
technology.''
---------------------------------------------------------------------------

    \827\ NHTSA-2018-0067-11873.
    \828\ NHTSA-2018-0067-11873.
    \829\ NHTSA-2018-0067-12039.
    \830\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    ICCT and CARB also took issue with the agencies' justification for 
not using the HCR2 engine map as a simulation proxy for other new 
engine technology, specifically the statement that:

    It is important to conduct a thorough evaluation of the actual 
new production engines to measure the brake specific fuel 
consumption and to characterize the improvements attributable to 
friction and thermal efficiency before drawing conclusions. Using 
vehicle level data may misrepresent or conflate complex interactions 
between a high thermal efficiency engine, engine friction reduction, 
accessory load improvements, transmission technologies, mass 
reduction, aerodynamics, rolling resistance, and other vehicle 
technologies.\831\
---------------------------------------------------------------------------

    \831\ 83 FR 43038.

    Both commenters also took issue with the agencies' statement that 
existing technologies in the NPRM version of the CAFE model could work 
together appropriately to represent an HCR1 engine with additional 
efficiency improvements.\832\
---------------------------------------------------------------------------

    \832\ 83 FR 43038.
---------------------------------------------------------------------------

    ICCT stated that the complexity associated with the package of 
improvements in the Camry engine was common to all of the technology 
packages included in either OMEGA or CAFE modeling, and was neither a 
new issue nor an issue that precludes making reasonable engineering 
judgments. ICCT stated that the agencies projected efficiency estimates 
for other technology packages without engine maps from a production 
engine, citing the agencies' approach to modeling ADEAC technology, and 
concluded that the purpose of full vehicle simulation modeling is to 
project the efficiency impact when several different parts of the 
vehicle are simultaneously upgraded. ICCT stated that ``[i]f reasonable 
estimates could be made for ADEAC without fully validated engine maps, 
there is no reason to exclude other technologies on these grounds, 
especially considering the deep expertise by the agencies and their 
state-of-the-art technology simulation capabilities with the ALPHA 
modeling.'' Similarly, HDS noted that in contrast to the agencies' 
exclusion of HCR2 due to

[[Page 24411]]

unresolved issues associated with knock mitigation and cylinder 
deactivation, ``the 2018 analysis included Advanced Cylinder De-
activation (ADEAC) which has recently come to market readiness.'' \833\
---------------------------------------------------------------------------

    \833\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    Merriam-Webster's dictionary defines speculative as ``involving, 
based on, or constituting intellectual speculation,'' and also, 
``theoretical rather than demonstrable.'' \834\ To be clear, most 
engines maps used in this analysis--IAV engine maps included--are 
theoretical, although they are built based on benchmarked engine data, 
and additional fuel-economy-improving technologies are added through 
modeling and simulation. But that does not mean that these engines are 
speculative. Although the IAV engine maps are not meant to model any 
manufacturer's particular engine, many, if not all, technology 
combinations have been implemented in real-world engines.
---------------------------------------------------------------------------

    \834\ Definition of ``speculative,'' https://www.merriam-webster.com/dictionary/speculative.
---------------------------------------------------------------------------

    The agencies qualified the HCR2 engine as speculative because ``no 
production engine as outlined in the EPA SAE paper has ever been 
commercially produced or even produced as a prototype in a lab setting. 
Furthermore, the engine map has not been validated with hardware and 
bench data, even on a prototype level (as no such engine exists to test 
to validate the engine map).'' \835\ It is important to distinguish 
theoretical engines maps with technology combinations that have been 
proven through real-world testing and operation, from the HCR2 engine 
map, that was created using a combination of validated individual 
component models, but the resulting engine system model and generated 
engine map were not fully validated against actual hardware.
---------------------------------------------------------------------------

    \835\ 83 FR 43038.
---------------------------------------------------------------------------

    The Alliance and individual automakers have repeatedly provided 
comments on agency actions with their assessment of the feasibility of 
the HCR2 engine, including comments ICCT referenced, stating the EPA 
had addressed concerns brought by the Alliance in the Proposed 
Determination Technical Support Document.\836\ The agencies agree with 
ICCT that EPA provided responses to comments about HCR2 assumptions and 
engine maps in the Technical Support Document, the Proposed 
Determination, and the 2017 Final Determination. However, the agencies 
considered the matter further after receiving extensive comments on 
HCR2 for the NPRM. The agencies have concluded responses did not 
directly and fully address the technical concerns raised by the 
Alliance. Further, new data and information has become available since 
the Proposed and Final Determination that is directly relevant to the 
use of EPA's engine maps in this analysis.
---------------------------------------------------------------------------

    \836\ Also important to note regarding ICCT's comment, the 
Alliance comment cited in the NPRM came from a section of the 
Alliance's comments titled, ``EPA's Response to Alliance Comments 
Regarding Atkinson Cycle Engine Technology Benefits is Inadequate,'' 
which seems to suggest that EPA did not address concerns brought by 
the Alliance in the Proposed Determination Technical Support 
Document.
---------------------------------------------------------------------------

    First, it is important to provide background information about 
ICCT's comments referencing previous discussions from the TAR, Proposed 
Determination and Final Determination. For the 2016 Draft TAR, EPA 
initially created the ATK1 and ATK2 engine maps based on the MY 2014 
Mazda 2.0L SKYACTIV-G engine. The EPA benchmarked the Mazda engine, 
then modeled increasing the efficiency of the Mazda engine map by 
simulating the application of additional technologies using GT-Power 
models. The Alliance and FCA commented on the 2016 Draft TAR suggesting 
the EPA's development of the ATK1 and ATK2 engine maps were flawed 
because the maps were developed based on optimistic baseline engine 
characterization of the Mazda engine. The Alliance provided evidence of 
the flaws in EPA's characterization by comparing EPA's published base 
engine data, developed using Tier 2 certification gasoline, to engine 
data benchmarked by USCAR. USCAR benchmarked their own Mazda Skyactiv 
engine map using a 91 RON fuel. The comparison resulted in the creation 
of a ``difference map'' that showed where the two data sets diverged. 
The ``difference map'' implied there were areas of significant 
divergence, calling into question the data upon which the ATK1 and ATK2 
models are based. The EPA responded stating ``[the Alliance] did not 
provide data or other information to substantiate its claim that EPA's 
engine dynamometer fuel consumption measurements using a MY2014 Mazda 
OEM production 2.0L SKYACTIV-G, upon which the ATK2 packages from the 
TAR analysis are based, were in any way unrepresentative of this 
engine's actual performance.'' \837\ ICCT cited in their NPRM comments 
that the EPA's discussion of these ``difference maps'' supported their 
statement that ``[i]n fact, in the Technical Support Document for EPA's 
Proposed and 2017 Final Determination, EPA addressed all these concerns 
brought forth by the Alliance [regarding HCR2] (including the costs and 
effectiveness impacts of using regular octane fuel instead of premium 
fuel).''
---------------------------------------------------------------------------

    \837\ EPA PD TSD at 2-299.
---------------------------------------------------------------------------

    It is understandable why ICCT may have thought this discussion 
addressed concerns raised about the HCR2 map; however, review of the 
Alliance's original Draft TAR comments makes it clear the Alliance's 
initial comments addressed the benchmarking of the MY 2014 Mazda 13:1 
SKYACTIV-G engine itself. The Alliance's original comments, expressed 
concern over the modeled effectiveness of the advanced Atkinson 
technology packages because of the baseline engine data used. The 
Alliance suggested the effectiveness is likely overestimated due to 
multiple flaws in the benchmarking and modeling approaches taken by 
EPA. Only the benchmarking is addressed by EPA's response to the 
``difference maps,'' not the concerns about modeling approach.
    The Alliance's concerns about modeling included the accuracy of the 
base engine fuel consumption maps (to the extent the baseline engine 
maps were overly optimistic, the modeled ATK maps were optimistic), 
limitations for cEGR to mitigate engine knock, limitations of cylinder 
deactivation, and the impact of fuels.\838\ After further review, the 
agencies determined the Alliance's concerns were not fully addressed, 
resulting in a closer review of the ATK model development process.
---------------------------------------------------------------------------

    \838\ EPA-HQ-OAR-2015-0827-4089.
---------------------------------------------------------------------------

    Review of the engine model development showed the engine map was 
generated assuming the use of high octane fuel, and the follow-up 
engine dynamometer validation testing also used high octane fuel.\839\ 
The characterization of the baseline Mazda Skyactiv engine showed 1-3 
percent increase in thermal efficiency across a large portion of the 
engine map when operated on Tier 2 fuel versus lower octane 
fuel.840 841 The increase in engine

[[Page 24412]]

thermal efficiency, caused by the higher octane fuel, is anticipated to 
be amplified when applying ATK technologies. ATK technologies increase 
efficiency by increasing the pressure in cylinder during combustion; 
however, at the same time the increased pressure increases risk of 
knock. For more discussion on engine knock, see Section VI.C.1.a). 
Ultimately, it is expected that the ATK1 and ATK2 engines would show a 
larger improvement in thermal efficiency as a result of being developed 
assuming a high-octane fuel versus the 1-3 percent improvement observed 
on the baseline Mazda Skyactiv engine.
---------------------------------------------------------------------------

    \839\ Ellies, B., Schenk, C., and Dekraker, P., ``Benchmarking 
and Hardware-in-the-Loop Operation of a 2014 MAZDA SkyActiv 2.0L 
13:1 Compression Ratio Engine,'' SAE Technical Paper 2016-01-1007, 
2016, doi:10.4271/2016-01-1007.
    \840\ The engine was first run on LEVIII-compliant certification 
fuel which has a 7 psi vapor pressure and 88aki. This fuel is 
similar to Tier 3 fuel with exception of the vapor pressure which is 
required to be 9 psi to meet Tier 3 certification. It was then 
tested on Tier 2 certification fuel (93aki) to assess effects of 
higher octane fuel on engine operation and efficiency.
    \841\ Ellies, B., Schenk, C., and Dekraker, P., ``Benchmarking 
and Hardware-in-the-Loop Operation of a 2014 MAZDA SkyActiv 2.0L 
13:1 Compression Ratio Engine,'' SAE Technical Paper 2016-01-1007, 
2016, doi:10.4271/2016-01-1007.
    Schenk, C. and Dekraker, P., ``Potential Fuel Economy 
Improvements from the Implementation of cEGR and CDA on an Atkinson 
Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
2017-01-1016.
---------------------------------------------------------------------------

    A further limitation was revealed during the agencies review of the 
ATK model development. The limitation was in how COV of IMEP, an 
important indicator of combustion stability, was not accounted for 
directly in the model. The 0-D/1-D models used for investigating cEGR 
effectiveness could not adequately simulate changes to COV of IMEP. To 
compensate for the lack of an appropriate model, limits on cEGR were 
based on literature values for unrelated engine technologies.\842\ As a 
result, there was no direct evaluation of combustion stability while 
evaluating the feasibility of the engine concept.
---------------------------------------------------------------------------

    \842\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy 
Improvements from the Implementation of cEGR and CDA on an Atkinson 
Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
2017-01-1016.
---------------------------------------------------------------------------

    In contrast, for the NPRM and final rule analysis, IAV engines were 
optimized using Tier 3 fuel, to balance performance and fuel 
consumption. The majority of baseline vehicles are specified to operate 
on 87 AKI fuel, therefore lower octane fuel was used to maintain 
baseline functionality. The IAV engine maps were all derived from a 
consistent baseline engine and were also optimized using a validated 
kinetic knock model, and using a COV of IMEP threshold of 3 percent.
    These differences in model construction caused an inconsistency 
that resulted in unrealistic improvements in fuel economy and 
CO2 emissions for the HCR engine technologies, whereas the 
IAV engine maps reflect more realistic accounting for the improvements. 
The use of high octane fuel and lack of combustion stability modeling 
are complimentary issues that have compounded effects when combined. 
For example, the use of high octane fuel allows more advanced spark 
timing which both increases efficiency and improves combustion 
stability, allowing higher cEGR rates before reaching acceptable limits 
for drivability. The compound effect is greater than the simply adding 
together individual effects, causing a potentially further unrealistic 
increase in effectiveness. At a minimum, it is uncertain how using Tier 
3 fuel in the HCR2 engine would impact the BSFC of the engine, as there 
was no direct evaluation of the feasibility of the engine concept's 
ability to operate on regular octane fuel. The cost for the 
effectiveness of the HCR2 technology also is inconsistent with the cost 
of the effectiveness improvement values for the technologies in the 
2015 NAS report.\843\ In considering all of this information, the 
agencies, believe the HCR2 engine map overstates the capabilities of 
the technology and decided not to use that engine map for the final 
rule analysis.
---------------------------------------------------------------------------

    \843\ 2015 NAS at p. 90 and 91.
---------------------------------------------------------------------------

    However, the agencies believe the HCR1 engine map does reflect 
improvements that are representative of the technology in the 
rulemaking timeframe. For the final rule, to reflect better the 
incremental effectiveness for a low-cost version of HCR technology, the 
agencies added the HCR0 engine for the analysis. The specification of 
this engine was provided in the NPRM PRIA as Eng22b. Using this engine 
improves the estimated incremental effectiveness because the 
incremental engine changes from were directly specified for the 
modeling. HCR0 is the first engine in the HCR path that a manufacturer 
could adopt. Accordingly, the non-HEV Atkinson engine maps used for the 
NPRM and final rule central analysis fit into the three defined 
categories as follows: (1) Eng26 is an HEV Atkinson Cycle engine; (2) 
in the NPRM analysis, Atkinson-mode engines were characterized by Eng24 
(HCR1), and for the final rule analysis, Atkinson-mode engines are 
characterized by Eng22b (HCR0) and Eng24 (HCR1); and (3) Atkinson-
enabled engines are characterized by the different VVT engine 
technologies identified earlier in basic engine discussions and shown 
on Table VI-41 and Table VI-42.
    Regarding the ability of manufacturers to adapt the engine 
architecture to practical use, the agencies see merit in observations 
from both manufacturers and other groups. ICCT is correct in their 
observation that some production engines have integrated combinations 
of the technologies, including SGDI, VVT and cEGR. Furthermore, the 
agencies agree with ICCT that an engine could be built integrating all 
the technologies represented in the HCR2 engine model. However, the 
agencies also agree with the Alliance's comments to the 2016 Draft TAR 
that applying all the technologies to an engine that only has some of 
the technologies would require a significant redesign of the powertrain 
package. The redesign would need to accommodate the new hardware 
integration, controls and emissions calibration, OBD development and 
other major efforts. As discussed further in Section VI.C.1.e), the 
agencies believe these considerations impact how quickly and widely the 
technology could be implemented in the rulemaking timeframe.
    The agencies also disagree with commenters that the HCR2 engine map 
should be used as a proxy for other vehicles in the fleet that achieve 
high thermal efficiency. None of the existing vehicles that commenters 
cited, like the 2019 Toyota Camry and Corolla with cEGR or the 2019 
Mazda CX5 and Mazda 6 with cylinder deactivation, include the same 
combination of technologies as the HCR2 engine. Unlike other engine 
technologies in the NPRM and the final rule analysis, no engines in the 
market or in prototype stages exist that have the combined technology 
specifications of the HCR2. Accordingly, there is no production vehicle 
that demonstrates the combination of technologies as applied in the 
HCR2 engine that (1) is feasible, and (2) can achieve the same 
effectiveness as the modeled HCR2 engine. The NPRM highlighted concerns 
about using the HCR2 engine map as a proxy for new engine technologies 
that achieve high thermal efficiency, specifically that:

    It is important to conduct a thorough evaluation of the actual 
new production engines to measure the brake specific fuel 
consumption and to characterize the improvements attributable to 
friction and thermal efficiency before drawing conclusions. Using 
vehicle level data may misrepresent or conflate complex interactions 
between a high thermal efficiency engine, engine friction reduction, 
accessory load improvements, transmission technologies, mass 
reduction, aerodynamics, rolling resistance, and other vehicle 
technologies.\844\
---------------------------------------------------------------------------

    \844\ 83 FR 43038.

    The agencies continue to believe this is true, and Toyota's 
comments that the Camry improvements were due to more than just the 
engine improvements, as discussed further below, provide further 
support to this conclusion.
    Several commenters cited EPA's SAE paper discussing the use of the 
HCR2 engine model and comparing it to the benchmarking of a 2018 Toyota 
Camry

[[Page 24413]]

2.5L engine.845 846 The commenters cited the HCR2 engine's 
similarities to the Toyota Camry engine as a reason to employ the 
technology model broadly across the entire vehicle fleet, including 
applying it to pickup trucks such as the Toyota Tacoma. In the paper, 
EPA benchmarked a 2018 Toyota Camry 2.5L Atkinson cycle engine equipped 
with cEGR. EPA created a full vehicle model (the exemplar vehicle) 
based on the benchmarked data for use in the ALPHA modeling tool. The 
full vehicle simulation was used to compare the HCR2 engine to the 
Camry's 2.5L engine, and showed some similarities. The paper implied 
that it is possible to adopt more technologies to the MY 2018 Camry, 
like cylinder deactivation, to meet future standards.
---------------------------------------------------------------------------

    \845\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al., 
``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine 
with Cooled-EGR,'' SAE Int. J. Adv. & Curr. Prac. in Mobility 
1(2):601-638, 2019, https://doi.org/10.4271/2019-01-0249.
    \846\ Duleep, K.G., ``Review of the Technology Costs and 
Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report, 
H-D Systems, October 2018, at p. 37.
---------------------------------------------------------------------------

    This paper, and the comments relying on it--specifically that it 
shows that additional technologies can be added to the MY 2018 Camry 
engine to meet future standards--were the subject of considerable 
debate in the rulemaking docket. Toyota provided supplemental comments 
regarding issues Toyota had with the modeling and simulation. These 
included a detailed discussion on why HCR2 is not a reasonable model of 
the 2018 Toyota Camry engine. Toyota identified other technologies that 
contributed to the overall thermal efficiency of the 2018 Camry 
compared to previous generation.\847\ Toyota stated that the 2018 
Toyota Camry employed numerous technologies like SGDI, cEGR, optimized 
intake system, optimized exhaust system, optimized piston design, 
laser-cladded valve seats, VVT, engine friction reduction, variable oil 
pump, and electric coolant pump, that all contributed to the engine's 
improved efficiency over the previous version.\848\
---------------------------------------------------------------------------

    \847\ NHTSA-2018-0067-12431. Supplemental Comments of Toyota 
Motor North America, Inc. (7/15/19) at 1-2; NHTSA-2018-0067-12376. 
Supplemental Comments of Toyota Motor North America, Inc. (3/25/19) 
at 1.
    \848\ Hakariya, M., Toda, T., and Sakai, M., ``The New Toyota 
Inline 4-Cylinder 2.5L Gasoline Engine,'' SAE Technical Paper 2017-
01-1021, 2017, available at https://doi.org/10.4271/2017-01-1021.
---------------------------------------------------------------------------

    In addition, Toyota stated:

[T]he 2018 Exemplar Vehicle that is based on the baseline 2018 
Toyota Camry was equipped with engine start stop that doesn't exist 
on the production vehicle. Cylinder deactivation was added to the 
2025 exemplar vehicle as a protentional enhancement. We acknowledged 
that adding cylinder deactivation to the Atkinson-cycle engines is 
technically possible and would provide some fuel economy benefits. 
However, the primary function of cylinder deactivation is to reduce 
engine pumping losses which the Atkinson cycle and EGR already 
accomplish. The diminishing return on the cylinder deactivation, 
Atkinson cycle and EGR are further exaggerated by smaller 4-cylinder 
engines.

    This assessment aligns with the 2015 NAS committee report that 
estimated a 0.7 percent fuel consumption improvement for adoption of 
cylinder deactivation for DOHC and SOHC V6 and V8 engines.\849\ The 
agencies agree with Toyota and the NAS assessment that applying 
cylinder deactivation in small cylinder count engines is subject to 
diminishing returns.
---------------------------------------------------------------------------

    \849\ 2015 NAS at p. 34.
---------------------------------------------------------------------------

    The agencies agree with Toyota that the presence of the advanced 
technologies, in addition to the HCR technology, contributed to the 
performance of the Camry. The analysis already provides benefits for 
the other advanced technologies individually, and risks, if not 
ensures, double counting these benefits if the HCR2 model is used (as 
discussed above and in VI.B). Likely double counting of technology 
effectiveness further supported the agencies' choice not to use the 
HCR2 model for the final rule analysis.
    The agencies disagree that the approach taken to modeling ADEAC 
technology should similarly apply to modeling the HCR2 engine, or that 
because ADEAC just recently entered the market and was employed in the 
modeling, HCR2 should be as well. As discussed further below, the 
effectiveness estimates for ADEAC were based on extensive discussions 
with suppliers and manufacturers that provided CBI data, and technical 
publications.\850\ The effectiveness estimates provided for ADEAC 
represented the effects of applying a single technology, and not a 
combined estimate for several technologies applied at once. Moreover, 
as commenters noted, ADEAC had recently ``come to market readiness,'' 
\851\ compared to the HCR2 technology which cannot be found, as 
modeled, in the market, or even in prototype form. As discussed 
throughout this document, the preferred approach for the NPRM and final 
rule was to isolate the effectiveness improvement attributable to 
specific technologies and apply those through full vehicle simulations 
to capture technology synergies and dis-synergies appropriately.
---------------------------------------------------------------------------

    \850\ Eisazadeh-Far, K. and Younkins, M., ``Fuel Economy Gains 
through Dynamic-Skip-Fire in Spark Ignition Engines,'' SAE Technical 
Paper 2016-01-0672, 2016, doi:10.4271/2016-01-0672.
    \851\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    The agencies also disagree with ICCT's comment that the agencies 
were simply deferring to stakeholders, or that EPA was simply deferring 
to NHTSA regarding the feasibility of the HCR2 engine. It is reasonable 
to assume that the automobile manufacturers that belong to the Alliance 
employ some engine experts that are qualified to speak on the 
feasibility of an engine. Not just one or two manufacturers objected to 
the HCR2 engine; the Alliance commented on behalf of its members in 
support of the exclusion of the engine from the analysis,\852\ and this 
exclusion was further supported by comments from individual automakers 
as well. Toyota, the automaker cited by several commenters as closest 
to implementing HCR2 technology stated in supplemental comments that 
(1) the HCR2 is not representative of its engine technology; \853\ and 
(2) Toyota believes there are diminishing returns for implementing the 
HCR2 technologies.\854\ The agencies received no comments from 
stakeholders that manufacture engines in support of the HCR2 
technology's feasibility and potential future adoption.
---------------------------------------------------------------------------

    \852\ NHTSA-2018-0067-12073, at 139.
    \853\ Comment from Toyota NHTSA-2018-0067-12376 (``While the 
agencies' definitions for the different levels of Atkinson 
technology seem to have evolved, the 2018 Camry is clearly not 
equipped with HCR2 technology.'').
    \854\ Comment from Toyota NHTSA-2018-0067-12376 (``advanced 
cylinder deactivation has not yet been established when packaged 
with an Atkinson-cycle engine. Both technologies play similar roles 
in reducing engine pumping losses which can led to diminishing 
returns when combined.'').
---------------------------------------------------------------------------

    For HCR technology, the agencies carefully considered comments to 
the NPRM and the available data, and concluded it is appropriate to 
include HCR0 and HCR1 engine models for the final rule analysis. The 
engine maps for those technologies provide the best estimates for the 
effectiveness of HCR technology relative to the engine maps for the 
other engine technologies used for the analysis. The agencies have 
reconsidered issues associated with the HCR2 engine models and maps. 
The agencies find that significant technical questions and issues 
remain and the engine maps very likely overstate the feasible amount of 
effectiveness that could be achieved by the represented technologies. 
Therefore, HCR2 technology is not included for the final rule analysis.
(4) HEV Atkinson Cycle Engines
    Three types of Atkinson technology were discussed in the previous 
section.

[[Page 24414]]

HEV Atkinson cycle engines fall in the first category, operating solely 
or primarily in Atkinson mode, supported by an electric drive.
    Engine map 26 (Eng26) is the model of the HEV/PHEV Atkinson cycle 
engine used for the NPRM and final rule analysis. The engine was based 
on Argonne's Advanced Mobility Technology Laboratory (AMTL) 2010 Toyota 
Prius test data and published literature.\855\ Argonne's AMTL is 
continuously involved in research and testing of advanced technologies, 
especially in areas of electrification, and has a large existing 
database of test data from advanced technology vehicles.\856\ As a 
result of Argonne's continued research, a 2017 Toyota Prius was 
characterized for an independent project. Argonne updated the HEV 
Atkinson cycle engine using the new Prius data to reflect the 41 
percent thermal efficiency of the new 2017 system.\857\ The 
electrification technology groups that used Eng26 include powersplit 
hybrid vehicles (SHEVPS) and plug-in powersplit hybrid vehicles 
(PHEV20/50).
---------------------------------------------------------------------------

    \855\ ``2010 Toyota Prius.'' http://www.anl.gov/energy-systems/group/downloadable-dynamometer-database/hybrid-electric-vehicles/2010-toyota-prius. Last accessed April, 2018.
    \856\ ANL AMTL Downloadable Dynamometer Database (D3). https://www.anl.gov/es/downloadable-dynamometer-database. Last accessed Dec. 
05, 2019.
    \857\ Carney, D. ``Toyota unveils more new gasoline ICEs with 
40% thermal efficiency.'' SAE. April 4, 2018. https://www.sae.org/news/2018/04/toyota-unveils-more-new-gasoline-ices-with-40-thermal-efficiency. Last accessed Dec. 5, 2019.
---------------------------------------------------------------------------

(5) Advanced Cylinder Deactivation Technologies
    Advanced cylinder deactivation (ADEAC) systems, also known as 
rolling or dynamic cylinder deactivation systems, allow a further 
degree of cylinder deactivation than the base DEAC. ADEAC allows the 
engine to vary the percentage of cylinders deactivated and the sequence 
in which cylinders are deactivated, essentially providing 
``displacement on demand'' for low load operations.
    ADEAC systems may be integrated into the valvetrains with moderate 
modifications on OHV engines. However, while the ADEAC operating 
concept remains the same on DOHC engines, the valvetrain hardware 
configuration is very different, and application on DOHC engines is 
projected to be more costly per cylinder due to the valvetrain 
differences.
    The agencies discussed assumptions and effectiveness for the ADEAC 
package in the NPRM preamble.\858\ The initial review of this 
technology was based on a technical publication that used a MY 2010 
engine design that had incorporated a SOHC VVT basic engine.\859\ Other 
preproduction 8-cylinder OHV prototype vehicles with ADEAC were briefly 
evaluated for this analysis, but no production versions of the 
technology have been studied.\860\ For ADEAC fuel consumption 
effectiveness values, no engine map was available at the time of the 
NPRM analysis. Accordingly, the agencies took the effectiveness values 
as predicted by full vehicle simulations of a DEAC engine with SGDI, 
VVL, and VVT, and added 3 percent or 6 percent respectively for I-4 
engines and V-6 or V-8 engines, and cross-referenced CBI data to 
quality check this approach.
---------------------------------------------------------------------------

    \858\ 83 FR 43038-39.
    \859\ Wilcutts, M., Switkes, J., Shost, M., and Tripathi, A., 
``Design and Benefits of Dynamic Skip Fire Strategies for Cylinder 
Deactivated Engines,'' SAE Int. J. Engines 6(1):278-288, 2013, 
available at https://doi.org/10.4271/2013-01-0359. Eisazadeh-Far, K. 
and Younkins, M., ``Fuel Economy Gains through Dynamic-Skip-Fire in 
Spark Ignition Engines,'' SAE Technical Paper 2016-01-0672, 2016, 
available at https://doi.org/10.4271/2016-01-0672.
    \860\ EPA, 2018. ``Benchmarking and Characterization of a Full 
Continuous Cylinder Deactivation System.'' Presented at the SAE 
World Congress, April 10-12, 2018. Retrieved from https://www.regulations.gov/document?D=EPA-HQOAR-2018-0283-0029.
---------------------------------------------------------------------------

    The agencies noted two potential approaches to including advanced 
cylinder deactivation in the full-scale Argonne simulation modeling 
analysis for the final rule. First, the agencies proposed using IAV 
Eng25a, which was developed to capture the maximum benefits of advanced 
cylinder deactivation with several constraints that could include 
emissions, cold start, NVH, and durability. Second, the agencies 
proposed using a technique developed by Argonne in coordination with 
NHTSA to split the overall engine data into individual cylinder data 
and compute overall torque and the fuel consumption rate by accounting 
for whether each cylinder is active or inactive. The agencies sought 
comment on using either approach in the final rule analysis to capture 
best the benefits of advanced cylinder deactivation.
    CARB, ICCT, Meszler Engineering Services, HDS, and UCS provided a 
mixed set of comments on numerous aspects of ADEAC in the NPRM 
analysis.\861\ Broadly, HDS commented on a need to describe ADEAC 
technology better: ``The 2018 analysis also utilized Advanced Cylinder 
Deactivation in its analysis but the package components were not 
completely explained in the PRIA.'' \862\ Other stakeholders provided 
comments on ADEAC adoption features, effectiveness, and cost, which are 
discussed below.
---------------------------------------------------------------------------

    \861\ ICCT Docket # NHTSA-2018-0067-11741 at I-12, Duleep Docket 
# NHTSA-2018-0067-11873 at 108, Meszler Docket # NHTSA-2018-0067-
11723 at p. 26.
    \862\ Duleep, K.G., ``Review of the Technology Costs and 
Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report, 
H-D Systems, October 2018, at p. 17.
---------------------------------------------------------------------------

    The agencies discussed assumptions and effectiveness for the ADEAC 
package in the NPRM preamble.\863\ The initial review of this 
technology was based on a technical publication that used a MY 2010 
engine design incorporating SOHC and VVT.\864\ After determining the 
MY2010 engine design was not representative of the analysis fleet, the 
agencies used effectiveness values based on CBI data. The MY2017 
baseline fleet reflects technology updates such as SGDI and DEAC that 
could adopt ADEAC incrementally in the final rule analysis. The cost 
and effectiveness for ADEAC reflects the baseline engine. The 2015 NAS 
Committee estimated an 0.7 percent fuel consumption improvement for 
adoption of cylinder deactivation for V6s and V8s 
engines.865 866
---------------------------------------------------------------------------

    \863\ 83 FR 43038-39.
    \864\ Wilcutts, M., Switkes, J., Shost, M., and Tripathi, A., 
``Design and Benefits of Dynamic Skip Fire Strategies for Cylinder 
Deactivated Engines,'' SAE Int. J. Engines 6(1):278-288, 2013, 
available at https://doi.org/10.4271/2013-01-0359. Eisazadeh-Far, K. 
and Younkins, M., ``Fuel Economy Gains through Dynamic-Skip-Fire in 
Spark Ignition Engines,'' SAE Technical Paper 2016-01-0672, 2016, 
available at https://doi.org/10.4271/2016-01-0672.
    \865\ Applied after VVT and VVL.
    \866\ Applied before VVT and VVL.
---------------------------------------------------------------------------

    The agencies requested comments on alternative methods to estimate 
ADEAC effectiveness but received no comments regarding either approach 
mentioned in the NPRM. For the final rule analysis, the agencies used 
effectiveness values as predicted by full vehicle simulations of a DEAC 
engine with SGDI, VVL, and VVT, and added 3 percent or 6 percent 
respectively for I-4 engines and V-6 or V-8 engines for the naturally 
aspirated engines. Effectiveness for turbocharged engines used 1.5 
percent and 3 percent values, as predicted by full vehicle simulation 
of a TURBOD engine for I4 and V6/V8, respectively. Without sufficient 
data to simulate ADEAC, both the IAV and Argonne methodologies 
described in the NPRM provided questionable estimates for ADEAC. These 
errors would have propagated across other technology combinations in 
the analysis. The estimates used for ADEAC and TURBOD for the final 
rule analysis are also in line with EPA

[[Page 24415]]

estimates discussed in their SAE technical publications.\867\
---------------------------------------------------------------------------

    \867\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al., 
``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine 
with Cooled-EGR,'' SAE Int. J. Adv. & Curr. Prac. in Mobility 
1(2):601-638, 2019, https://doi.org/10.4271/2019-01-0249 at pp. 19-
21.
---------------------------------------------------------------------------

    For the final rule analysis, the agencies used the same 
effectiveness values for ADEAC applied to naturally aspirated engines 
as in the NPRM, and incorporated estimated effectiveness values for 
TURBOAD to represent ADEAC on downsized turbocharged engines.
(6) Miller Cycle Engines
    In the proposed rule, the agencies provided two engine maps 
representative of Miller cycle and Eboost engines with 48V battery 
systems. The Miller cycle engine (Eng23b) and Miller cycle engine with 
Eboost (Eng23c) specifications were provided in the PRIA but were not 
used in the NPRM analysis,\868\ although the agencies sought comment on 
the specifications used for the modeling.
---------------------------------------------------------------------------

    \868\ NPRM PRIA at p. 307-09.
---------------------------------------------------------------------------

    Roush on behalf of CARB, ICCT, Meszler Engineering on behalf NRDC, 
HDS, and UCS, commented that the agencies did not consider the 
combination of turbocharging and Miller cycle.\869\ Specifically, Roush 
argued that the agencies' omission of an engine that utilizes a 
combination of turbocharging and Miller cycle was unreasonable because 
it is already in production, specifically on the VW 2.0L EA888 Gen3B--
DI. Roush stated this omission would limit the effectiveness for 
turbocharged engines and cause the adoption of more expensive 
solutions, thereby overstating the cost to achieve target fuel economy 
levels. Similarly, Roush pointed to the omission of an engine that uses 
a variable geometry turbocharger as an error in the agencies' vehicle 
modeling; Roush pointed to VW's EA211 TSI Evo engine available in 
Europe in 2017 as an example of an engine in production that enables 
cost-effective Miller cycle applications.
---------------------------------------------------------------------------

    \869\ NHTSA-2018-0067-11985. HD systems at p, 34; ICCT at p. 
102; NRDC Attachment 2 at p.16.
---------------------------------------------------------------------------

    In response to these comments, the agencies added and used both 
Miller cycle-type engines and Miller cycle engines with electric assist 
for the final rule analysis. Discussed earlier in this section, the 
agencies developed engine maps for additional combinations of 
technologies for the final rule, including engine maps that became 
available after the NPRM analysis was completed but before the NPRM was 
published. For the final rule analysis, the agencies have included a 
Miller cycle engine, Eng23b (VTG), as another available engine 
technology. The specification of this engine was discussed in PRIA 
Chapter 6.3.2.2.20.20.2.2 and the costs are based on the 2015 NAS 
estimates for this technology.
(7) Variable Compression Ratio Engines
    Variable compression ratio (VCR) engines work by changing the 
length of the piston stroke of the engine to operate at a more optimal 
compression ratio and improve thermal efficiency over the full range of 
engine operating conditions. Engines using VCR technology are currently 
in production, but appear to be targeted primarily towards limited 
production, high performance and very high BMEP (27-30 bar) 
applications.
    A few manufacturers and suppliers provided information about VCR 
technologies, and several design concepts were reviewed that could 
achieve a similar functional outcome. In addition to design concept 
differences, intellectual property ownership complicates the ability of 
the agencies to define a VCR hardware system that could be widely 
adopted across the industry.
    For the NPRM analysis, the agencies provided specifications of a 
VCR engine (Eng26a) in the PRIA for review and comment.\870\ However 
the VCR engine was not used in the NPRM analysis.
---------------------------------------------------------------------------

    \870\ NPRM PRIA at pp. 304-06.
---------------------------------------------------------------------------

    The Alliance commented in support of the exclusion of variable 
compression ratio engines from the analysis, stating that the 
technology is still in early development, and too speculative to be 
included at this time. The Alliance also stated that the technology is 
unlikely to attain significant penetration in the MY 2026 timeframe due 
to intellectual property protection associated with early 
implementations and its likely application primarily to high-
performance vehicles. The Alliance also cited the technology's price as 
a potential barrier to adoption.\871\ Similarly, Ford commented that:
---------------------------------------------------------------------------

    \871\ NHTSA-2018-0067-12073 (``At least one source also 
indicates a steep price to this technology--``at least $3,000 more 
to produce than a standard 16-valve double-overhead-camshaft four-
cylinder.'').

    [VCR technology] is likely to be adopted only for premium/
limited-market vehicles in the near future. We also agree that 
intellectual property protections on early implementations will 
further inhibit significant fleet penetration. Incorporation of VCR 
requires a new or highly modified engine architecture, necessitating 
major investment from both the engineering and manufacturing 
standpoints. Sharing/commonality across engine families would be 
greatly limited.'' 872 873
---------------------------------------------------------------------------

    \872\ NHTSA-2018-0067-11928.
    \873\ NHTSA-2018-0067-11928 at p. 9.

    Similarly, other automakers commented on a confidential basis that 
several main hurdles prevented them from employing VCR engines, 
including the complexity of VCR engines and the associated cost of 
those complex parts.
    UCS commented that the agencies did not consider VCR engine 
technologies in the NPRM analysis.\874\ They stated that the technology 
was not modeled, nor was it incorporated into the CAFE model. UCS 
argued that Nissan's VC-Turbo engine is part of a strategy to improve 
fuel efficiency for Nissan's luxury vehicles by 30-35 percent over 
previous models, which would be enough to exceed the vehicle's 
regulatory targets without any credits. UCS concluded that given VCR 
technology is being put into production in a high-volume vehicle, there 
is no reason for the agencies to exclude its adoption.
---------------------------------------------------------------------------

    \874\ NHTSA-2018-0067-12039 at p. 6.
---------------------------------------------------------------------------

    The agencies agreed with comments to include VCR engine 
technologies in the final rule analysis and on further technical 
consideration, the agencies have added a VCR engine to the engine 
technologies list manufacturers could adopt. However, the agencies 
limited the adoption of the VCR engine technology to Nissan only. VCR 
engines are complex, costly by design, and synergetic with mainstream 
technologies like downsize turbocharging, making it unlikely that a 
manufacturer that has already started down an incongruent technology 
path would adopt VCR technology.
(8) Diesel Engines
    Diesel engines have several characteristics that result in superior 
fuel efficiency over traditional gasoline engines, including reduced 
pumping losses due to lack of (or greatly reduced) throttling, high 
pressure direct injection of fuel, a combustion cycle that operates at 
a higher compression ratio, and a very lean air/fuel mixture relative 
to an equivalent-performance gasoline engine.\875\ However, diesel 
technologies requires additional enablers, such as a NOX 
adsorption catalyst system or a urea/ammonia selective catalytic 
reduction system, for control of NOX emissions.
---------------------------------------------------------------------------

    \875\ Diesel cycle is also a four-stroke cycle like the Otto 
Cycle, except in the Intake stroke no fuel is injected and fuel is 
injected late in the compression stroke at higher pressure and 
temperature.
---------------------------------------------------------------------------

    For the NPRM, the agencies modeled one diesel engine, represented 
by

[[Page 24416]]

Eng17,\876\ which was termed ``ADSL'' in the CAFE modeling. DSLI, a 
more advanced diesel engine, was modeled using a 4.5 percent fixed 
effectiveness improvements over ADSL.
---------------------------------------------------------------------------

    \876\ Docket ID NHTSA-2018-0067-1972. NPRM PRIA at p. 295.
---------------------------------------------------------------------------

    CARB commented that diesel technologies are essentially locked out 
of being selected in the CAFE model because of the high cost.\877\ They 
state that diesel technology is only selected in rare instances.
---------------------------------------------------------------------------

    \877\ Docket ID NHTSA-2018-0067-11873. CARB at 108.
---------------------------------------------------------------------------

    The agencies agree that diesel technology is rarely selected. The 
technologies required to meet diesel emissions standards are costlier 
compared to gasoline technologies, particularly in the rulemaking 
timeframe. For example, the 2015 NAS report determined that in the 
current market, ``vehicles with diesel engines are priced an average of 
more than $4,000 more than comparably equipped gasoline vehicles.'' 
\878\ Furthermore, the NAS report stated that the ``Carbon Penalty'' 
makes it harder for manufactures to meet CO2 standards 
because of the higher carbon density in the diesel fuel compared to 
gasoline that results in higher CO2 per gallon.\879\ In 
addition, the market for diesel vehicles has stagnated at around 1 
percent for many years after it peaked at 5.9 percent in 1981, 
according to the EPA Trends Report.\880\ The agencies believe that the 
modeled cost of diesel engines appropriately prevents their widespread 
adoption in the analysis.
---------------------------------------------------------------------------

    \878\ 2015 NAS at 123-24.
    \879\ 2015 NAS Findings 3.3 and 3.4 at p. 120.
    \880\ EPA, ``The 2018 EPA Automotive Trends Report.'' March 
2019. EPA-420-R-19-002. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at pp. 5 & 6. Last accessed 
December 16, 2019.
---------------------------------------------------------------------------

    UCS commented that the agencies restricted cylinder deactivation 
technologies to only naturally aspirated gasoline engines.\881\ In 
response to this and other comments, the agencies have allowed diesel 
engines to adopt ADEAC for this final rule analysis. These engines were 
designated as DSLIAD to represent diesel engines with ADEAC, and were 
modeled using a 7.5 percent fixed effectiveness improvement on top of 
DSLI. This effectiveness improvement of ADEAC on diesel engines is 
based on the review of technical publications discussed earlier in 
Section VI.C.1.c)(5).
---------------------------------------------------------------------------

    \881\ Docket ID NHTSA-2018-0067-12039, at p. 3.
---------------------------------------------------------------------------

(9) Alternative Fuel Engines
    CNG engines use compressed natural gas as a fuel source. The fuel 
storage and supply systems for these engines differ tremendously from 
gasoline, diesel, and flex fuel vehicles. CNG engines were a baseline-
only technology and were not applied to any vehicle that was not 
already CNG-based in NHTSA's analysis, per EPCA/EISA's restrictions on 
considering dedicated alternative fueled vehicles to set fuel economy 
standards.882 883 However, for the EPA program the agencies 
allowed any vehicle to adopt CNG engines. The NPRM MY 2016 analysis 
fleet did not include any dedicated CNG vehicles to simulate in the 
CAFE Model.
---------------------------------------------------------------------------

    \882\ NHTSA's provisions for dedicated alternative fuel vehicles 
in 49 U.S.C. 32905(a) state that the fuel economy of any dedicated 
automobile manufactured after 1992 shall be measured based on the 
fuel content of the alternative fuel used to operate the automobile. 
A gallon of liquid alternative fuel used to operate a dedicated 
automobile is deemed to contain 0.15 gallon of fuel. Under EPCA, for 
dedicated alternative fuel vehicles, there are no limits or phase-
out for this special fuel economy calculation, unlike for duel-
fueled vehicles, as discussed below.
    \883\ EPA's provisions for dedicated alternative fuel vehicles 
that are able to run on compressed natural gas (CNG) currently are 
eligible for an advanced technology multiplier credit for MYs 2017-
2021.
---------------------------------------------------------------------------

    In addition, for the NPRM and this final rule analysis, NHTSA 
modified the CAFE model to include the specific provisions related to 
AFVs under the CO2 standards. In particular, the CAFE model 
now carries a full representation of the production multipliers related 
to electric vehicles, fuel cell vehicles, plug-in hybrids, and CNG 
vehicles, all of which vary by year through MY 2021.
(10) Emerging Gasoline Engine Technologies
    Manufacturers, suppliers, and researchers continue to create a 
diverse set of fuel economy technologies, some of which are still in 
the early stages of the development and commercialization process. Due 
to uncertainties in the cost and capabilities of emerging technologies, 
some new and pre-production technologies are not a part of the CAFE 
model simulation. As discussed throughout this section and in VI.B.3, 
the agencies declined to include technologies in the analysis where the 
agencies did not believe those technologies would be feasible in the 
rulemaking timeframe, or the agencies did not have appropriate data 
upon which to generate an estimate of how effective the technology is 
that could be applied across the ten vehicle classes. Evaluating and 
benchmarking promising fuel economy technologies as they enter 
production-intent stages of development continues to be a priority as 
commercial development matures.
    UCS and ICCT commented that the agencies should consider novel 
engine designs.\884\ Specifically, ICCT stated that the agencies should 
consider a more advanced HCR technology called HCCI (similar to Mazda's 
Skyactiv-X) by estimating efficiency and cost to EPA's process that 
assigned effectiveness estimates using LPM. They stated that ``the 
agencies developed estimates for ADEAC in the NPRM and the associated 
modeling even without conclusive and independently verifiable 
effectiveness.''
---------------------------------------------------------------------------

    \884\ ICCT, Full Comments Summary. Docket No. NHTSA-2018-0067-
117411, at I-17 to I-19.
    UCS, Comment. Docket No. NHTSA-2018-0067-12039, at pp. 6 & 7.
---------------------------------------------------------------------------

    In response to comments, a number of technologies were added for 
the final rule analysis, and adoption features were refined 
accordingly, as discussed further in Section VI.C.1.e). New engine 
technologies and combinations include Atkinson engine technology 
allowed with P2 HEV, new high compression ratio engine (HCR0), variable 
compression ratio engine, variable geometry turbo engine, variable 
geometry turbo with electric assist engine, diesel with advanced 
cylinder deactivation engine, turbo with cylinder deactivation engine, 
diesel with manual transmission, diesel with start-stop, and PHEV-turbo 
with 20 mile range, and PHEV-turbo with 50 mile range.
    The agencies also disagree with ICCT's comment that because ADEAC 
was developed without ``conclusive and independently verifiable 
effectiveness'' estimates, and as such the agencies should allow HCCI 
technology as well. First, conclusive estimates for ADEAC effectiveness 
were based on CBI data from both manufacturers and suppliers, technical 
publications, and engineering judgement. The references can be reviewed 
in the previous Section VI.C.1.c)(5) Advanced Cylinder Deactivation 
Technologies. In addition, the agencies benchmarked the first prototype 
vehicle equipped with skip-fire, and discussed potential application of 
it for other engines. A similar level of data has not been made 
available for HCCI engine technologies.
    The agencies also believe that the technology associated with Mazda 
SkyActiv-X has been mischaracterized by ICCT and other commenters, and 
declined to include a specific representation of the SkyActiv-X family 
of technologies in the analysis for two reasons. The engine known as 
Skyactiv-X is characterized by Mazda as a unique spark plug controlled 
compression ignition (SPCCI) technology, 2-liter displacement, 4-
cylinder engine with mechanical compression ratio of 16.3:1 operating 
on 95 RON fuel (91 AKI) with

[[Page 24417]]

a mild hybrid system.\885\ The NPRM and this final rule analysis may 
not have the exact technology combination associated with this vehicle, 
but the analysis does include technologies that are representative of 
them, that could enable the benefits employed by the Mazda engine. A 
mild hybrid system is available for adoption in both the NPRM and this 
final rule analysis.
---------------------------------------------------------------------------

    \885\ Mazda Press Release. ``Revolutionary Mazda Skyactiv-x 
engine details confirmed sales start.'' May 6, 2019. https://www.mazda-press.com/eu/news/2019/revolutionary-mazda-skyactiv-x-engine-details-confirmed-as-sales-start/. Last accessed Dec, 11, 
2019.
---------------------------------------------------------------------------

    Also, the effectiveness associated with this engine was from 
European test cycles and cannot be compared for U.S. application. 
European compliance tests are significantly different than those in the 
U.S., especially when it comes to fuel type and test cycles. Any 
effectiveness data provided for this engine or any non-U.S. engine 
cannot be used for U.S. vehicle application without an adjustment for 
fuel and emissions. For example, the higher-octane fuel used in Europe 
enables engines to operate at higher compression ratios across wider 
areas of engine operation.
    The agencies further believe that with the technology additions for 
the final rule discussed in previous sections, the analysis reasonably 
represents the suite of engine technologies that could be available in 
the rulemaking time frame. Manufacturers, suppliers, and researchers 
continue to create a diverse set of fuel economy technologies. However, 
due to the uncertainties in the cost, manufacturing, and intellectual 
property concerns like those identified by commenters, the agencies did 
not consider prototype technologies in the final rule analysis.
(11) Engine Lubrication and Friction Reduction Technologies
    Manufacturers have already widely adopted both lubrication and 
friction reduction technologies. Previous agency analysis considered 
these improvements in combination as Improved Low Friction Lubricants 
and Engine Friction Reduction (LUBEFR). The NPRM analysis included 
advanced engine maps that already assume application of low-friction 
lubricants and engine friction reduction technologies, and therefore 
additional levels of friction reduction were not considered. Low-
friction lubricants including low viscosity and advanced low-friction 
lubricant oils are now available, and widely used. Manufacturers may 
make engine changes and conduct durability testing to accommodate the 
lubricants. The level of low-friction lubricants exceeded 85 percent 
penetration in the MY 2016 fleet.\886\ Reduction of engine friction can 
be achieved through low-tension piston rings, roller cam followers, 
improved material coatings, more optimal thermal management, piston 
surface treatments, and other improvements in the design of engine 
components and subsystems that improve efficient engine operation.
---------------------------------------------------------------------------

    \886\ NPRM CAFE Model Market Data file.
---------------------------------------------------------------------------

    Meszler Engineering on behalf of NRDC commented that ``the NPRM 
CAFE model no longer considers advanced lubricants and evolutionary 
friction reduction (LUBEFR) to be adoptable. As a result, no fuel 
efficiency improvement credits are available. Engine friction reduction 
is an ongoing evolutionary process that should generate benefits on the 
order of 5 percent or so increase in fuel economy over a multiyear 
forecast period, with costs totaling approximately $100. Moreover, the 
technology is a benefit of ongoing industry research and evolutionary 
engine improvements so that it is easily `adoptable' and deployed 
throughout the fleet. Accordingly, NHTSA should revise the NPRM CAFE 
model to reinstate the ability to adopt evolutionary friction reduction 
technology.'' \887\
---------------------------------------------------------------------------

    \887\ Meszler Engineering. Docket ID NHTSA-2018-0067-11723, at 
p. 32.
---------------------------------------------------------------------------

    The agencies disagree with Meszler that a five percent fuel economy 
improvement attributable to lubricants and evolutionary friction 
reduction is continuously feasible. The MY 2017 baseline vehicles have 
incorporated many technologies like low viscosity engine oil, 
integrated exhaust manifold for faster oil warmup, and internal 
component friction reduction.\888\ \889\ \890\ The LUB and EFR 
technologies are a legacy of the existing rulemaking work going back to 
the 2010 CAFE and CO2 rule for MY 2012 to MY 2016.\891\ The 
agencies believe that many of these technologies have been incorporated 
in many of the engines in the baseline fleet, and therefore the engine 
maps used for the NPRM and final rule analysis incorporated them as 
well. Furthermore, manufactures have raised concerns over issues with 
further decreasing oil viscosity; specifically, manufacturers have 
articulated concerns that damage caused by low speed pre-ignition 
(LSPI) \892\ can damage an engine.\893\ \894\ \895\
---------------------------------------------------------------------------

    \888\ Wards Auto. ``Infiniti's Brilliantly Downsized V-6 Turbo 
Shines.'' July 11, 2017. Available at https://www.wardsauto.com/print/engines/infiniti-s-brilliantly-downsized-v-6-turbo-shines. 
Last accessed Dec. 11, 2019. Nissan Motor Corp. ``Mirror Bore 
Coating.'' Available at https://www.nissan-global.com/EN/TECHNOLOGY/OVERVIEW/mirror_bore_coating.html. Last accessed Dec 11, 2019.
    \889\ Toyota's 2AR-FE I4 and 2GR-FE V6 use 0-W20.
    \890\ Audi Media Center. ``Efficiency and driving pleasure: 
innovative V engines at Audi.'' Available at https://www.audi-mediacenter.com/en/techday-on-combustion-engine-technology-8738/efficiency-and-driving-pleasure-innovative-v-engines-at-audi-8748. 
Last accessed Dec.11, 2019.
    \891\ 75 FR 25373.
    \892\ LSPI is an abnormal combustion event in which the fuel-air 
mixture ignites before intended, causing excessive pressures inside 
the engine's cylinders. In mild cases, this can cause engine noise, 
but when severe enough, LSPI can cause engine damage. There are 
several factors that contribute to LSPI, of which lubricating oil 
has been observed to be one.
    \893\ Motor Magazine. ``Will ILSAC GF-6 Ever Be Approved?'' Nov, 
20, 2018. Available at http://newsletter.motor.com/2018/20181120/!ID_Infineum_ILSAC_GF-6.html. Last accessed Dec 11, 2019.
    \894\ Chevron. ``Low Speed Pre-ignition.'' Available at https://www.oronite.com/about/news/low-speed-pre-ignition.aspx. Last 
accessed Dec. 11, 2019.
    \895\ Elliott, I., Sztenderowicz, M., Sinha, K., Takeuchi, Y. et 
al., ``Understanding Low Speed Pre-Ignition Phenomena across Turbo-
Charged GDI Engines and Impact on Future Engine Oil Design.'' SAE 
Technical Paper 2015-01-2028, 2015, available at https://doi.org/10.4271/2015-01-2028.
---------------------------------------------------------------------------

    In response to the comment that engine friction reduction 
technology is evolutionary technology, the agencies introduced one 
level of friction reduction (EFR) for the final rule analysis. The 
agencies estimated a 1.4 percent effectiveness for this type of 
technology based on the 2015 NAS report assessment of further 
improvements in lubrication and friction.\896\
---------------------------------------------------------------------------

    \896\ 2015 NAS at pp. 28 & 29.
---------------------------------------------------------------------------

d) How the Agencies Assign Engine Technologies to the Baseline Fleet

[[Page 24418]]

    Manufacturers have made significant improvements in fuel economy 
and CO2 emissions reductions since the MY 2012 rulemaking 
analysis.\897\ \898\ The agencies expended substantial effort to update 
the analysis fleet from the MY 2016 representative fleet used for the 
NPRM to a MY 2017 analysis fleet used for this final rulemaking to 
capture the technologies manufacturers have used to increase their 
fleet's fuel economy and CO2 emissions performance. Detailed 
discussion of the model year 2017 fleet development and application can 
be found in VI.B.1. The agencies extensively updated the new MY 2017 
fleet engine technologies using available manufacturer final model year 
CAFE compliance submissions to the agencies, as well as manufacturer 
press release specifications, agency-sponsored vehicle benchmarking 
studies, review of available technical publications, and through 
manufacturer CBI.\899\
---------------------------------------------------------------------------

    \897\ EPA. ``2018 EPA Automotive Trends Report'' 12 pp, 421 K, 
EPA-420-S-19-001, March 2019. https://www.epa.gov/automotive-trends/download-automotive-trends-report#Full%20Report last accessed Feb. 
12, 2020
    \898\ FOTW #1108, Nov 18, 2019: Fuel Economy Guide Shows the 
Number of Conventional Gasoline Vehicle Models Achieving 45 miles 
per gallon or Greater is Increasing. DOE VTO. Available at https://www.energy.gov/eere/vehicles/articles/fotw-1108-november-18-2019-fuel-economy-guide-shows-number-conventional. Last accessed Nov 18, 
2019.
    \899\ NPRM CAFE Market Data file.
---------------------------------------------------------------------------

    The data for each manufacturer was used to determine which 
platforms shared engines and to establish the leader-follower 
relationships between vehicles. Within each manufacturer's fleet, 
engines were assigned unique identification designations based on 
configuration, and technologies applied, along with other 
characteristics. The data were also used to identify the most similar 
engine among the IAV engine maps, as discussed in Section VI.C.1.
    Just like the real-world vehicle variants, the CAFE model considers 
differences between each vehicle like base performance and higher 
performance levels. For example, the 2017 Ford F150 has many variants 
with different types of engines like the 2.7L turbocharged V6, 3.3L 
naturally-aspirated V6, 3.5L turbocharged V6, and 5L naturally-
aspirated V8. In contrast to the LPM, the CAFE model rosters each 
variant level and powertrain application individually. This variation 
is accounted for as engine technologies are assigned in the analysis 
fleet.
    As a result of new information available since publication of the 
NPRM and comments received to the NPRM, the agencies included 
additional engine technologies in the compliance analysis, expanding 
the total number of engine technologies available from 16 to 23. This 
expansion is a direct result of comments received to the NPRM and 
further enables the agencies' capabilities to accurately and, 
realistically, characterize the technologies present on an engine found 
in the analysis fleet. This collection of technologies represents the 
best available information the agencies have, at the time of this 
action, regarding both currently available engine technologies and 
engine technologies that could be feasible for application to the U.S. 
fleet during the rulemaking timeframe. The agencies believe this effort 
has yielded the most technology-rich and accurate analysis fleet 
utilized by the CAFE model to date.
    In some cases, however, it was necessary for the agencies to 
substitute an engine map that closely represented an engine technology 
that were effectively the same, or, based on engineering judgement, 
were the best available proxy at the time of the analysis. For example, 
many manufacturers offer their own proprietary VVT engine technologies 
and so the agencies assigned the same engine map for all of these VVT 
in the baseline fleet. The CAFE model uses compliance CAFE and 
CO2 values for baseline vehicles and so it's not as relevant 
to have exact technology assignment type as it more important to 
provide the advanced vehicle have adopted to date. For further 
discussion of this see section VI.A.3 Fuel-Savings Technologies. This 
substitution was necessary, in some cases, where an ``exact-match'' 
engine map was not available for application to a specific vehicle and/
or vehicle specific engine application. The agencies leveraged a series 
of engine operating characteristic maps developed by industry suppliers 
and, in some cases, the agencies themselves, to assign the closest 
baseline engine map for the analysis.
    As discussed in Section VI.C.1.b), these engine maps provide 
operational characteristics such as horsepower, torque, or efficiency 
at a specified point in an engine's operational range. These 
operational maps are developed based on a given set of engine 
characteristics and technologies applied to that engine. Engine maps 
are closely held by vehicle manufacturers and are typically considered 
intellectual property. As such, vehicle manufacturers are not typically 
willing provide the operational maps to the agencies, where it would 
ultimately be in the purview of competitors. In some instances, 
manufacturer engine maps are published in media such as technical 
papers or conference presentation materials. However, these publicly 
available engine maps are, in nearly all instances, void of critical 
information that would enable their use for meaningful simulation and 
modeling.
    Therefore, the agencies are generally limited to the catalog of 
engine maps they have developed through contracts and, where possible, 
in-house which, in turn, yields the need for sound, engineering 
judgement-based substitution of an engine map as a proxy for an engine 
application in the marketplace. Unfortunately, this is necessary as the 
agencies are unable to fund the development of engines maps for every 
possible engine and technology combination available for sale. However, 
it is important to note the agencies do have a substantial catalog of 
engine maps to leverage and continue to fund the development of new 
maps as new technologies enter the marketplace. Additional information 
on the agencies' catalog of engine maps used for this this final 
rulemaking can be found in Section VI.C.1.b).
    Some engine technologies are designated in the CAFE Model as 
``baseline only'' technologies, meaning these are characteristics such 
as engine configuration, architecture, or a technology that is 
considered inherent to the fleet for the given model year, an example 
for the MY 2017 fleet used in this analysis is variable-valve-timing 
(VVT). Beyond the aforementioned configurations and technology, engine 
technologies that can be applied to a future engine and, eventually, to 
a vehicle in the compliance modeling are only available at a vehicle 
redesign. As such, a vehicle will only adopt a new engine according to 
the application schedule defined as a CAFE model input.
e) Engine Adoption Features
    Engine adoption features are defined through mechanisms like 
technology path logic or the application of selection logic, refresh 
and redesign cycles, and phase-in capacity limits. Most of the 
technology adoption features from the NPRM have been carried over for 
the final rule analysis. However, the final rule analysis also included 
adoption features for the new technologies incorporated in the final 
rule analysis. For a detailed discussion of CAFE model path logic for 
the final rule analysis, including technology supersession logic and 
technology mutual exclusivity logic, please see Section IV.
    Figure VI-18 and Figure VI-19 below show the engine technology 
paths used for the NPRM and this final rule analysis, respectively. The 
engine

[[Page 24419]]

technology paths have increased to incorporate new advanced 
technologies manufacturers could adopt into their fleet.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.152

[GRAPHIC] [TIFF OMITTED] TR30AP20.153

BILLING CODE 4910-59-C
    Similar to the 2012 final rule for MYs 2017-2025, this final rule 
analysis also considered real-world limits when the defining the rate 
at which technologies can be deployed.\900\ During the rulemaking 
timeframe, manufacturers are expected to go through the normal 
automotive business cycle of redesigning and upgrading their light-duty 
vehicle products. This allows manufacturers the time needed to 
incorporate fuel economy improving and CO2 reducing 
technologies into their normal business cycle. This is important 
because it has the potential to avoid the much higher costs that could 
occur if manufacturers need to add or change technology at times other 
than their scheduled vehicle redesigns. This time period also provides 
manufacturers the opportunity to plan for compliance using a multi-year 
time frame, again consistent with normal business practice.
---------------------------------------------------------------------------

    \900\ 77 FR 62712.
---------------------------------------------------------------------------

    Section II.G.3.a of the NPRM provided substantial discussion of how 
an ``application schedule'' is used by the CAFE model to determine when 
manufacturers are assumed to be able to apply a given technology to a 
vehicle. The NPRM application schedule for engine technologies is 
reproduced in Table VI-43, which shows that all of the

[[Page 24420]]

engine technologies may only be applied (for the first time) during 
redesign.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.154

    For this final rulemaking action, a similar schedule is employed, 
and has been updated with information gathered since the NPRM and 
through comments provided to the agencies.
    Table VI-44 presents the engine technology application schedule 
used for the final rule CAFE modeling.

[[Page 24421]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.155

    Fuel economy improving and CO2 reducing technologies for 
vehicle applications vary widely in function, cost, effectiveness, and 
availability. Some of these attributes, like cost and availability, 
vary from year to year. New technologies often take several years to 
become available across the entire market. The agencies use phase-in 
caps to manage the maximum rate that the CAFE model can apply new 
technologies. Phase-in caps are intended to function as a proxy for a 
number of real-world limitations in deploying new technologies in the 
auto industry. These limitations can include but are not limited to, 
engineering resources at the OEM or supplier level, restrictions on 
intellectual property that limit deployment, and/or limitations in 
material or component supply as a market for a new technology develops. 
Without phase-in caps, the model may apply technologies at rates that 
are not representative of what the industry is actually capable of 
producing, which would suggest that more stringent standards might be 
feasible than actually would be. Table VI-45 and Table VI-46 below 
shows the phase-in caps between the NPRM and this final rule analysis, 
respectively.
    Most engine technologies are available at a rate of 100 percent in 
MY2017 for the final rule analysis. Some advanced technologies that 
have been recently introduced for one or two vehicle models are phased 
in at lower rates. Technologies such as ADEAC and TURBOD are phase in 
at rates that represent manufacturers' adoption capability and 
typically have complementary effectiveness compared to other advanced 
technologies. These lower phase-in caps also represent intellectual 
property and functional performance concerns.
BILLING CODE 4910-59-P

[[Page 24422]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.156


[[Page 24423]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.157

BILLING CODE 4910-59-C
    Comments received on engine adoption features were mixed, with 
manufacturers generally supporting the NPRM methodology, and CARB and 
NGOs opposing it. Several manufacturers commented, both in their public 
comments or on a CBI basis, that many of the emerging engine 
technologies had the potential to improve vehicle fuel economy, but 
were technically complex and addressed many of the same issues as other 
existing engine technologies.
    We agree with manufacturers that broadly, there are technologies 
that, in theory, present large potential effectiveness improvements 
like VCR, ADEAC, and others. However, the agencies believe it is 
important to assure realistic adoption of these technologies into the 
fleet in the rulemaking time frame, so that the rulemaking analysis 
accurately represents the costs and benefits of different regulatory 
alternatives considered. If the agencies were to select stringency 
based on an assumption that an emerging technology would see widespread 
adoption, and then it does not, the benefits of that stringency level 
would not be realized. The agencies have taken steps in the NPRM and 
this final rule analysis to consider the manufacturability and 
feasibility of these technologies for different vehicle types and 
manufacturers. Discussed earlier, the analysis considers these and 
other concerns by accounting for product cadence, and by implementing 
phase-in caps and skips, and by designating technology phase-in and 
phase-out years. Similar to the 2012 final rule, this final rule 
analysis employed these strategies to reflect better the real-world 
considerations faced by manufacturers.
    EDF commented, referencing EPA's statutory command prescribed in 
Section 202(a) of the Clean Air Act that:

    EPA's task is thus to identify the major steps necessary for 
`development and application of the requisite technology,' and then 
the respective standard `shall take effect.' These individual 
decisions are highly consequential: As noted above, without changing 
anything else about the agencies' analysis, allowing HCR2 would 
reduce augural compliance costs by $619--or about 30% of the total 
difference between the augural and rollback scenarios. The 
proposal's rejection of these technologies nowhere justifies how the 
(unfounded and cursorily justified) concerns accord with the 
agency's limited discretion under Section 202(a)(2) and duty to 
`press for the development and application of improved technology 
rather than be limited by that which exists today.' If the agency is 
to predict more than the results of merely assembling pre-existing 
components, it must have some leeway to deduce results that are not 
represented by present data.\901\
---------------------------------------------------------------------------

    \901\ NHTSA-2018-0067-12108 at 104.

    CARB also commented that the CAFE Model prevents manufacturers 
``from

[[Page 24424]]

switching between a turbocharged and HCR pathways under the premise 
that manufacturers either would not develop both or would be committed 
irreversibly to one path or the other. This assumption is not based in 
reality and is not reflective of actual industry practice--
manufacturers who have pursued turbocharging have also already pursued 
HCR engines for other vehicles in their line-up. For example, General 
Motors (GM) utilizes downsized turbocharging in some vehicles, such as 
the newly designed 2019MY Silverado pick-up and the Malibu sedan which 
has two different turbocharged engine options. GM also has a third 
offering in the Malibu sedan which is an HCR naturally aspirated 1.8L 
equipped with cooled exhaust gas recirculation (CEGR) mated to a hybrid 
electric system.'' \902\
---------------------------------------------------------------------------

    \902\ NHTSA-2018-0067-11873 at 109.
---------------------------------------------------------------------------

    CARB's observation was true for the NPRM analysis, however for the 
final rule analysis the agencies allowed manufacturers to adopt engine 
technologies from alternate tree paths, when incorporating 
electrification technology, see Section VI.C.3.c). The agencies still 
believe that if manufacturers have invested in one type of engine 
technology for their vehicles that they would not transition to another 
technology except in the case of a major vehicle powertrain redesign, 
such as the inclusion of an HEV system. Additional discussion on this 
issue is presented in Section VI.B.1.
    The following sections discuss adoption features specific to 
individual engine technologies, including comments received and updates 
(or not) for the final rule analysis.
(1) Basic Engines
    Most vehicles in the MY 2017 analysis fleet that are DOHC or SOHC/
OHV spark ignited engines and are not downsized turbocharged engines 
have any two combinations of VVT, VVL, SGDI or DEAC.\903\ For the NPRM, 
only engines with 6-cylinders or more could adopt DEAC and ADEAC.
---------------------------------------------------------------------------

    \903\ EPA. ``2018 EPA Automotive Trends Report'' 12 pp, 421 K, 
EPA-420-S-19-001, March 2019. https://www.epa.gov/automotive-trends/download-automotive-trends-report#Full%20Report (last accessed Feb. 
12, 2020) p. 72.
---------------------------------------------------------------------------

    HDS on behalf of CARB commented that in the NPRM analysis VVL, 
which is cost ineffective compared to other conventional technologies, 
was always included in an adopted technology package.\904\ HDS further 
stated that the ``effectiveness of VVL is even smaller when the 
technology is combined with turbocharged downsized engines.'' 
Accordingly, HDS stated that removing VVL from the base pathway would 
save $314 but reduce fuel economy by only 1.4 percent, according to the 
LPM.
---------------------------------------------------------------------------

    \904\ NHTSA-2018-0067-11985 at p.34.
---------------------------------------------------------------------------

    The agencies did not agree with HDS' assessment of the NPRM 
analysis. The agencies do not agree VVL was forced to be adopted in the 
analysis fleet and do not agree with how technology effectiveness 
values compare to LPM estimates. As discussed earlier in the 
effectiveness and modeling section, each engine technology was modeled 
independently and the CAFE model was allowed to adopt the most cost 
effective technology. Therefore, it is inaccurate to state, a 
technology is less effective, especially when comparing LPM. 
Particularly because VVL technologies reduce pumping losses in engines, 
so it is realistic that other technologies, that also reduce pumping 
losses, have synergetic effect. This is specifically true for 
turbocharged engines.
    ICCT commented that DEAC technology should be available for every 
engine, and should not be limited to 6-cylinder and higher cylinder 
count engines. ICCT and CARB also commented that DEAC should be allowed 
on turbocharged engines. ICCT also commented that ADEAC should be 
widely available as it can be a viable technology application for 
various other powertrain technology combinations.\905\ Furthermore, 
CARB commented ``automakers will combine technologies like 
turbocharging, HCR and DEAC as well as more technologies when they have 
cost-effectiveness synergies.'' \906\
---------------------------------------------------------------------------

    \905\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-13.
    \906\ CARB at p. 6.
---------------------------------------------------------------------------

    The agencies agree with ICCT that DEAC and ADEAC could be applied 
to additional engine types, including turbocharged engines. However, 
the agencies disagree with ICCT that ADEAC should be widely applied to 
all powertrain technology combinations in this analysis. The agencies 
have updated the final rule analysis to allow DEAC and ADEAC for 
various engine cylinder counts and for turbocharged engines.
    For the final rule analysis, both DEAC and ADEAC technologies can 
be adopted by any naturally aspirated engine. Similarly, any 
turbocharged engine can also adopt cylinder deactivation technology, as 
characterized by TURBOD and TURBOAD in the CAFE model. In this final 
rule analysis, the agencies distinguished cylinder deactivation 
technologies between naturally aspirated and forced air induction 
systems.
    For the final rule analysis, the agencies allow any combination of 
VVT, VVL, SGDI and DEAC to be adopted for any engine displacement and 
cylinder count. Figure VI-18 below shows the basic engine paths a 
vehicle could traverse for the final rule analysis. Similar to the 
NPRM, the agencies have not changed the adoption features of the 
technologies shown in Figure VI-18, with one exception. Vehicles that 
are SOHC or DOHC configuration that do not have VVT in the baseline can 
now adopt it.
    Finally, the agencies disagree with ICCT and CARB that these DEAC, 
ADEAC, TURBOD, and TURBOAD should apply beyond these configurations. 
DEAC's fundamental benefits are driven by reducing pumping losses and 
by enabling the engine to operate in a more thermal efficient region of 
the engine fuel map. Conventional spark-ignited engines control airflow 
into the cylinders via a throttle operated by the driver to provide the 
level of power that is delivered.\907\ In an 8-cylinder engine, when 
driving in light load conditions such as highway driving, there are 
lower engine power requirements. In a throttle controlled system, 
engine pumping losses increase as air flow decreases. A way to reduce 
pumping loss in an engine is by increasing the airflow into the 
cylinders. By deactivating a set of cylinders, the same power output 
can be delivered by a ``smaller'' engine. Many technologies modeled for 
this analysis work to reduce pumping losses, but through other 
mechanisms like VVT, VVL, downsized engines with turbochargers, high 
compression Atkinson mode cycle, and Miller Cycle.\908\ Transmissions 
with a higher number of gears also provide the opportunity to reduce 
pumping work of the engine.\909\
---------------------------------------------------------------------------

    \907\ A throttle is the mechanism by which fluid flow is managed 
by constriction or obstruction. An engine's power can be increased 
or decreased by the restriction of inlet gases, but usually 
decreased.
    \908\ 2015 NAS at p. 23.
    \909\ 2015 NAS at p.173.
---------------------------------------------------------------------------

    As discussed earlier, DEAC can reduce pumping losses, so when 
combined with other technologies that also reduce pumping losses, like 
downsized turbocharged engines, the benefits for cylinder deactivation 
are lower than for naturally aspirated engines because downsized 
turbocharged engines already have lower pumping losses due to having a 
downsized engine.\910\
---------------------------------------------------------------------------

    \910\ 2015 NAS at p. 34.

---------------------------------------------------------------------------

[[Page 24425]]

(2) Turbocharged Downsized Engines
    About 23 percent of vehicles in the MY 2017 baseline fleet had 
turbocharged engines. For the final rule analysis, the agencies allowed 
any basic engine to adopt turbo engine technology (TURBO1, TURBO2 and 
CEGR1) from the Turbo path similar to the NPRM analysis. This includes 
any combination of VVT, VVL, SGDI and DEAC for both SOHC and DOHC 
configurations. Vehicles that have turbocharged engines in the baseline 
fleet will stay on the turbo engine path to prevent unrealistic engine 
technology change in a short timeframe considered in the rulemaking 
analysis. Turbo path is a mutually exclusive technology in that it 
cannot be adopted for HCR, diesel, ADEAC, CNG and powersplit PHEVs.
(3) Non-HEV Atkinson Mode Engines
    The NPRM analysis allowed limited application of HCR engines (HCR1 
and HCR2) to vehicles in the MY 2016 baseline fleet.\911\ As discussed 
above, applying HCR1 or HCR2 technologies to a vehicle resulted in 
overstated effectiveness values relative to the baseline VVT 
engine,\912\ because of differences in how those maps were developed 
compared to the IAV engine maps used for the majority of the technology 
analysis. In an attempt to avoid unrealistic results in the NPRM, 
adoption of HCR1 (Eng24) technology was limited to only manufacturers 
that demonstrated existing use of high compression ratio technology. 
HCR was disallowed for other manufacturers that demonstrated an intent 
to develop other advanced technologies incompatible with HCR 
technology. In addition, the agencies disallowed HCR engines from being 
applied to vehicles with greater performance requirements, like 6- and 
8-cylinder vehicles, because the higher load requirements from these 
vehicles would force the engine to exit the Atkinson mode, where 
maximum efficiency is achieved.
---------------------------------------------------------------------------

    \911\ 83 FR 43037.
    \912\ 83 FR 43029 Figure II-1--Simulated Technology 
Effectiveness Value.
---------------------------------------------------------------------------

    The Alliance commented in agreement with the application 
restrictions for HCR1 in the NPRM, listing the following 
justifications: ``Packaging and emission constraints associated with 
intricate exhaust manifolds needed to mitigate high load/low 
revolutions per minute knock; Inherent performance limitations of 
Atkinson cycle engines; and Extensive capital and resources required 
for manufacturers to shift to HCR from other established technology 
pathways (e.g., downsized turbocharging).'' \913\ Ford similarly 
commented in support of ``the more restrained application of HCR1 in 
the Proposed Rule, an approach that recognizes the investment, 
packaging, performance and emissions factors that will limit 
penetration of this technology.'' \914\
---------------------------------------------------------------------------

    \913\ NHTSA-2018-0067-12073.
    \914\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    In contrast, CARB stated that the constraint on HCR1 engines was 
inappropriate and did not reflect reality,\915\ and stated that the 
agencies failed to supply any detailed rationale as to why HCR 
applications were so constrained in the CAFE Model. Specifically, CARB 
took issue with the justification that HCR1 is limited in the CAFE 
model because it is ``not suitable for MY 2016 baseline vehicle models 
that have 8-cylinder engines and in many cases 6-cylinder engines.'' 
\916\ CARB stated that ``the HCR1 technology is declared not suitable 
on 207 of the 288 engines cumulatively used by all of industry 
including over 50 percent of the 4 cylinder engines and nearly 90 
percent of the 6 cylinder engines instead of only being restricted from 
8 cylinder and `in many cases 6 cylinder engines.' '' CARB also stated 
that the implied rationale for not allowing HCR1 to be applied to 6- 
and 8-cylinder engines because trucks or larger vehicles could not 
utilize it is unreasonable, as the Toyota Tacoma used a 3.5L V6 HCR 
Atkinson-like engine since MY 2016. CARB stated that the Toyota Tacoma 
was properly assigned a HCR1 engine in the MY 2016 analysis fleet file, 
but the engine was disallowed from other Toyota V6 engines utilized in 
vehicles like the Sienna minivan and 4Runner SUV. CARB commented that 
``[i]f the intended rationale is that HCR engines will have 
insufficient low end torque to satisfy truck-like towing demands, it 
would be inappropriate to restrict the engine from minivan and SUV 
applications which have a lower tow rating and lower expected towing 
demands.'' Finally, CARB stated that the HCR1 package restrictions were 
inappropriate, as there was no mechanism in the CAFE model to represent 
appropriately the MY 2019 Dodge Ram 1500 5.7L V8 that uses ``a higher 
compression ratio than earlier versions and using its VVT system to 
reduce pumping losses via delayed, or late, intake valve closing--
resulting in an HCR-like engine with an over-expanded or Atkinson 
cycle.''
---------------------------------------------------------------------------

    \915\ NHTSA-2018-0067-11873.
    \916\ 83 FR 43038.
---------------------------------------------------------------------------

    Similarly, Meszler Engineering Services, commenting on behalf of 
NRDC, commented that HCR1 appears as a baseline technology on vehicles 
representing about 4 percent of the baseline non-hybrid vehicle market, 
and is subsequently applied to only 23 percent of the market. Meszler 
stated that the ``relative cost effectiveness of the technology is 
perhaps best illustrated by the fact that the market penetration of HCR 
technology on non-hybrid vehicles under the augural standard is modeled 
to be 27 percent of 2032 sales, exactly equal to the baseline 
penetration of 4 percent and the allowable adoption fraction of 23 
percent. In other words, the technology was adopted by every vehicle 
that was not explicitly prohibited (by NHTSA) from doing so.'' EDF 
commented that ``NHTSA has further imposed artificial and unreasonable 
constrains on the use of certain technologies that does not match how 
automakers are applying them in vehicles today,'' stating that HCR1 
represented a technology that had been in the marketplace for many 
years and had been applied by several manufacturers, ``[y]et, even for 
MY 2030 vehicles and beyond, NHTSA only allows the use of HCR1 by about 
30 percent of the U.S. fleet.'' \917\
---------------------------------------------------------------------------

    \917\ NHTSA-2018-0067-12108.
---------------------------------------------------------------------------

    In considering the comments, the agencies agree with commenters 
that the HCR1 engine application was overly limited for the NPRM 
analysis. As a result, the agencies have expanded the availability of 
HCR1 technology for the final rule analysis. The refined adoption 
features for HCR1 are discussed below. The new adoption features do 
maintain considerations for performance neutrality. Comments about how 
the characterization of engine technologies in the analysis fleet 
impacted HCR technology adoption in subsequent model years are 
addressed in Section VI.C.1.d) Baseline Fleet Engine Tech.
    Regarding HCR2, the Alliance commented in support of ``the decision 
to exclude the speculative HCR2 technology from the analysis.'' \918\ 
The Alliance continued, ``[a]s previously documented in Alliance 
comments, the inexplicably high benefits ascribed to this theoretical 
combination of technologies has not been validated by physical 
testing.'' Similarly, Ford stated that ``[t]he effectiveness of the 
`futured' Atkinson package (HCR2) that includes cooled exhaust gas 
recirculation (CEGR) and cylinder deactivation (DEAC) is excessively 
high, primarily due to overly-optimistic efficiencies in the base 
engine map, insufficient accounting of CEGR and DEAC integration 
losses, and no accounting of the impact of 91RON

[[Page 24426]]

Tier 3 test fuel. Given the speculative and optimistic modeling of this 
technology combination, Ford supports limiting the use of HCR2 
technology to reference only, as described in the Proposed Rule.'' 
\919\
---------------------------------------------------------------------------

    \918\ NHTSA-2018-0067-12073.
    \919\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    In contrast, several commenters disagreed with the agencies' 
decision to limit the adoption of HCR2 engines, stating that the 
technology was clearly applicable during the rulemaking timeframe, as 
the technology was already being applied by manufacturers, and that the 
technology was cost-effective, as shown by the agencies' own modeling.
    ICCT commented that ``[i]t is clear that the agencies have 
artificially excluded a known technology that is applicable in the 
timeframe of the rulemaking.'' \920\ ICCT commented that ``[d]espite 
the facts that (as discussed above) the agencies have cost and 
effectiveness data for this technology, many automakers are already 
deploying the HCR1 technology, and the 2018 Camry has already put most 
of the HCR2 technologies into production, the agencies did not allow 
any application of HCR2 by 2025.'' \921\ ICCT concluded that the ``only 
explanations . . . for the agencies' system of omissions and 
constraints are that the agencies have biased the analysis against 
including all the viable technologies by inserting their own artificial 
constraints (either for lack of research, lack of analytical effort, or 
not fully utilizing all the agencies' best analytical tools and data) 
or that the auto industry is providing information that erroneously 
suggests their innovation is far less than what is demonstrated both 
above and in the agencies' own previous analyses.'' ICCT stated that 
``[t]he great lengths the agencies have gone to artificially impose 
`skip' constraints for HCR in the CAFE modeling system demonstrates 
that the agencies have exerted an explicable and apparently deliberate 
bias towards forcing most of the automaker compliance technology toward 
higher cost, non-HCR turbocharging paths.'' \922\
---------------------------------------------------------------------------

    \920\ NHTSA-2018-0067-11741.
    \921\ NHTSA-2018-0067-11741.
    \922\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    Several commenters also stated that HCR should not have been 
restricted because it is clearly a cost-effective technology, citing 
the sensitivity runs conducted that allowed unrestricted HCR 
application in the analysis. For example, ICCT commented that allowing 
HCR2 application across the fleet reduced total per-vehicle cost of 
compliance with the augural standards by $690, which ``shows that the 
agencies intentionally excluded a highly cost-effective technology (by 
their own analysis) in the rulemaking analysis.'' \923\ Similarly, EDF 
performed software modifications of the CAFE model, including allowing 
the use of both HCR1 and HCR2 technology for all manufacturers by MY 
2028. The analysis performed by EDF using their modified version of the 
CAFE model, showed reductions in the per-vehicle compliance cost 
projections by nearly $600.\924\
---------------------------------------------------------------------------

    \923\ NHTSA-2018-0067-11741.
    \924\ NHTSA-2018-0067-12108.
---------------------------------------------------------------------------

    ICCT concluded that ``[t]he only reasonable and technically valid 
assumption is that HCR be allowed for application to all vehicle 
models' engine redesigns through all the model years of the compliance 
modeling analysis.'' \925\ ICCT stated that ``[f]or the agencies to 
constrain HCR technology for use by other automakers, they have a 
responsibility to demonstrate why each of the other automakers cannot 
adopt this known technology in their fleet.''
---------------------------------------------------------------------------

    \925\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    The agencies agree with commenters' observations about the results 
of the sensitivity runs performed as part of the NPRM analysis. 
However, the agencies also believe the adoption features for HCR1 and 
HCR2 were appropriate for the NPRM analysis. Had the agencies not 
applied adoption features in that way, the agencies would have shown 
unrealistic pathways for compliance for manufacturers that would have 
understated costs and overstated benefits of potential CAFE and 
CO2 standards.
    The agencies disagree with commenters' statements that HCR has been 
widely available in the automotive market and that the HCR technology 
accordingly should not be limited in the CAFE model. For reasons 
discussed in the NPRM and explained in more detail in Section 
VI.C.1.c)(3), depending on vehicle type and use, Atkinson cycle 
operation may be enabled for low and moderate engine demand conditions, 
whereas Otto cycle operation may be needed for higher load conditions 
to meet performance needs, such as to move more passengers, cargo, or 
for towing. In addition, there may be issues on some platforms to 
package the larger exhaust manifolds needed to enable Atkinson 
operation, particularly with V6 and V8 engines. Manufacturers have 
applied Atkinson technologies in unique ways to meet the needs and 
capabilities of their vehicles to operate using the Atkinson and Otto 
cycles. The agencies agree with comments from stakeholders, including 
Toyota, who observed HCR technology is not suitable for all vehicle 
configurations, and may not meet performance requirements for high-load 
applications. As discussed earlier, the agencies believe the variation 
of technologies can be categorized into three different forms of 
Atkinson engine technologies for this analysis: (1) Atkinson engines, 
(2) Atkinson-mode engines, and (3) Atkinson-enabled engines using 
variable valve timing with late intake closing. Manufacturers typically 
apply one of these technologies and tune that technology for specific 
applications. Some commenters have consistently conflated the 
technologies and asserted the capabilities of all three types of 
Atkinson technologies can be represented by a single engine model. The 
agencies do not agree with stakeholder assertions that a single HCR 
engine map should be applied to every technology class or vehicle 
platform.
    To reflect better the incremental effectiveness for a low-cost 
version of HCR technology, the agencies added the HCR0 engine for the 
analysis. The specification of this engine was provided in the NPRM 
PRIA as Eng22b. Using this engine improves the estimated incremental 
effectiveness because the incremental engine changes were directly 
specified for the modeling and are relative to the other engine 
technologies in the analysis.\926\ HCR0 is the first engine in the HCR 
path that a manufacturer could adopt. HCR0 represents technology that 
could incrementally be adopted to the VVT engine, increasing 
compression ratio and adding Atkinson cycle capability. The use of the 
HCR0 technology, applied in the final rule analysis, allowed the 
agencies to update HCR adoption features. Once a basic engine adopts 
HCR technology (i.e., HCR0 and HCR1 for the central analysis, or HCR2 
for a sensitivity case) the vehicle will not switch to a different 
engine technology path. For example, if a vehicle had adopted HCR or is 
equipped with HCR technology it is not allowed to adopt turbocharged 
engine technologies. The HCR0 technology appropriately captures the 
benefits of applying transitional Atkinson technologies to conventional 
basic engine technologies. The agencies note that VVT technology valve 
control has late intake valve closing under some operating conditions 
to take some advantage of Atkinson cycle-like operation; however, that 
operation is not as extensive as HCR technology and is not coupled with 
a higher

[[Page 24427]]

compression ratio as is the case for HCR technologies.
---------------------------------------------------------------------------

    \926\ PRIA 6.3.2.2.21.20.2.1 IAV Engine 22b--High Compression 
Atkinson Cycle Engine at p. 307.
---------------------------------------------------------------------------

    The agencies also allowed all 4-cylinder engines on the basic 
engine path to adopt HCR technology similar to turbocharged 
technologies. This allowed any small and midsize vehicles, including 
small and midsize SUVs, that had any combinations of basic engine path 
technologies to move to the HCR path. However, there are two exceptions 
to this feature, including: (1) When the vehicle is a pickup including 
both standard and performance class; and (2) when the base engine is 
shared with a pickup including both standard and performance class. The 
agencies discussed earlier in the non-HEV Atkinson section why HCR 
technology cannot be applied to all vehicle applications.
    Finally, engines with advanced engine technology already in the 
baseline vehicle such as turbocharged engines are not allowed to adopt 
HCR technology. The agencies continue to believe this constraint is 
reasonable given the extensive capital resources and stranded capital 
that would be involved if a manufacturer who focused on and invested 
heavily in non-HCR advanced technologies were to abandon those 
technologies abruptly and switch to HCR technologies.\927\ For example, 
Ford has incorporated turbocharged engines across 75 percent to 80 
percent of their fleet in MY2017, and these engines are shared across 
multiple technology classes.\928\ The abovementioned modeling, 
limitation for this analysis assumes that manufacturers will not change 
advanced engine technology applied to a platform due to the high cost 
and lead time required for research and development, and for the 
development and implementation of new manufacturing plants and 
equipment to implement an entirely new powertrain in the rule making 
time frame. For further discussion see Section VI.B.1.
---------------------------------------------------------------------------

    \927\ 83 FR 43038.
    \928\ The 2018 EPA Automotive Trends Report figure 4.23. at 
p.68.
---------------------------------------------------------------------------

    In response to ICCT's comment that agencies must discuss the 
reasoning for allowing and disallowing HCR technology for each 
individual manufacturer, these updated adoption features now allow more 
manufacturers to adopt HCR engine technology. The agencies no longer 
apply adoption features based on manufacturer, but now base them on 
individual platforms. The agencies believe a manufacturer that has 
already invested in advanced engine technologies for a specific 
platform would face very high costs and incur significant stranded 
capital to switch that platform to another advanced technology. And 
doing so would not be reasonable given the small incremental fuel 
economy improvement that would be gained, for example, for switching 
from advanced turbocharging to HCR technologies. Specifically, 
manufacturers that have invested in turbocharging technology for 
certain platforms, like Honda, Ford, and the German manufacturers, 
would incur unreasonable costs to switch to another advanced technology 
path. However, manufacturers that use turbo technology on one platform 
are not precluded from implementing HCR technology on another of its 
platforms. HCR adoption is still limited for all manufacturers based on 
vehicle performance requirements discussed earlier.
(4) Advanced Cylinder Deactivation Technology
    In the NPRM, any basic engine technology could adopt ADEAC. 
Commenters stated that the agencies restricted ADEAC technologies in 
the NPRM analysis to naturally aspirated engines.
    ICCT provided a broad comment regarding the treatment of advanced 
technologies, including ADEAC, and criticized how the NPRM ``removed 
many technologies that are viable and being actively deployed by the 
auto industry.'' ICCT specifically criticized ``cases where viable 
technology combinations are disallowed'' such as ``turbocharging and 
cylinder deactivation (DEAC).'' \929\
---------------------------------------------------------------------------

    \929\ NHTSA-2018-0067-11741 at p.6.
---------------------------------------------------------------------------

    UCS also commented on how ADEAC technology was applied in the NPRM, 
stating ``While the agencies have acknowledged the existence of dynamic 
cylinder deactivation, they have not appropriately included it as an 
available technology, dramatically limiting its availability.'' UCS 
specifically disagreed with adoption features of the ADEC, noting the 
technology ``is restricted to naturally aspirated, low-compression 
ratio engines--it cannot be combined with turbocharged engines, high 
compression ratio engines, or variable compression ratio engines due to 
pathway exclusivity in the Volpe model.'' \930\ CARB and Meszler 
mirrored these concerns.\931\
---------------------------------------------------------------------------

    \930\ NHTSA-2018-0067-12039 at p.4
    \931\ NHTSA-2018-0067-12039 at p.4.
---------------------------------------------------------------------------

    The agencies agreed with commenters and in response have allowed 
both naturally aspirated engines and turbocharged engines to adopt 
ADEAC in the final rule analysis. The new Advanced Turbocharging path 
includes TURBOD and TURBOAD, while naturally aspirated engines use the 
same ADEAC engine designation. There is some potential for this type of 
technology to improve fuel economy and reduce CO2 emissions, 
however, the technology provides diminishing returns if it is included 
with engine downsizing or other technologies that already reduce 
pumping losses. Accordingly, once a vehicle has adopted ADEAC, TURBOD, 
or TURBOAD, the agencies did not allow further adoption of other engine 
technologies that reduce pumping losses such as VCR and VTG.
(5) Miller Cycle Engines
    Miller cycle engine technologies (VTG and VTGe) are new for this 
final rule analysis, and VTG engines could be applied to any basic and 
turbocharged engine. Discussed earlier, the VTGe technology is enabled 
by the use of a 48V system that presents an improvement from 
traditional turbocharged engines, and accordingly VTGe could only be 
applied with a mild hybrid system.
(6) Variable Compression Ratio Engines
    In the NPRM analysis, variable compression ratio (VCR) technology 
was not available for adoption, but the engine map and specifications 
were provided for review. For this final rule analysis, VCR engines are 
included in the analysis and can be applied to basic and turbocharged 
engines, however the technology is limited to Nissan. VCR technology 
requires a complete redesign of the engine, and in MY2020, only two of 
Nissan's models had incorporated this technology. In addition, the 
technology showed lower fuel savings than expected.\932\ The agencies 
do not believe any other manufacturers will invest to develop and 
market this technology in their fleet in the rulemaking time frame.
---------------------------------------------------------------------------

    \932\ VanderWerp, D. ``Why Nissan's Holy-Grail VC-T Engine 
Doesn't Achieve Better Fuel Economy,'' C/D Nov 1, 2018. Available at 
https://www.caranddriver.com/features/a24434937/nissan-new-vc-t-engine-fuel-economy/. Last accessed Dec. 19, 2019.
---------------------------------------------------------------------------

(7) Diesel Engines
    Diesel engine adoption and features have been carried from the NPRM 
analysis for this final rule analysis for ADSL and DSLI. Any basic 
engine technologies (VVT, VVL, SGDI, and DEAC) can adopt ADSL and DSLI 
engine technologies. New for the final rule analysis is the adoption of 
advanced cylinder deactivation for diesel engines (DSLIAD). Any basic 
engine and diesel engine can adopt this technology in the final rule 
analysis;

[[Page 24428]]

however, the agencies have applied a phase in cap and year for this 
technology at 34 percent and MY 2023, respectively. In the agencies' 
engineering judgement, the agencies have concluded that this is a 
rather complex and costly technology to adopt and think that it could 
take significant investment to develop. For more than a decade, diesel 
engine technologies have been used in less than one percent of the 
total light-duty fleet production,\933\ and the investment for this 
cylinder deactivation technologies may not be justifiable.
---------------------------------------------------------------------------

    \933\ The 2018 EPA Automotive Trends Report Table 4.1 at p. 72.
---------------------------------------------------------------------------

(8) Alternative Fuel Engines
    Adoption features for alternative fueled compressed natural gas 
(CNG) engines have been carried over from the NPRM for this final rule 
analysis. Because CNG is considered an alternative fuel under EPCA/
EISA, it cannot be adopted during the rulemaking timeframe for NHTSA's 
standard setting analysis. The EPA analysis was modeled separately in 
the CAFE model without such constraints.
(9) Engine Lubrication and Friction Reduction
    Finally, new for this analysis is the addition of EFR. The agencies 
allow EFR to apply to any engine technology except for DSLI and DSLIAD. 
DSLI and DSLIAD inherently have incorporated engine friction 
technologies from ADSL. In addition, friction reduction technologies 
that apply to gasoline engines cannot necessarily be applied to diesel 
engines due to the higher temperature and pressure operation in diesel 
engines.
f) Engine Effectiveness Modeling and Effectiveness Values
    Figure VI-20 below shows the effectivness estimates from all the 
vehicle types for the NPRM analysis using Autonomie full vehicle 
modeling and simulation.
[GRAPHIC] [TIFF OMITTED] TR30AP20.158

    Roush commented that they had observed wide variations in estimated 
incremental effectiveness associated with individual technology 
packages between the 2016 Draft TAR and NPRM analysis.\934\
---------------------------------------------------------------------------

    \934\ NHTSA-2018-0067-11984. Roush at p. 16.
---------------------------------------------------------------------------

    The agencies agree that to predict potential incremental 
improvements in fuel efficiency accurately, it is extremely important 
to understand the nature of the improvements being sought by each 
increment (improved thermodynamics, reduced friction, reduced vehicle 
weight, etc.). The technology modeling

[[Page 24429]]

and large scale simulation used for the proposal and updated for the 
final rule does exactly that. In fact, the NPRM and final rule use 
these methods more expansively than any previous CAFE and 
CO2 rulemaking, including the 2016 Draft TAR and 2016 EPA 
Proposed Determination.
    One commenter stated the effectiveness for ADEAC was overestimated 
for the NPRM, and that data from compliance shows much lower 
effectiveness. The agencies disagree with this comment, as it is 
invalid to compare effectiveness of full vehicle compliance data 
directly to the incremental effectiveness modeled for ADEAC. For 
reasons discussed in Section VI.B.3 data from full vehicle benchmarking 
cannot be used as a comparison for specific technology effectiveness. 
The effectiveness estimated for this technology is in line with test 
data, CBI, and engineering analysis.\935\
---------------------------------------------------------------------------

    \935\ Boha, Stani. ``Benchmarking and Characterization of a Full 
Continuous Cylinder Deactivation System.'' EPA. April 10-12, 2018 
SAEA World Congress. https://www.epa.gov/sites/production/files/2018-10/documents/deact-sae-world-congress-bohac-2018-04.pdf last 
access Feb 12, 2020.
---------------------------------------------------------------------------

    Engine effectiveness estimates remained the same for most 
technologies from the NPRM analysis, with the exception of some 
technologies that had characteristics updated, and the new added engine 
technologies. For the final rule analysis, the agencies used the same 
effectiveness values for ADEAC applied to naturally aspirated engines 
as in the NPRM, and incorporated estimated effectiveness values for 
TURBOAD to represent ADEAC on downsized turbocharged engines.
    Other technology-specific comments and the agencies' responses are 
provided within the discussion of each technology throughout this 
section, as those comments tended to be predicated on issues 
surrounding the engine maps used to model technologies or technology-
specific adoption features. For the final rule analysis, the technical 
merits of the substantive comments and any accompanying publications 
and information were carefully considered and discussed in the 
subsections where appropriate.
    Figure VI-21 below shows the effectivness estimates from compact 
car and midsize car vehicle types for the final rule analysis using 
Autonomie full vehicle modeling and simulation.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.159

g) Engine Costs
    Discussed in the PRIA, the agencies spent millions of dollars 
sponsoring research to determine direct manufacturing costs (DMCs) for 
fuel saving technologies since the 2012 rule.\936\ Because a major 
objective of the studies was to consider costs in the rulemaking 
timeframe, the agencies believed that these costs were appropriate to 
use for the NPRM and final rule analysis. Table VI-47 below shows the 
DMC used for IC engine technologies for the NPRM analysis.
---------------------------------------------------------------------------

    \936\ FEV prepared several cost analysis studies for EPA on 
subjects ranging from advanced 8-speed transmissions to belt 
alternator starter, or Start/Stop systems. NHTSA also contracted 
with Electricore, EDAG, and Southwest Research on teardown studies 
evaluating mass reduction and transmissions. The 2015 NAS report on 
fuel economy technologies for light-duty vehicles also evaluated the 
agencies' technology costs developed based on these teardown 
studies, and the technology costs used in this proposal were updated 
accordingly. These studies are discussed in detail in Chapter 6 of 
the RIA accompanying the NPRM proposal.

---------------------------------------------------------------------------

[[Page 24430]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.160


[[Page 24431]]


    CARB commented that costs associated with IC engines were not 
excluded from the final costs of BEV vehicles.\937\ CARB continued, 
stating that ``the final costs of BEV vehicles are higher due to the 
inclusion of the base absolute costs, to which the assigned BEV 
incremental cost would be added.''
---------------------------------------------------------------------------

    \937\ NHTSA-2018-0067-11873 at p.122.
---------------------------------------------------------------------------

    The agencies agree with CARB that inclusion of IC engine costs in 
the BEV cost was an error in the analysis. In response to this comment, 
the agencies have developed absolute costs for baseline engines for the 
CAFE model in order to account for appropriate cost of removing engines 
from BEVs. In the final rule analysis, once a vehicle adopts BEV 
technology, the costs associated with powertrain systems are removed. 
Due to the extensive variations in engine technologies in real world 
production, the agencies relied on discrete publication costs and 
historical studies to assign costs for base engines.938 939 
For this final rule analysis, the agencies have included these costs 
for base engines shown in Table VI-48.
---------------------------------------------------------------------------

    \938\ FEV P311732-02 Oct13, 2015 at p. 259.
    \939\ UBS Limited. ``UBS Evidence Lab Electric Car Teardown--
Disruption ahead?'' May 18, 2017.
[GRAPHIC] [TIFF OMITTED] TR30AP20.161

BILLING CODE 4910-59-C
    Commenters compared engine cost data from the NPRM to other 
sources, in many cases to support their comments that the technology 
costs used in the NPRM were too high. ICCT commented that the agencies 
did not consider the latest reports on technology cost data, and 
specifically referenced an ICCT-sponsored FEV cost study for the 
European EU6b regulations in MY 2025,\940\ as well as prior EPA cost 
estimates for several engine technologies including SGDI, cEGR, HCR, 
and others, to point out differences in cost.\941\ ICCT also commented 
on the difficulty they had in locating the cost data used in the NPRM, 
stating that ``because the agencies present cost data in so many 
different ways in dozens of different places in the NPRM, impact 
assessment, and supporting data files, the precise agencies' costs are 
obscured and not transparent.'' ICCT stated that ``[w]ithout a clear 
explanation of the methodology, it is unclear precisely how price 
increases are determined, as well as the relationship between 
technology costs, fines, and price increases.'' Despite this claim, 
ICCT was able to provide several pages comparing engine technology 
costs.
---------------------------------------------------------------------------

    \940\ FEV. '' 2025 Passenger Car and Light Commercial Vehicle 
Powertrain Technology Analysis'' September 2015. https://theicct.org/sites/default/files/publications/PV-LCV-Powertrain-Tech-Analysis_FEV-ICCT_2015.pdf.
    \941\ NHTSA-2018-0067-11741 at p. I-68.
---------------------------------------------------------------------------

    In the NPRM PRIA Chapter 6.3.2.2.20.22, the agencies provided DMCs 
for all engine technologies in 2016 dollars without inclusion of RPE 
and learning for review. In the same chapter, the agencies also 
provided absolute costs that incorporated costs in 2016 dollars, RPE 
and learning data as used by the CAFE model to assess cost 
effectiveness for future MY vehicles. Where appropriate, the agencies 
discussed in the individual technology sections where costs were 
updated for this final rule analysis with the latest data. This also 
includes cost data for new technologies available in the CAFE model for 
the final rule analysis.
    Some engine costs were carried over from prior rulemakings, but may 
have looked different because they were updated to current dollars 
(2016 for the NPRM and 2018 for the final rule), and for engine 
architecture and cylinder count. In addition, costs were updated based 
on appropriate vehicle class. This was important to consider to 
maintain performance neutrality, as technology effectiveness associated 
with one engine technology type for a vehicle class cannot be used for 
the same engine technology for higher performance vehicle class. This 
affected total costs. For further discussion on the cost-effectiveness 
metric used in the CAFE model, see discussions in the Section VI.A 
Overview of the CAFE model and VI.B.3 Technology Effectiveness Values.
    The agencies do not believe that the FEV report referenced by ICCT 
is applicable for this analysis for a few reasons. First, the primary 
focus of the FEV study ``is the European Market according to the EU6b 
regulation as well as the consideration of emissions under both the 
NEDC and WLTP test procedures.'' This final rule analysis specifically 
considered the U.S. automotive market during the rulemaking timeframe 
based on U.S.-specific regulatory test cycles. Accordingly, the costs 
reflect incremental technology effectiveness for achieving improvements 
as measured through U.S. regulatory test methods. The agencies had 
discussed these test cycles and methods further in Section VI.B.3 
Technology Effectiveness Values.
    Second, FEV did not conduct original teardown studies for this 
report, as indicated by project tasks, but rather used engineering 
judgement and external studies in assessing incremental costs.\942\ The 
FEV report did not provide sources for each individual cost and it is 
unclear how costs in many scenarios were developed since no teardowns 
were used. Note that for this final rule analysis, the agencies have 
used previously conducted FEV cost teardown studies and the referenced 
2015 NAS costs that referenced FEV teardowns. The agencies are not 
concluding that FEV is an unreliable source. The agencies preferred to 
specifically identify incremental costs of adding technology to account 
appropriately for the costs of those technologies in the analysis.
---------------------------------------------------------------------------

    \942\ FEV EU Costs Tasks: ``Definition of reference hardware or 
description made by experience of development and design engineers 
as well as additional research as base for cost analysis (no 
purchase of hardware)''.
---------------------------------------------------------------------------

    Finally, the cost for different vehicle classes identified by the 
FEV study does not line up with the vehicle classes discussed in the 
NPRM and this final rule analysis. FEV stated specifically, ``the 
configuration of the vehicles has not been optimized for the US market 
and may not be representative of this market.''\943\ The agencies have 
discussed the importance of aligning the CAFE vehicle models with the 
U.S. market earlier in Section VI.B.3

[[Page 24432]]

Technology Effectiveness Values and Section VI.C.1.d) Baseline Fleet. 
All of these factors make it difficult to compare directly the 
agencies' estimates and estimates presented in the FEV report cited by 
ICCT in their comments.
---------------------------------------------------------------------------

    \943\ Id. at p.141.
---------------------------------------------------------------------------

    HDS provided a variety of costs and effectiveness comparisons 
between the NPRM and previous 2012 final rule and the 2016 Draft 
TAR.\944\ Specifically, HDS stated that the data presented in the 2016 
TAR indicated a $60 per CO2/mile reduction for most 
conventional engine technologies.
---------------------------------------------------------------------------

    \944\ Duleep, K.G., ``Review of the Technology Costs and 
Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report, 
H-D Systems, October 2018, at p. 18-19.
---------------------------------------------------------------------------

    Although the comparison was technically sound, there are 
significant differences between the Draft TAR and NPRM analyses that 
clearly account for the differences in engine cost. First, the NPRM 
analysis used the MY 2016 fleet as a starting point to model 
manufacturers' potential responses to CAFE and CO2 
standards, whereas the 2012 final rule and Draft TAR used older 
baseline fleets. Vehicles in the MY 2016 fleet already included more 
advanced technologies than their predecessors in prior MY fleets, which 
would make it more expensive for vehicles that have already adopted 
advanced technologies to adopt more advanced technology. Second, the 
agencies refined the engine modeling from previous analysis to the NPRM 
to account for engine configurations and cylinder count more precisely. 
For the final rule analysis, the same approach was taken to account 
appropriately for costs for different type engine designs and 
configurations.
    Aside from these updates, engine costs were carried over from the 
NPRM analysis, except for newly added technologies, where costs were 
obtained from various sources such as NAS studies, technical 
publications, and CBI data. Finally, the cost estimates have been 
updated to account for dollar year (updated from 2016 dollars to 2018 
dollars), and learning rate.
(1) Basic Engines
    DMCs used for the final rule analysis for basic engine technologies 
were the same as NPRM costs. Table VI-49 below shows the basic engine 
DMC used for this final rule analysis.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.162

[GRAPHIC] [TIFF OMITTED] TR30AP20.163


[[Page 24433]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.164

BILLING CODE 4910-59-C
(2) Turbocharged Downsized Engines
    DMCs used for the final rule analysis for the turbocharged engine 
technologies were the same as NPRM costs. When these technologies are 
applied to V6 and V8 non-turbocharged engines, the incremental I4 and 
V6 turbocharged costs are applied, respectively. Table VI-52 below 
shows the DMC used for turbochared technologies for FRM analysis in 
2018 dollars.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.165

[GRAPHIC] [TIFF OMITTED] TR30AP20.166


[[Page 24434]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.167

BILLING CODE 4910-59-C
(3) Non-HEV Atkinson and Atkinson Engines
    DMCs used for the final rule analysis for HCR0 and HCR1 were based 
on HCR1 and HCR2 from NPRM, respectively. Discussed in Section 
VI.C.1.c).(3), the agencies aligned the cost of HCR technologies to 
align with 2015 NAS effectiveness and costs.
    Stakeholders commented on the costs of HCR technology compared to 
previous analysis. ICCT compared the NPRM costs to EPA's Proposed 
Determination costs, stating that ``[t]his is a clear case where the 
agencies appear to have not used the best available data from EPA which 
has extensively analyzed this technology and its associated cost, nor 
have the agencies justified how they have increased the associated 
costs, apparently by a factor of three.'' Similarly, Roush Industries 
commenting on behalf of CARB stated that the costs for implementing HCR 
technology were 5-6 times the 2016 Draft TAR estimated costs, which are 
``extremely high'' and ``will significantly overstate the incremental 
cost and bias technology pathways.''\945\ HDS also commented that the 
costs for HCR technology were higher than the costs from the 2016 Draft 
TAR, and speculated that was due to ``the bulky exhaust system used in 
the Mazda ATK1 engine, which apart from being expensive also requires 
the vehicle to be modified to accommodate the exhaust system.''\946\ 
HDS cited the 2018 Camry as an example of a vehicle that does not use 
the same exhaust system, but stated the sources of the new cost data 
were not documented in the PRIA. ICCT stated that ``[t]he agencies 
should reinstate the better justified and more deeply analyzed original 
Proposed Determination HCR cost numbers from EPA for this rulemaking.''
---------------------------------------------------------------------------

    \945\ NHTSA-2018-0067-11984.
    \946\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    The NPRM analysis and the final rule analysis used the same DMCs 
established by the 2015 NAS report for the Atkinson cycle technologies. 
However, because there are many various engine configurations in the 
market, the agencies do not use the same fixed costs that were set for 
each type of vehicle described in the 2015 NAS report, such as pickup 
and sedan. The agencies have expanded costs by taking into account the 
type of technology in the baseline, like SGDI, and the configuration of 
the engine, such as SOHC versus DOHC. In addition, the cost used in the 
NPRM also included updated dollar year, learning rate, and RPE. 
Although EPA also used costs from the 2015 NAS report for the Proposed 
Determination analysis, they used a different approach to account for 
components.\947\ For the final rule analysis the agencies continued to 
use the same DMC for HCR technologies. Table VI-55 below shows HCR DMCs 
used for the final rule analysis in 2018 dollars.
---------------------------------------------------------------------------

    \947\ EPA PD TSD at 2-307 to 2-308 ``Note that the NAS costs 
include the costs of gasoline direct injection (shown as ``DI'' in 
the NAS report row header). EPA has removed those costs (using the 
NAS reported values) since EPA accounts for those costs separately 
rather than including them in the Atkinson-2 costs. Note also that 
EPA always includes costs for direct injection, along with variable 
valve timing and other costs, when building an Atkinson-2 package.''
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.168


[[Page 24435]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.169

[GRAPHIC] [TIFF OMITTED] TR30AP20.170


[[Page 24436]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.171


[[Page 24437]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.172


[[Page 24438]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.173


[[Page 24439]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.174


[[Page 24440]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.175

BILLING CODE 4910-59-C
(4) Advanced Cylinder Deactivation Technologies
    DMCs used for the final rule analysis for the advanced cylinder 
deactivation technologies were the same as NPRM costs.
    Roush commented that in the NPRM analysis, the agencies did not 
properly

[[Page 24441]]

consider the ``very cost-effective benefits of skip-fire technology,'' 
referred to in the analysis as ADEAC. Roush stated that ``due to 
extremely high estimated cost ($1,250.00 in MY2016), the benefits of 
this technology will likely not be chosen in any reasonable technology 
pathway. If included, the predicted cost for that pathway will be 
overestimated by $750-$1,000.''\948\ Similarly, Meszler commented on 
the cost for the ADEAC system stating ``advanced cylinder deactivation 
paths are assumed (by NHTSA) to be expensive, and are selected only in 
rare instances.'' \949\ ICCT also stated ``The agencies estimated a 
greatly exaggerated cost of advanced cylinder deactivation for that 
level of the technology.'' \950\
---------------------------------------------------------------------------

    \948\ Roush at p.13.
    \949\ Meszler Comments, Attachment 2, NHTSA Docket No. NHTSA-
2018-0067-11723.
    \950\ ICCT comments, NHTSA-2018-0067-11741, Page I-71.
---------------------------------------------------------------------------

    The agencies do not agree with the commenter's statement that the 
analysis did not consider ADEAC as a cost effective technology or that 
the agencies overestimated costs for the technology. The agencies 
considered the most up to date information and data for the NPRM and 
final rule analysis.\951\ The agencies rely on the CAFE model to 
determine technology cost effectiveness, and if the technology was cost 
effective for a manufacturer to adopt, then the model would apply it to 
a manufacturer's vehicle. The adoption of ADEAC was applied to vehicles 
with corresponding technology combinations to reflect appropriate cost 
and effectiveness, as discussed in the paragraph above. The purpose of 
ADEAC is to reduce pumping losses, but if the engine has been 
downsized, or has already incorporated technologies that also reduce 
pumping loss, then it is likely the ADEAC has reached a point of 
diminishing return. As far as the agencies are aware, Roush did not 
provide alternative DMCs for ADEAC technology. Table VI-58 below shows 
the examples of advanced cylinder deactivation DMC used for both 
naturally aspirated and turbocharged engines for the final rule 
analysis in 2018$.
---------------------------------------------------------------------------

    \951\ Boha, Stani. ``Benchmarking and Characterization of a Full 
Continuous Cylinder Deactivation System.'' EPA. April 10-12, 2018 
SAEA World Congress. https://www.epa.gov/sites/production/files/2018-10/documents/deact-sae-world-congress-bohac-2018-04.pdf. (last 
accessed Feb 12, 2020).
    CARB. ``Tula Technology's Dynamic Skip Fire.'' September 28, 
2016. CARB_2016 Tula ppt skipfire_NHTSA-2018-0067-11985.pdf
---------------------------------------------------------------------------

(5) Miller Cycle Engines
    The agencies estimated costs for Miller cycle engines with VTG from 
2016 ICCT-sponsored FEV technology cost assessment report. The agencies 
considered costs from 2015 NAS study that referenced a NESCCAF 2004 
report,952 953 but believed that the reference material from 
the ICCT report had more updated cost estimates for this technology 
that represented what was discussed in the NPRM and modeled in the 
final rule analysis.
---------------------------------------------------------------------------

    \952\ ``Reducing Greenhouse Gas Emissions from Light-Duty Motor 
Vehicles.'' NESCCAF. September 23, 2004 Report. Available at https://www.nesccaf.org/documents/rpt040923ghglightduty.pdf/. Last accessed 
Dec. 22, 2019.
    \953\ ``VGT gasoline turbo, charge air cooler, piston upgrade, 
piston cooling, steel crankshaft, cooling system upsize, plumbing, 
rings, pressure sensor & bearing upgrade. Excludes any needed 
increase in transmission torque capacity or modifications to 
aftertreatment system.'' NESCCAF Report comment (2004).
---------------------------------------------------------------------------

    NAS estimated the incremental cost for VTG as $525 in 2010$, but 
this cost assumes many of the traditional turbocharged components and 
adds VVT, VVL and SGDI. In addition, VTG (Eng23b) and VTGe (Eng23c) 
engines both have similar modeled BMEP levels and a cooled EGR system 
to CEGR1 (Eng14), implying that the components such as cooling systems 
and piping will have similar costs.
    The NAS template to calculating the final DMCs for the Miller cycle 
engines for the different engine configuration is the $525 (2010$) plus 
cost of cEGR1 minus cost of VVT, VVL, and SGDI. The agencies estimated 
the cost for electrically-assisted variable supercharger VTGe (Eng23c) 
engines based on the 2015 NAS study that uses a cost of $1050 (2010$) 
plus the cost of the mild hybrid battery. For the final rule analysis, 
the total costs for these technologies are shown below.
BILLING CODE 4910-59-P

[[Page 24442]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.176


[[Page 24443]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.177


[[Page 24444]]


BILLING CODE 4910-59-C
(6) Variable Compression Ratio Engines
    DMCs used for the final rule analysis for the VCR engines were 
based on the 2015 NAS report.\954\ The 2015 NAS reported cost for VCR 
in MY2025 used a naturally aspirated engine; however, for this final 
rule analysis the agencies have added cEGR and other engine 
technologies to the engine. Total costs were updated to reflect 2018 
dollars and MY2017 learning rate which is based on the NPRM ADEAC 
learning rate. Table VI-67 below shows examples of VCR DMCs used for 
this this final rule analysis in 2018 dollars.
---------------------------------------------------------------------------

    \954\ 2015 NAS at p. 93.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24445]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.178


[[Page 24446]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.179


[[Page 24447]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.180

BILLING CODE 4910-59-C
(7) Diesel Engines
    DMCs used for the final rule analysis for diesel engine 
technologies were the same as the NPRM analysis. For DSLIAD 
technologies, the agencies have added the incremental cost of ADEAC to 
DSLI.

[[Page 24448]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.181

[GRAPHIC] [TIFF OMITTED] TR30AP20.182

(8) Alternative Fuel Engines
    DMCs used for the final rule analysis for CNG engine technologies 
were the same as the NPRM analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.183

[GRAPHIC] [TIFF OMITTED] TR30AP20.184


[[Page 24449]]


(9) Engine Lubrication and Friction Reduction Technologies
    EFR costs used for the final rule analysis are based on the 2015 
NAS assessment for low friction lubrication and engine friction 
reduction level 2 (LUB2_EFR2). The 2015 NAS report provided estimates 
of $51 (I4 DOHC), and $72 (V6 SOHC and DOHC) for midsize cars, in 2015 
dollars, relative to level 1 engine friction reduction (EFR1), which 
costs about $12 per cylinder. For this analysis, EFR technologies DMCs 
are estimated to be $14.05 per cylinder in 2016 dollars. Total costs 
were updated to reflect 2018 dollars and MY 2017 learning rate. Table 
VI-74 shows the EFR DMC used for the final rule analysis in 2018 
dollars.

[[Page 24450]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.185


[[Page 24451]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.186


[[Page 24452]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.187

BILLING CODE 4910-59-C

[[Page 24453]]

2. Transmission Paths
    Transmissions transmit torque from the engine to the wheels. 
Transmissions primarily use two mechanisms to improve fuel efficiency: 
(1) A higher gear count, as more gears allow the engine to operate 
longer at higher efficiency speed-load points; and (2) improvements in 
friction or shifting efficiency (e.g., improved gears, bearings, seals, 
and other components), which reduce parasitic losses.
    There are two major categories of transmission types modeled in the 
analysis: Automatic and manual. Automatic transmissions automatically 
select and shift between transmission gears for the driver during 
vehicle operation. The automatic transmission category is further 
subdivided into four subcategories: Traditional automatic 
transmissions, dual clutch transmissions, continuously variable 
transmissions, and direct drive transmissions. Manual transmissions 
require direct control by the driver to select and shift between gears 
during vehicle operation.
    Conventional planetary gear automatic transmissions (AT) are the 
most popular transmission.\955\ ATs typically contain three or four 
planetary gear sets that provide the various gear ratios. Gear ratios 
are selected by activating solenoids which engage or release multiple 
clutches and brakes as needed. ATs with gear counts ranging from five 
speeds to ten speeds were considered in the NPRM and final rule 
analysis.\956\
---------------------------------------------------------------------------

    \955\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
    \956\ Specifically, the agencies considered five-speed automatic 
transmissions (AT5), six-speed automatic transmissions (AT6), seven-
speed automatic transmission (AT7), eight-speed automatic 
transmissions (AT8), nine-speed automatic transmissions (AT9), and 
ten-speed automatic transmissions (AT10).
---------------------------------------------------------------------------

    ATs are packaged with torque converters, which provide a fluid 
coupling between the engine and the driveline, and provide a 
significant increase in launch torque. When transmitting torque through 
this fluid coupling, energy is lost due to the churning fluid. These 
losses can be eliminated by engaging the torque convertor clutch to 
directly connect the engine and transmission (``lockup'').
    Conventional continuously variable transmissions (CVT) consist of 
two cone-shaped pulleys, connected with a belt or chain. Moving the 
pulley halves allows the belt to ride inward or outward radially on 
each pulley, effectively changing the speed ratio between the pulleys. 
This ratio change is smooth and continuous, unlike the step changes of 
other transmission varieties. CVTs were not initially chosen in the 
fleet modeling for the 2012 rulemaking analysis for MYs 2017 and later 
because of the predicted low effectiveness associated with CVTs (due to 
the high internal losses and narrow ratio spans of CVTs in the fleet at 
that time).\957\ However, improvements in CVTs in the current fleet 
have increased their effectiveness, leading to increased adoption rates 
in the fleet. In its 2015 report, the NAS recommended CVTs be added to 
the list of considered technologies. The agencies included CVT 
technology for the NPRM and this final rule analyses.
---------------------------------------------------------------------------

    \957\ Morihiro, S., ``Fuel Economy Improvement by 
Transmission,'' presented at the CTI Symposium 8th International 
2014 Automotive Transmissions, HEV and EV Drives.
---------------------------------------------------------------------------

    Dual clutch transmissions (DCT), like automatic transmissions, 
automate shift and launch functions. DCTs use separate clutches for 
even-numbered and odd-numbered gears, allowing the next gear needed to 
be pre-selected, resulting in faster shifting. The use of multiple 
clutches in place of a torque converter result in lower parasitic 
losses than ATs. However, DCTs are seeing limited penetration in the 
fleet, and because of the low penetration rate, only two DCTs were 
considered in the analysis.
    Direct drive (DD) transmissions are a direct connection between the 
wheels and a drive motor. In a DD transmission, the ratio between wheel 
speed and motor speed remains constant. A DD transmission is only used 
in battery electric vehicles, and in the NPRM the agencies provided the 
specification for comments.\958\
---------------------------------------------------------------------------

    \958\ NHTSA-2018-0067-0003. ANL Autonomie Summary of Main 
Component Assumptions. Aug 21, 2018. NHTSA-2018-0067-0007. Islam, E. 
S, Moawad, A., Kim, N, Rousseau, A. ``A Detailed Vehicle Simulation 
Process To Support CAFE Standards 04262018--Report'' ANL Autonomie 
Documentation. Aug 21, 2018.Aug 21, 2018 NHTSA-2018-0067-0004. ANL 
Autonomie Data Dictionary. Aug 21, 2018.
---------------------------------------------------------------------------

    Manual transmissions (MT) are transmissions that require direct 
control by the driver to operate the clutch and shift between gears. 
Manual transmissions have seen a significant reduction in application 
by automakers over recent years. As a result of the reduced market 
presence, only three variants are used in the analysis.
a) Transmission Modeling in the CAFE Model
    The NPRM analysis modeled pathways for applying improved technology 
for each of the transmission categories and subcategories, except for 
the direct drive, which was only available in the battery electric 
vehicles. The MT and DCT pathways only included increasing gear counts 
(e.g., 5-speed manual transmission, 6-speed manual transmission, and 7-
speed manual transmission) as improved technologies.
    The traditional ATs and CVTs included both increased gear counts 
and high efficiency gearbox (HEG) technology improvements as options. 
HEG improvements for transmissions represent incremental advancement in 
technology that improves efficiency, such as: Reduced friction seals, 
bearings and clutches, super finishing of gearbox parts, and improved 
lubrication. All these advancements are aimed at reducing frictional 
and other parasitic loads in transmissions to improve efficiency. Three 
levels of HEG improvements are considered in this analysis, based on 
2015 NAS recommendations and based on CBI data.\959\ HEG efficiency 
improvements were applied to ATs and CVTs, as those transmissions 
inherently have higher friction and parasitic loads related to 
hydraulic control systems and greater component complexity, compared to 
MTs and DCTs.
---------------------------------------------------------------------------

    \959\ 2015 NAS Report, at 191.
---------------------------------------------------------------------------

    In total, 18 unique transmission technology combinations were 
simulated, using explicit input values for gear ratios, gear 
efficiencies, gear spans, shift logic, and transmission 
architecture.\960\ \961\ Table VI-77 shows a list of the multi-gear 
transmissions used for the NPRM.\962\
---------------------------------------------------------------------------

    \960\ See PRIA Chapter 6.3.
    \961\ Ehsan, I.S., Moawad, A., Kim, N., & Rousseau, A., ``A 
Detailed Vehicle Simulation Process To Support CAFE Standards.'' 
ANL/ESD-18/6. Energy Systems Division, Argonne National Laboratory. 
2018.
    \962\ The NPRM and final rule also included a direct drive 
transmission (single ratio) for BEVs.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24454]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.188

    The technologies that made up the four transmission/level paths 
defined by the modeling system for the NPRM analysis are shown in 
Figure VI-22. Each vehicle model in the analysis fleet is assigned an 
initial transmission type and level that most closely matches its 
configuration and characteristics. The baseline-level technologies 
(AT5, MT5 and CVT) appear in gray boxes and are only used to represent 
the initial configuration of a vehicle's transmission in the analysis 
fleet. Because there are only a few manual transmissions with less than 
five forward gears in the analysis fleet, for simplicity, all manual 
transmissions with five forward gears or fewer were designated MT5 for 
the analysis. Similarly, all automatic transmissions with five forward 
gears or fewer have been assigned the AT5 technology. For the NPRM 
analysis, the agencies included a 7-speed automatic and a 9-speed 
automatic to account for effectiveness of those transmissions in the 
analysis fleet. These two transmissions were not available for adoption 
but were available as initial configurations, and appear in gray boxes 
in Figure VI-22.

[[Page 24455]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.189

BILLING CODE 4910-59-C
    The model generally may apply any of the more efficient 
transmission technologies that are contained within the pathway of the 
baseline vehicle initial transmission configuration. The model 
prohibits manual transmissions from becoming automatic transmissions. 
Automatic transmissions may become CVT level 2 after progressing though 
the 6-speed automatic, as shown in Figure VI-22. While the structure of 
the model could allow automatic transmissions to consider applying a 
DCT, the market data file was used to preclude the application of DCTs 
to automatic transmission vehicles, as discussed

[[Page 24456]]

further in Section VI.C.2.c) Transmission Adoption Features, below.
    The model does not attempt to simulate ``reversion'' to less 
advanced transmission technologies, such as replacing a 6-speed AT with 
a DCT and then replacing that DCT with a 10-speed AT. The agencies 
invited comment on whether the model should be modified to simulate 
``reversion'' and, if so, how this possible behavior might be 
practicably simulated. Richard Rykowski, supporting comments from the 
Environmental Defense Fund (EDF), broadly discussed the concept of 
reversion in the CAFE model, and included an example relating to the 
transmission technology paths.\963\ Mr. Rykowski stated that it is 
``possible that the model could add a 10-speed transmission to a 
vehicle with a very basic engine'' and then as the simulation 
progressed and ``the manufacturer required greater fuel or 
CO2 emission control, the Volpe Model might move to a TURBO1 
or HCR engine'' and the vehicle would no longer need the 10-speed 
transmission to meet standards, and a 6-speed or 8-speed transmission 
might be more cost effective.
---------------------------------------------------------------------------

    \963\ Comments from Environmental Defense Fund, Attachment B, 
NPRM Docket No. NHTSA-2018-0067-12108, at 70.
---------------------------------------------------------------------------

    The scenario discussed by Mr. Rykowski is very unlikely. The CAFE 
model cost optimization algorithm considers both current and future 
standard requirements when selecting current MY technologies. The 
algorithm will look multiple years into the future and compare multiple 
potential technology paths going forward for the most cost-effective 
path. For a more detailed discussion on the cost optimization algorithm 
see Section VI.A.4, Compliance Simulation.
    Regarding the types of transmission technologies modeled, Meszler 
Engineering Services provided a comment criticizing the limited number 
of manual transmission model options and the limited technology paths 
available to vehicles with manual transmissions.\964\ The agencies do 
not agree with Meszler Engineering Service's assessment. The manual 
transmission path includes three model options and allows for the 
vehicles to receive electrification in the form of SS12V and BISG 
technologies. The agencies believe the technology paths dedicated to 
manual transmission was appropriate for vehicles that typically 
represent manufacturers' specialty performance cars, such as the Subaru 
STI or BMW M-series, that comprise an overall fleet share of less than 
2 percent.
---------------------------------------------------------------------------

    \964\ Comments from Meszler Engineering Services, 
Attachment2_CAFE Model Tech Issues, Docket No. NHTSA-2018-0067-
11723, at 33.
---------------------------------------------------------------------------

    Commenters also discussed potential missing transmission 
technologies in the NPRM analysis. ICCT stated that the agencies failed 
to consider transmission warm-up technologies, which are available in 
3.7 million new vehicles in the MY 2016 fleet, that are being deployed 
due to regulatory test-cycle benefits and off-cycle credits.\965\ In 
addition, the Fiat Chrysler Automobiles (FCA) also expressed concern 
over the lack of inclusion of thermal bypass devices in the modeling of 
transmission technologies.\966\
---------------------------------------------------------------------------

    \965\ Comments from ICCT, NPRM Docket No. NHTSA-2018-0067-11741 
full comments, at I-28.
    \966\ Comments from Fiat Chrysler Automobiles, Attachment 1, 
NPRM Docket No. NHTSA-2018-0067-11943, at 97.
---------------------------------------------------------------------------

    The agencies agree with parts of ICCT's and the FCAs comments and 
disagree with other parts. The agencies do agree with ICCT and the Auto 
Alliance that the analysis should consider the off-cycle benefits of 
transmission warm-up technology. For the final rule analysis, the 
agencies applied off-cycle technologies in the CAFE model. For the 
final rule analysis, the agencies applied off-cycle technologies at the 
maximum menu regulatory value of 10 g/mile for all manufacturers by MY 
2023. The modeled adoption included benefits of transmission warm-up as 
a menu item. The modeling of off-cycle technologies is further 
discussed in Section VI.C.8. The agencies disagree with ICCT and the 
Auto Alliance comments that transmission warm-up technologies were not 
included in the NPRM on-cycle analysis. For the NPRM, and for the final 
rule, the HEG level 2 technology package includes rapid transmission 
oil warm-up technology.\967\ The inclusion of the HEG2 technology 
package in AT and CVT models accounts for impacts of this technology to 
performance on the standard test-cycle.
---------------------------------------------------------------------------

    \967\ 2015 NAS Report, at 191.
---------------------------------------------------------------------------

    For the final rule analysis the transmission model paths are shown 
in Figure VI-23. For the final rule analysis, the baseline-only 
technologies (MT5, AT5, AT7L2, AT9L2, and CVT) are grayed and are only 
used to signify initial vehicle transmission configurations. For 
simplicity, all manual transmissions with five forward gears or fewer 
are assigned the MT5 technology in the analysis fleet. Similarly, all 
automatic transmissions with five forward gears or fewer are assigned 
the AT5 technology.
BILLING CODE 4910-59-P

[[Page 24457]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.190

    Since the Manual Transmission path terminates with MT7, the system 
assumes that all manual transmissions with seven or more gears are 
mapped to the MT7 technology. Moreover, all dual-clutch (DCT) or auto-
manual (AMT) transmissions with five or six forward gears are mapped to 
the DCT6 technology, and all DCTs or AMTs with seven or more forward 
gears are mapped to DCT8.
    For the final rule analysis, the naming convention for the 
transmission technology models was updated to identify better the 
technologies represented in each transmission. Although the 
technologies in each transmission configuration were described in the 
NPRM, there appears to have been confusion among some commenters about 
the technology content of some transmission configurations. Some 
commenters compared the NPRM AT10 to the NPRM AT8, and commented on 
unexpected differences in effectiveness relative to the differences in 
transmission gear count.\968\ For the given example, the NPRM AT8 
represented a baseline 8-speed automatic transmission, with level 1 HEG 
technology applied, and the NPRM AT10 represented a 10-speed automatic 
transmission with level 2 HEG technology applied. A direct comparison 
of gear count would occur by comparing the NPRM AT8L2 to the NPRM AT10. 
The updated naming convention identifies the transmission technology 
type, gear count and HEG technology level. Table VI-78 shows the final 
rule names for transmission models compared to the names used for the 
NPRM analysis.
---------------------------------------------------------------------------

    \968\ Comments from CARB, Attachment 2018-10-26 FINAL CARB 
Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067 at 110-
13.

---------------------------------------------------------------------------

[[Page 24458]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.191

BILLING CODE 4910-59-C
b) Transmission Analysis Fleet Assignments
    The agencies discussed in the NPRM the process for developing the 
2016 analysis fleet, including how the agencies weighed using 
confidential business information versus publicly-releasable sources, 
the use of compliance data, and decision to use a 2016 analysis fleet 
over other alternatives.\969\ As discussed above, this final rule 
analysis used the 2017 vehicle fleet as the analysis fleet input, and 
the agencies followed largely the same process for assigning initial 
transmission assignments as in the NPRM.
---------------------------------------------------------------------------

    \969\ 83 FR 43003.
---------------------------------------------------------------------------

    For the 2017 analysis fleet, transmission data was gathered from 
the manufacturer final model year CAFE compliance submissions to the 
agencies as well as manufacturer press releases. The data for each 
manufacturer was used to determine which platforms shared transmissions 
and to establish the leader-follower relationships between vehicles. 
Within each manufacturer fleet, transmissions were assigned unique 
identification designations based on technology type, drive type, gear 
count, and technology version. The data were also used to identify the 
most similar transmission among the Autonomie transmission models, as 
discussed further below.
    The transmission characteristics of vehicles in the analysis fleet 
show manufacturers use transmissions that are the same or similar on 
multiple vehicle models. Manufacturers have told the agencies they do 
this to control component complexity and associated costs for 
development, manufacturing, assembly, and service. Both the NPRM and 
final rule analyses account for this sharing. To identify common 
transmissions, the agencies considered the transmission type (manual, 
automatic, dual-clutch, continuously variable), number of gears, and 
vehicle architecture (front-wheel-drive, rear-wheel-drive, all-wheel-
drive based on a front-wheel-drive platform, or all-wheel-drive based 
on a rear-wheel-drive platform). If multiple vehicle models shared 
these attributes, the transmissions were treated as single group for 
the analysis. Vehicles in the analysis fleet with the same transmission 
configuration adopted transmission technology together.
    For ATs and CVTs, the identification of the most similar Autonomie 
transmission model required additional steps beyond just assigning gear 
count for ATs, or just assigning the CVT model. A review of the age of 
the transmission design, relative performance versus previous designs, 
and technologies incorporated was conducted, and the information 
obtained was used to assign a HEG level. Engineering judgment was used 
to compare the technologies and performance improvements reported 
versus descriptions of HEG technology discussed in the NAS report.\970\
---------------------------------------------------------------------------

    \970\ 2015 NAS Report, at 191.
---------------------------------------------------------------------------

    In addition, no automatic transmissions in the 2017 analysis fleet 
were determined to be initially at a HEG Level 3. However, all 7-speed 
automatic transmissions, all 9-speed automatic transmissions, all 10-
speed automatic transmissions and some 8-speed automatic transmissions 
were found to be advanced transmissions operating at a Level 2 HEG 
equivalence. All other transmissions were assigned at the minimum 
level.
c) Transmission Adoption Features
    The agencies included several transmission adoption features in the 
NPRM that have been carried over for the final rule analysis. For a 
detailed discussion of path logic applied in the final rule analysis, 
including technology supersession logic and technology mutual 
exclusivity logic, please see FRM CAFE Model Documentation Section 
S4.5, Technology Constraints

[[Page 24459]]

(Supersession and Mutual Exclusivity).\971\
---------------------------------------------------------------------------

    \971\ Available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

(1) Automatic Transmissions
    Automatic transmission technology adoption is defined by path logic 
and technology availability. The transmission path precludes adoption 
of other transmission types once a platform progresses past an AT6. 
This restriction is used to avoid the significant level of stranded 
capital that could result from adopting a completely different 
transmission type shortly after adopting an advanced transmission, 
which would occur if a different transmission type was adopted after 
AT6 in the rulemaking timeframe. Stranded capital is discussed in more 
detail in Section VI.B.4.c), Stranded Capital Costs. In addition, any 
automatic transmissions that use HEG3 technology cannot be phased in 
until the 2020 model year. The technology phase-in year is based on the 
estimated availability of HEG3 technology from the NAS (2015) report 
and confidential data obtained from OEM's and suppliers. Finally, all 
P2HEVs are paired with an AT8 transmission, which is also discussed 
further in Section VI.C.3.c).
    One commenter expressed concern that all P2HEVs were paired with an 
AT8 transmission, and argued that the full slate of transmission 
technology should be available for adoption with that powertrain 
technology.\972\ The commenter correctly observed a limit of 
transmission technologies for use only with the P2HEV technology 
option; all other HEV based technology options did not have this 
limitation.
---------------------------------------------------------------------------

    \972\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723 at 32.
---------------------------------------------------------------------------

    The agencies disagree that a greater variety of transmission 
technologies are necessary to model the P2HEV technology reasonably. 
The P2HEV demonstrated limited response to transmission technologies 
beyond the AT8L2, and access to those technologies were limited to 
reflect the diminishing returns anticipated for higher gear counts used 
in conjunction with the P2 system, and trends in industry.\973\ 
Adopting P2HEV to a conventional vehicle provides a significant fuel 
consumption improvement, agnostic of transmission type, based on the 
agencies' full vehicle simulation results.
---------------------------------------------------------------------------

    \973\ Greimel, H. ``ZF CEO--We're not chasing 10-speeds,'' 
Automotive News, November 23, 2014, http://www.autonews.com/article/20141123/OEM10/311249990/zf-ceo:-were-not-chasing-10-speeds.
---------------------------------------------------------------------------

(2) Continuously Variable Transmissions
    Application of CVTs in the NPRM and final rule analysis was not 
allowed for high torque vehicle applications. The launch, acceleration, 
and ratio variation characteristics of powertrains with CVTs may be 
significantly different than ATs leading to potential consumer 
acceptance issues and/or complaints. Several manufacturers have told 
the agencies that they employ strategies that mimic AT shifting under 
some conditions to address these issues. Some manufacturers have also 
encountered significant engineering challenges in employing CVTs for 
use in high torque or high load applications.
    In addition, the CVT adoption was limited by technology path logic. 
CVTs cannot be adopted by vehicles that do not start with a CVT or by 
vehicles beyond the AT6 in the baseline fleet which have a greater 
number of gear ratios and therefore increased ability to operate the 
engine at a highly efficient speed and load. Once on the CVT path the 
platform is only allowed to apply improved CVT technologies. This 
restriction is used to avoid the significant level of stranded capital 
that could result from adopting a completely different transmission 
type shortly after adopting an advanced transmission, which would occur 
if a different transmission type was adopted in the rulemaking 
timeframe. Stranded capital is discussed in more detail in Section 
VI.B.4.c), Stranded Capital Costs.
    The Alliance commented that the analysis ``appropriately restricts 
the application of CVT technology on larger vehicles.'' \974\ The 
agencies concurred with the Alliance's observations and thus the 
limitations on CVT application were continued in the final rule 
analysis.
---------------------------------------------------------------------------

    \974\ Comments from Auto Alliance, Attachment 1, NHTSA-2018-
0067-12073, at 142.
---------------------------------------------------------------------------

(3) Dual Clutch Transmission
    For DCTs, while the structure of the model could allow automatic 
transmissions to consider applying a DCT, the market data file was used 
to preclude the application of DCTs to vehicles that had already 
adopted an automatic transmission with six or more gears (e.g., AT6 
through AT10). The model allows baseline vehicles that have DCTs to 
apply an improved DCT (if opportunities to do so exist), and allows 
vehicles with an AT5 to consider DCTs. This was done to ensure vehicle 
functionality is maintained as technologies are applied, and accounts 
for consumer acceptance issues related to the drivability and launch 
performance tradeoffs. These issues with DCTs resulted in a low 
relative adoption rate over the last decade.\975\ It also is broadly 
consistent with manufacturers' technology choices.
---------------------------------------------------------------------------

    \975\ ``The 2018 EPA Automotive Trends Report,'' Page 60, figure 
4.18, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
---------------------------------------------------------------------------

(4) Manual Transmissions
    Manual transmission technology adoption in the CAFE model remained 
unchanged from the NPRM and is only limited by the technology path 
limits discussed above. Manual transmissions cannot be adopted by 
vehicles that do not start with a manual transmission in the analysis 
fleet. Vehicles with manual transmissions cannot receive an alternate 
transmission technology, and may only progress to more advanced manual 
transmissions. These restrictions are in recognition of the low 
customer demand for manual transmissions.\976\
---------------------------------------------------------------------------

    \976\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
---------------------------------------------------------------------------

d) Transmission Effectiveness Modeling and Resulting Effectiveness 
Values
    For the NPRM and final rule analysis, full vehicle simulation was 
used to understand how transmissions work within the full vehicle 
system to improve fuel economy, and how changes to the transmission 
subsystem influence the performance of the full vehicle system.
    The Autonomie tool models transmissions as a sequence of mechanical 
torque gains. The torque and speed are multiplied and divided, 
respectively, by the current ratio for the selected operating 
condition. Furthermore, torque losses corresponding to the torque/speed 
operating point are subtracted from the torque input. Torque losses are 
defined based on a three-dimensional efficiency lookup table that has 
as inputs: Input shaft rotational speed, input shaft torque, and 
operating condition.\977\
---------------------------------------------------------------------------

    \977\ Detailed discussion of transmission modeling can be found 
in the ANL Model Documentation at Chapter 4 and Chapter 5.
---------------------------------------------------------------------------

    The general transmission models are populated with characteristics 
data to model specific transmissions. Characteristics data are 
typically provided in the form of tabulated data for transmission gear 
ratios, maps for transmission efficiency, and maps for torque converter 
performance, as applicable. The quantity of data needed

[[Page 24460]]

depends on the transmission technology being modeled. The 
characteristics data for these models was collected from peer-reviewed 
sources, transmission and vehicle testing programs, results from 
simulating current and future transmission configurations, and 
confidential data obtained from OEMs and suppliers.\978\
---------------------------------------------------------------------------

    \978\ Downloadable Dynamometer Database.: https://www.anl.gov/energy-systems/group/downloadable-dynamometer-database, Kim, N., 
Rousseau, N., Lohse-Bush, H., ``Advanced Automatic Transmission 
Model Validation Using Dynamometer Test Data,'' SAE 2014-01-1778, 
SAE World Congress, Detroit, April 2014. Kim, N., Lohse-Bush, H., 
Rousseau, A., ``Development of a model of the dual clutch 
transmission in Autonomie and validation with dynamometer test 
data,'' International Journal of Automotive Technologies, March 
2014, Volume 15, Issue 2, pp 263-271.
---------------------------------------------------------------------------

    The level of HEG improvement applied to a given transmission was 
modeled by improvements made to the efficiency map of the transmission. 
As an example, the 8-speed automatic transmission models show how a 
model can be incrementally improved with the addition of the HEG 
enhancement. The AT8 is the model of a baseline transmission developed 
from a transmission characterization report.\979\ The AT8L2 has the 
same gear ratios as the AT8, however the gear efficiency map has been 
improved to represent application of the HEG level 2 technologies. The 
AT8L3 models the application of HEG level 3 technologies using the same 
principle, further improving the gear efficiency map over the AT8L2 
improvements.
---------------------------------------------------------------------------

    \979\ See PRIA Section 6.3.3.2
---------------------------------------------------------------------------

    The NPRM and final rule analysis, using the Autonomie tool, 
comprehensively simulated each of the 18 transmission technologies. 
Each transmission was modeled with explicit gear ratios, gear 
efficiencies, gear spans, adaptive shift logic, and transmission 
architecture individually for each of the ten vehicle types. The NPRM 
and final rule analysis clearly showed the specific contributions to 
effectiveness provided by each transmission technology combination and 
the associated cost. This provided greater transparency for public 
review and comment.
    The implementation of the full vehicle simulation approach used in 
the NPRM analysis, and carried forward to the final rule analysis, 
clearly defines the contribution of individual transmission 
technologies and separates those contributions from other technologies. 
This modeling approach comports with the National Academy of Science 
2015 recommendation to use full vehicle modeling supported by 
application of collected improvements at the sub-model level.\980\ The 
approach allows the isolation of technology effects in the analysis 
which contributes to an accurate cost assessment.
---------------------------------------------------------------------------

    \980\ 2015 NAS Report, at 292.
---------------------------------------------------------------------------

    This approach was supported by the Auto Alliance, who commented in 
support of the agencies' explicit and transparent modeling of the cost 
and effectiveness for each of the transmission technologies. The 
Alliance contrasted the NPRM approach with the transmission modeling 
methodology used in the Proposed Determination--which they strongly 
objected to--which had lumped together fundamentally different 
transmission technologies into bundles with identical cost and 
efficiencies, ``making it impossible to fully comprehend the 
rationale'' for the Proposed Determination's high effectiveness 
estimates.\981\
---------------------------------------------------------------------------

    \981\ Comments from Alliance of Automobile Manufacturers, NHTSA-
2018-0067-12073, at 142.
---------------------------------------------------------------------------

    However, other stakeholders were not supportive of the modeling 
approach used in the NPRM. The Union of Concerned Scientists (UCS) 
thought a level of abstraction was necessary to account for 
unpredictability in the market, such as the failure of the dual-clutch 
transmission to reach widespread use as anticipated in the agencies 
2012 analysis for MYs 2017 and later. UCS thought that keeping the 
transmission technology generalized would avoid the pitfalls of 
potentially picking the wrong technology leader, but would still 
predict the general trend of behavior, stating that ``[i]ncidentally, 
this is an example of why we supported EPA's move to a more generic 
representation of transmissions in its OMEGA modeling.'' \982\
---------------------------------------------------------------------------

    \982\ Comments from Union of Concerned Scientists, NHTSA-2018-
0067-12039, at 20-21.
---------------------------------------------------------------------------

    The agencies disagree with UCS's suggestion to generalize the 
transmission technology groupings for the analysis. By grouping the 
technologies into overly broad, generic categories, the analysis loses 
accuracy on the costs and the effectiveness for specific systems. The 
OMEGA model used general transmission categories, asked for by UCS's 
comments, as part of the CO2 analysis in the Draft TAR and 
in the Proposed Determination, and the assumptions and limitations were 
acknowledged at the time.983 984 One assumption used by the 
OMEGA model approach was ``[t]he incremental effectiveness and cost for 
all automated transmissions are based on data from conventional 
automatics.'' \985\ In response, the Alliance observed that the 
transmission groups used ``do not recognize unique efficiencies of 
different transmission technologies.'' \986\ At the time EPA stated 
``the potential effectiveness gains between TRX levels, while arising 
from different technology packages within each transmission type, will 
be very similar among the transmission types.'' \987\ However, as shown 
in Table VI-81 and Table VI-82, there are nontrivial differences in the 
costs of different transmission technologies.
---------------------------------------------------------------------------

    \983\ ``Midterm Evaluation of Light duty Vehicle Greenhouse Gas 
Emission Standards and Corporate Average Fuel Economy Standards for 
Model Years 2022-2025,'' Paragraph 5.3.4.2.1, EPA-420-D-16-900, July 
2016.
    \984\ ``Proposed Determination on the Appropriateness of the 
Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions 
Standards under the Midterm Evaluation, Technical Support 
Document,'' Pages 2-328--2-329, EPA-420-R-16-021, November 2016.
    \985\ ``Proposed Determination on the Appropriateness of the 
Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions 
Standards under the Midterm Evaluation, Technical Support 
Document,'' Pages 2-327, EPA-420-R-16-021, November 2016.
    \986\ ``Proposed Determination on the Appropriateness of the 
Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions 
Standards under the Midterm Evaluation, Technical Support 
Document,'' Pages 2-329, EPA-420-R-16-021, November 2016.
    \987\ ``Proposed Determination on the Appropriateness of the 
Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions 
Standards under the Midterm Evaluation, Technical Support 
Document,'' Pages 2-329, EPA-420-R-16-021, November 2016.
---------------------------------------------------------------------------

    The approach used in the NPRM analysis and this final rule analysis 
is an evolution of the approach used for the Proposed Determination 
model, and avoids the issue described above. The NPRM and final rule 
analyses reduce the span of transmission technology groupings, with the 
intent to provide an increase in fidelity and precision for cost and 
performance, as was requested by stakeholders such as the Auto 
Alliance, while including tools to mitigate market effects, which 
addresses other concerns such as those expressed by UCS. In the 
analysis for the final rule the transmissions are grouped by technology 
type (AT, DCT, CVT, etc.) and gear count (5,6,7, etc.). The level of 
HEG technology applied as a separate factor further subdivided the 
transmission groups. Defining technology adoption features addresses 
the potential for market forces, such as those that affected the sales 
of DCTs, and supports the narrower technology groupings. Technology 
adoption features are defined through market research, historic and 
current fleet composition analysis, and dialogue with manufacturers.
    Commenters also provided general comments regarding the values of 
effectiveness for advanced transmissions used for the NPRM

[[Page 24461]]

analysis versus values used for the Draft TAR. For example, CARB noted 
a ``2 percent-3 percent lower efficiency assumed for advanced 8- and 9-
speed transmissions relative to the data EPA itself previously 
developed with back to back testing on FCA vehicles,'' \988\ with 
similar concerns expressed by other commenters.\989\ Meszler 
Engineering Services wondered ``why the AT10 technology was being so 
widely adopted when its associated benefits appeared negligible for a 
particular vehicle'' and noted ``[t]he wide ranging effectiveness 
estimates were unexpected.'' \990\ Senator Tom Carper also noted ``the 
most advanced eight speed transmission technology are assigned 
unrealistically low fuel efficiency effectiveness values for some 
vehicle types.'' \991\
---------------------------------------------------------------------------

    \988\ Comments from CARB, Attachment 2018-10-26 FINAL CARB 
Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at 
110-113.
    \989\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 5; Comments from CARB, Attachment HDS 
Final Report, NPRM Docket No. NHTSA-2018-0067-11985, at 26, 47.
    \990\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723, at 5-6.
    \991\ Comments from Senator Tom Carper, Attachment 1, NPRM 
Docket No. NHTSA-2018-0067-11910, at 4.
---------------------------------------------------------------------------

    The Auto Alliance also provided comments with regards to the larger 
variation of effectiveness values that were of concern to commenters 
such as Meszler Engineering Services and Senator Tom Carper. The Auto 
Alliance acknowledged that the use of full vehicle simulation, with 
more details, results in greater diversity of results. The comment 
stated, ``Over an entire fleet, a more reasonable expectation is that 
there will be some vehicles with higher fuel economy than expected for 
a given technology set and some vehicles with a lower fuel economy than 
expected for a given technology set. As discussed above, these 
differences arise for a variety of reasons, and cannot simply be 
attributed to ``less than optimal technology integration.'' \992\
---------------------------------------------------------------------------

    \992\ Comments from Alliance of Automobile Manufacturers, 
Attachment 1, NPRM Docket No NHTSA-2018-0067-12385, at 9.
---------------------------------------------------------------------------

    The Auto Alliance also specifically commented on the FCA vehicle 
study used to support CARB's comment and used to generate the TAR 
analysis values. The Auto Alliance pointed out that the vehicles used 
in the study had other technology differences, however the study still 
``proceeds to compare the fuel economy of these variants to assert 
support for its own estimate of transmission effectiveness. This 
comparison neglects that the 2.4L engines in these variants are not the 
same and that the variant with the nine-speed transmission was a 
redesigned vehicle.'' The Alliance concluded, therefore, that ``the 
Chrysler 200 comparison provided by H-D Systems does not compare a 
transmission change in isolation from other changes that impact fuel 
economy and likely overestimates the benefits associated with the 
transmission change.'' The Auto Alliance summarized the analysis of the 
study by noting that ``[s]uch differences also impact fuel economy, 
confounding an analysis which purports to compare the fuel economy 
benefits associated directly with the transmission.'' \993\
---------------------------------------------------------------------------

    \993\ Comments from Alliance of Automobile Manufacturers, 
Attachment 1, NPRM Docket No NHTSA-2018-0067-12385, at 27-28.
---------------------------------------------------------------------------

    The agencies agree with the Auto Alliance assessment of the 8- and 
9-speed FCA vehicles, and have based analysis inputs on alternate 
information sources.\994\ However, the observations by commenters of a 
wider range of values for the NPRM effectiveness when compared to the 
Draft TAR compliance analyses are a direct result of the improvements 
in modeling approach. As discussed above the NPRM compliance analysis 
increased the number of transmission technology paths considered by 
further subdividing the technology groupings. The change resulted in a 
wider range of effectiveness, as the specific transmission technologies 
are paired across all the configurations of vehicle technologies. In 
addition to this greater range, there were also specific effectiveness 
issues identified for some of the transmission technologies, which are 
addressed in the sections below.
---------------------------------------------------------------------------

    \994\ See Data discussed in PRIA Section 6.3.3.2. and Kim, N., 
Rousseau, N., Lohse-Bush, H. ``Advanced Automatic Transmission Model 
Validation Using Dynamometer Test Data,'' SAE 2014-01-1778, SAE 
World Congress, Detroit, April 2014. Kim, N., Lohse-Bush, H., 
Rousseau, A. ``Development of a model of the dual clutch 
transmission in Autonomie and validation with dynamometer test 
data,'' International Journal of Automotive Technologies, March 
2014, Volume 15, Issue 2, pp 263-271.
---------------------------------------------------------------------------

    Commenters may also be observing, with comments like ``advanced 
transmissions have low effectiveness with some vehicles types,'' an 
expected effect when an advanced transmission is coupled to an advanced 
engine. The National Academy of Science, in their 2015 report, noted 
that ``as engines incorporate new technologies to improve fuel 
consumption, including variable valve timing and lift, direct 
injection, and turbocharging and downsizing, the benefits of increasing 
transmission ratios or switching to a CVT diminish.'' \995\ This is not 
to say that transmissions are not an important technology going 
forward, but rather a recognition that advanced engines have larger 
``islands'' of low fuel consumption that rely less on the transmission 
to improve the overall efficiency of the vehicle. Thus, effectiveness 
percentages reported for transmissions paired with unimproved engines 
would be expected to be reduced when the same transmission is paired 
with a more advanced engine.
---------------------------------------------------------------------------

    \995\ 2015 NAS Report, at 175.
---------------------------------------------------------------------------

    Commenters also expressed concern for the transmission gear set and 
final drive values used for the NPRM analysis, or, more specifically, 
that the gear ratios were held constant across applications. Roush 
commented that ``all transmissions with a given number of ratios (8-
speed, 10-speed) maintain the same individual step ratios'' and that 
this would lead to ``powertrain inefficiencies and under-predict 
potential fuel economy benefits.'' \996\ CARB, quoting a report from 
its contractor, noted that ``the final drive ratio was kept constant as 
powertrains were changed and that transmission gear ratios were not 
optimized,'' and suggested that manufacturers forgoing improvements 
from gear ratio or final drive ratio changes is unrealistic and results 
in an underestimation of the benefits from advanced transmissions.\997\
---------------------------------------------------------------------------

    \996\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 14-15.
    \997\ Comments from CARB, Attachment 1, NPRM Docket No. NHTSA-
2018-0067-11873, at 110.
---------------------------------------------------------------------------

    However, the Auto Alliance stated that ``[m]anufacturers share 
major technologies such as transmissions and engines across multiple 
vehicle models and platforms.'' The Auto Alliance also supported the 
agencies' approach of not including final drive ratio changes, 
particularly when only minor system changes are incurred. The Auto 
Alliance continued further stating that ``[i]n the case of passenger 
cars, the final drive ratio is frequently the same across multiple 
models that use the same transmission.'' \998\
---------------------------------------------------------------------------

    \998\ Comments from Auto Alliance, Attachment 1, NPRM Docket No. 
NHTSA-2018-0067-12073, at 142.
---------------------------------------------------------------------------

    The agencies disagree with Roush, Duleep, and CARB's assessment. It 
is an observable practice in industry to use a common gear set across 
multiple platforms and applications. The most recent example is the GM 
10L90, a 10-speed automatic transmission that used the same gear set in 
both pick-up truck

[[Page 24462]]

and passenger car applications.\999\ Optimization of performance is 
achieved through shift control logic rather than customized hardware 
for each vehicle line. The use of a single gear set for each 
transmission technology also supports the overall analysis approach. 
The level of technology performance modeled must reasonably represent a 
typical level of performance representative of the industry range of 
performance. If the systems were over-optimized for the agencies' 
modeling, such as applying a unique gear set for each individual 
vehicle configuration, the analysis would likely over-predict the 
reasonably achievable fuel economy improvement for the technology. 
Over-prediction would be exaggerated when applied under real-world 
large-scale manufacturing constraints necessary to achieve the 
estimated costs for the transmission technologies. Accordingly, the 
agencies used the NPRM approach for the final rule analysis.
---------------------------------------------------------------------------

    \999\ ``GM Global Propulsion Systems--USA Information Guide 
Model Year 2018'' (PDF). General Motors Powertrain. Retrieved 26 
September 2019. https://www.gmpowertrain.com/assets/docs/2018R_F3F_Information_Guide_031918.pdf.
---------------------------------------------------------------------------

    In response to comments related to the effectiveness of micro-HEV 
systems, which are discussed in Section VI.C.3.d)(2)(a), and comments 
related to the effectiveness of diesel engines, which are discussed in 
Section VI.C.1.c)(8), the agencies took a close look at NPRM 
effectiveness results. Two issues were identified related to the 
interaction between Autonomie transmission models and other Autonomie 
powertrain technology models. First, a logic issue was found in a 
transmission control subroutine and, second, there was an issue with a 
sub-model input. While these items were caused by issues in the 
transmission model sub-systems, the effects manifested in the 
effectiveness of the micro-HEV systems and the diesel engine systems. 
Autonomie uses a gearbox transient sub-model to control the simulated 
state of powertrain components during a transmission event, such as 
shifting or vehicle starting and stopping. The simulated powertrain 
component states include conditions such as clutch engagement, or 
engine operation mode. A detailed discussion of the Autonomie control 
model can be found FRM Argonne Model Documentation file at Section 4.4. 
Different versions of the sub-model are used for micro-HEV technologies 
(12VSS and ISG) than for conventional drivetrains, mild-HEV or Strong-
HEV systems.
    An issue was found in the control logic used in the micro-HEV 
version related to the sequence of powertrain component modes during 
shifting events for automatic transmissions, regenerative braking 
events for automatic transmissions, and stop start events for manual 
transmissions. While these issues reduced the effectiveness of the 
micro-HEV technology in the Argonne modeling results, they had very 
minimal effect on the overall NPRM Analysis. The control logic issue 
was resolved for the final rule analysis. There also was an issue with 
the gearbox transient sub-model used for micro HEVs that impacted 
calculation of the CVT best efficiency operating ratio targets under 
low torque conditions. This resulted in some negative effectiveness 
values for certain CVT technology combinations, but had very minimal 
effect on the overall NPRM results. This software item was also 
resolved for the final rule analysis.
    As discussed in the Autonomie model documentation, FRM Argonne 
Model Documentation file at Section 4, the full vehicle model is 
created from a network of subsystem models. The subsystems all interact 
through data connections transferring outputs from one subsystem model 
to the inputs of another. An issue was identified with the definition 
of the connection between the gearbox transient sub-model for DCT's 
with diesel engines, which impacted the values provided to the diesel 
control model. This caused reduced effectiveness values for the diesel 
engines with DCTs in the Argonne modeling results, however it had very 
minimal effect on the overall NPRM analysis. The data connection issue 
was resolved for the final rule analysis.
    Lastly, the agencies received several comments on transmission 
shifting logic, which are addressed in the following section.
(1) Shift Logic
    Transmission shifting logic has a significant impact on vehicle 
energy consumption and was modeled in Autonomie to maximize the 
powertrain efficiency while maintaining acceptable drive quality. The 
logic used in the Autonomie full vehicle modeling relied on two 
components: (1) The shifting controller, which provides the logic to 
select appropriate gears during simulation; and (2) the shifting 
initializer, an algorithm that defines shifting maps (i.e., values of 
the parameters of the shifting controller) specific to the selected set 
of modeled vehicle characteristics and modeled powertrain 
components.\1000\
---------------------------------------------------------------------------

    \1000\ See FRM ANL Model Documentation file at Paragraph 4.4.5.
---------------------------------------------------------------------------

(a) Shifting Controller
    The shift controller is the logic that governs shifting behavior 
during simulated operation. The shift controller performance was 
informed by inputs from the model. The inputs included: Specific engine 
or transmission used, and instantaneous conditions in the simulation. 
Instantaneous conditions included values such as vehicle speed, driver 
demand and a shifting map unique to the full vehicle 
configuration.\1001\ The shift controller logic was consistently 
applied for all vehicles simulated.
---------------------------------------------------------------------------

    \1001\ See FRM ANL Model Documentation file at Paragraph 4.4.5.
---------------------------------------------------------------------------

    Although no comments were received specifically on shift control 
logic, the agencies tracked several effectiveness concerns identified 
by commenters back to how the agencies modeled some transmissions 
paired with turbocharged engines. Meszler Engineering Services 
discussed an unexpected range of effectiveness observed for 
transmissions when coupled to different engine technologies, and 
concluded that ``[m]oreover, the variation across technology 
combinations is markedly different.'' \1002\ Senator Carper's comments 
mirrored Meszler's, noting that ``the more expensive version of an 
engine technology (TURBO2), which would be expected to be more fuel-
efficient, was instead assigned a negative fuel-efficiency value for 
some types of vehicles.'' \1003\ The Senator also observed the same 
phenomenon for cooled exhaust gas recirculation (CEGR I), which ``was 
assigned a fuel-efficiency effectiveness of at or near zero.'' 
Similarly, UCS noted that ``many simulations of improved transmissions 
and turbocharged engines show little incremental improvement over less 
complex technologies.'' \1004\
---------------------------------------------------------------------------

    \1002\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723, at 5-6.
    \1003\ Comments from Senator Tom Carper, Attachment 1, NPRM 
Docket No. NHTSA-2018-0067-11910, at 4.
    \1004\ Comments from UCS, Attachment 1, NPRM Docket No. NHTSA-
2018-0067-12039, at 32.
---------------------------------------------------------------------------

    In response to the comments, the agencies conducted an in-depth 
review of these technology combinations. The agencies determined the 
minimum lugging speed for turbocharged engines, which controls the 
minimum engine speed allowed before down-shifting, caused the observed 
behavior. The issue was isolated to some combinations of advanced 
transmissions and

[[Page 24463]]

turbocharged engines. For the final rule analysis, a modification was 
made to the shift controller logic of transmissions coupled to 
turbocharged engines. Specifically, the minimum lugging speed allowed 
for turbocharged engines was increased in the shift controller. An 
increase in lugging speed increases the minimum speed at which the 
shift controller will allow the engine to operate before down-shifting, 
resulting in increased operation in better efficiency regions of the 
engine map.\1005\ The updated lugging speeds are based on Argonne 
benchmarking data of the 2017 F150.\1006\ The updated values are shown 
in Table VI-79, the lugging speeds for naturally aspirated engines are 
shown as reference and remain unchanged from the NPRM.
---------------------------------------------------------------------------

    \1005\ See FRM ANL Model Documentation at Paragraph 4.4.5.1, for 
more details on lugging speed.
    \1006\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford 
F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812 
520.
[GRAPHIC] [TIFF OMITTED] TR30AP20.192

(b) Shift Initializer
    As defined above, the shifting initializer is an algorithm that 
defines shifting maps (i.e., values of the parameters of the shifting 
controller) specific to the selected set of modeled vehicle 
characteristics and modeled powertrain components.
    Commenters stated that the model did not customize shifting maps 
for each transmission application. Roush Industries commented, ``[t]he 
2018 PRIA analysis assumes that all transmissions with a given number 
of ratios maintain the same individual step ratios and shift maps.'' 
\1007\ Roush also commented that the effectiveness of transmissions 
were understated due to inaccurate transmission maps or ``the lack of 
vehicle system optimization and calibration.'' \1008\ UCS stated that 
the ``transmission shift strategy does not deploy gear-skipping or 
other more modern control strategies.'' \1009\ HDS provided similar 
comments to Roush, observing that the Autonomie models ``do not 
optimize engine efficiency after most changes in tractive load because 
the model employs fixed shift points, gear ratios, and axle ratios.'' 
\1010\ Finally, CARB expressed that ``[f]or the Autonomie modeling, a 
fixed final drive ratio was utilized and, presumably, a fixed shift 
logic based on the selected transmission.'' \1011\
---------------------------------------------------------------------------

    \1007\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 14-15.
    \1008\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 5.
    \1009\ Comments from UCS, Attachment 1, NPRM Docket No. NHTSA-
2018-0067-12039, at 23.
    \1010\ Comments from K. Gopal Duleep, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-12395, at 4-5.
    \1011\ Comments from CARB, Attachment 2018-10-26 FINAL CARB 
Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at 
185.
---------------------------------------------------------------------------

    The commenters seem to conflate the practice in the analysis of 
using the same gear sets across vehicle configuration with using the 
same shift maps. As commenters stated, they assumed the same maps were 
applied across vehicle models. However, the shift initializer routine 
was run for every unique Autonomie full vehicle model configuration and 
generated customized shifting maps. The algorithms' optimization was 
designed to balance minimization of energy consumption and vehicle 
performance.\1012\ This balance was necessary to achieve the best fuel 
efficiency while maintaining customer acceptability by meeting 
performance neutrality requirements, as discussed in Performance 
Neutrality, Section VI.B.3.a)(6).
---------------------------------------------------------------------------

    \1012\ See FRM ANL Model Documentation at Paragraph 4.4.5.2.
---------------------------------------------------------------------------

    While discussing shift logic, commenters also expressed concern 
about the capturing of fuel efficiency losses associated with shifting 
events. Roush stated, ``[t]he 2018 PRIA transmission modeling does not 
accurately capture the losses and FE penalty associated with a shift 
event.'' \1013\ The agencies disagree with this statement. While losses 
associated with a shifting event are not modeled as a single factor, 
the mechanisms that cause the loss are appropriately incorporated in 
the Autonomie transmission models.
---------------------------------------------------------------------------

    \1013\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 14-15.

---------------------------------------------------------------------------

[[Page 24464]]

The automatic transmission models have an associated torque converter 
model.\1014\ The torque converter model is designed to simulate the 
inertial and torque loads imposed on an engine because of shift events. 
Other clutch-based transmission models, MTs and DCTs, apply a general 
loss of efficiency across transmission efficiency maps to account for 
losses due to shift events.
---------------------------------------------------------------------------

    \1014\ See FRM ANL Model Documentation at Paragraph 4.5 and 
Paragraph 5.4.
---------------------------------------------------------------------------

(2) Transmission Effectiveness Values
    The NPRM technology effectiveness modeling results showed that the 
effectiveness of a technology often varies with the type of vehicle and 
the other technologies that are on the vehicle. Figure VI-24 shows the 
range of effectiveness for each transmission technology across the 
range of vehicle types and technology combinations in the NPRM 
analysis. The data reflect the change in effectiveness for applying 
each transmission technology by itself while all other technologies are 
held unchanged. The effectiveness improvement range is over a 5-speed 
automatic transmission.
[GRAPHIC] [TIFF OMITTED] TR30AP20.193

(a) Automatic Transmissions
    Regarding AT effectiveness values, commenters pointed out the 
unusually high level of effectiveness displayed by the AT6L2 
transmission. ICCT and UCS both specifically expressed concern with the 
effectiveness of the AT6L2 compared to other advanced 
transmissions.1015 1016 The performance of the AT6L2 was 
central to ICCT's analysis of the NPRM inputs, which highlighted the 
AT6L2 models' performance, showing the cost versus effectiveness of the 
AT6L2 outperformed more advanced transmission options.\1017\
---------------------------------------------------------------------------

    \1015\ Comments from International Council on Clean 
Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741, 
at I-26, I-64 (`` ``However, the impact of adding level 2 
transmission efficiency technologies varies wildly and produces 
absurd results. A 6-speed AT6L2 Is modeled as much more efficient 
(12.0% improvement) than a comparable 8-speed AT8L2 (9.1%) and even 
slightly more efficient than a comparable 10-speed AT10L2 
(11.5%).'')%).''.
    \1016\ Comments from Union of Concerned Scientists, Attachment 
1, NPRM Docket No. NHTSA-2018-0067-12039, at 32. (``[I]n the NPRM 
analysis, 0 percent of vehicles had an AT6L2 transmission while 52.4 
percent adopted AT10L2 transmissions, even though the latter 
supplies virtually identical modeled efficiency.'').
    \1017\ Comments from International Council on Clean 
Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741, 
at I-64--I-65.
---------------------------------------------------------------------------

    Evaluation of the AT6L2 transmission model in response to these 
comments revealed an overestimated efficiency map was developed for the 
NPRM model. The high level of efficiency assigned to the transmission 
surpassed benchmarked advanced transmissions.\1018\ To address the 
issue, the agencies replaced the effectiveness values of the AT6L2 
model for the final rule analysis with AT7L2 effectiveness values.
---------------------------------------------------------------------------

    \1018\ See PRIA Section 6.3.3.2. Sources of Transmission 
Effectiveness Data.
---------------------------------------------------------------------------

    The updated estimate of effectiveness is supported by values shown 
in the NAS 2015 analysis.\1019\ The study estimated the difference in 
effectiveness between a 6-speed automatic transmission and a 7-speed 
automatic transmission of approximately the same technology level to be 
0.8 percent. The difference is reduced further when application of high 
efficiency gear box technology ranges of effectiveness is applied. 
Because the 7-speed automatic transmission and the advanced 6-speed 
automatic transmission technologies are parallel on the technology 
tree, the agencies felt using the same effectiveness value was 
reasonable and appropriate.
---------------------------------------------------------------------------

    \1019\ 2015 NAS Report, at page 189.
---------------------------------------------------------------------------

    Commenters also pointed out a lack of skip-shift logic used in the 
NPRM analysis, and an increase in the shift busyness observed for the 
high gear count transmissions. Roush commented on the NPRM analysis 
``not incorporating the concept of `Skip shifting' which is important 
for reducing shift busyness and increasing FE especially in vehicles 
equipped with transmission with a large number of ratios (8-10).'' 
\1020\ Both CARB and UCS repeated similar concerns.\1021\
---------------------------------------------------------------------------

    \1020\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984 at 14-15.
    \1021\ Comments from CARB, Attachment 2018-10-26 FINAL CARB 
Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at 
110-113 (``Rogers found that the modeling did not consider `skip-
shifting' where a transmission can upshift or downshift in a non-
sequential manner''). Comments from UCS, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-12039, at 23 ``including that ANL's transmission 
shift strategy does not deploy gear-skipping'').''.
---------------------------------------------------------------------------

    After consideration of the comments and re-evaluation of the NPRM 
results, the agencies concurred with the commenters. The lack of skip-
shift logic and increased shift busyness can result in lower overall 
efficiency and decreased consumer acceptance. For the final rule 
analysis, a skip-shift logic was applied to the 10 speed automatic 
transmissions. The logic was based on the baseline 2017 Ford F150 10-
speed transmission benchmarking performed by Argonne.\1022\ The 
introduction of the skip-shift logic impacted effectiveness and reduced 
the number of shifts by 23 percent for the 10-speed automatic 
transmission over the UDDS cycle.\1023\
---------------------------------------------------------------------------

    \1022\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford 
F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812 
520.
    \1023\ See FRM ANL Model Documentation file at Paragraph 
4.4.5.5. This update reduced the number of shift events from 231 to 
178.
---------------------------------------------------------------------------

    In the NPRM analysis, transmission gear spans increased as the 
number of

[[Page 24465]]

gears increased.\1024\ However, to address further the comments related 
to optimization, the gear span of the AT10L3 was increased over the 
AT10L2, based on gear span data for the Honda 2018 10-speed 
transmission.\1025\ The AT10L3 span was increased to 10.10 in the final 
rule analysis from 7.34 in the NPRM analysis. However, the efficiency 
map for the AT10L3 remained the same for the final rule analysis.\1026\
---------------------------------------------------------------------------

    \1024\ See FRM ANL Model Documentation file at 5.3.2.1.
    \1025\ Sugino, S., SAE Internation Presentation., ``ALL-NEW 
HONDA 10-SPEED FWD TRANSMISSION.'' November 2017. ``2018 Honda 
Odyssey Press Kit--Overview.'' internet: Honda News, https://hondanews.com/en-US/releases/2018-honda-odyssey-press-kit-overview. 
Last accessed October 8, 2019.
    \1026\ See FRM ANL Model Documentation file at 5.3.4.1.
---------------------------------------------------------------------------

    Finally, in the agencies' review of NPRM model inputs, a weight 
discrepancy for the AT10 transmissions was identified. The weight 
assigned to the AT10 transmission in the NPRM analysis was too high. 
The weights were corrected for the final rule analysis. The AT10 
transmission weights were reduced by 20-45 kg, depending upon vehicle 
type.\1027\
---------------------------------------------------------------------------

    \1027\ See FRIA VI.C.2.d.2.
---------------------------------------------------------------------------

    The AT effectiveness values used for the final rule analysis can be 
seen in Figure VI-25. For automatic transmission technologies, the 
effectiveness improvement range is relative to a 5-speed automatic 
transmission. The new effectiveness values are a result of the 
aforementioned changes implemented to address comments. To summarize, 
the changes included an adjustment to the modeled effectiveness of the 
AT6L2, the use of skip-shift logic on the 10-speed transmissions, and 
the increase of the AT10L2 gear span.
    Figure VI-25 shows the automatic transmission's effectiveness 
increases progressively in a logical order and behaves in an expected 
manner. Gains in effectiveness can be observed increasing as gear count 
increases, and as HEG levels increase. The effects of diminishing 
returns can be observed as gear count reaches higher levels, and 
effectiveness effects for increased gear count are reduced. This agrees 
with observed data reported by the NAS and industry 
stakeholders.1028 1029
---------------------------------------------------------------------------

    \1028\ 2015 NAS Report, at 175.
    \1029\ Greimel, H., ``ZF CEO--We're not chasing 10-speeds,'' 
Automotive News, November 23, 2014, http://www.autonews.com/article/20141123/OEM10/311249990/zf-ceo:-were-not-chasing-10-speeds.
---------------------------------------------------------------------------

(b) Continuously Variable Transmissions
    For CVTs, the agencies also identified a discrepancy with the NPRM 
CVT weights. The weight assigned to the CVT class during the NPRM 
analysis was incorrect. Corrected values were assigned for the final 
rule analysis. The CVT weights were reduced by 9-10 kg based on vehicle 
type.\1030\
---------------------------------------------------------------------------

    \1030\ See FRIA VI.C.2.d.2.
---------------------------------------------------------------------------

    The CVT effectiveness values used for the final rule analysis can 
be seen in Figure VI-26, shown as an effectiveness improvement over a 
5-speed automatic transmission. The effectiveness values were not 
changed significantly from the values used in the NPRM analysis.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.194

(c) Dual Clutch Transmissions
    The DCT effectiveness values used for the final rule analysis can 
be seen in Figure VI-27, shown as an effectiveness improvement over a 
5-speed automatic transmission. The effectiveness values were not 
changed significantly from the values used in the NPRM analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.195


[[Page 24466]]


(d) Manual Transmission
    The MT effectiveness values used for the final rule analysis can be 
seen in Figure VI-28, shown as an effectiveness improvement over a 5-
speed manual transmission. The effectiveness values were not changed 
significantly from the values used in the NPRM analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.196

e) Transmission Costs
    For the NPRM, the transmission technology costs used as inputs for 
the CAFE model were retail price equivalent costs with learning curves 
applied. For a complete discussion on how the retail price equivalent 
and learning effects were applied to direct manufacturing costs see 
Section VI.B.4.b), Indirect Costs, and Section VI.B.4.d), Cost 
Learning. The direct manufacturing costs for the transmission 
technologies used in the NPRM were derived from technical sources and 
manufacturer's CBI.\1031\
---------------------------------------------------------------------------

    \1031\ See PRIA Section 6.3.7.3.
---------------------------------------------------------------------------

    Table VI-80 below shows the relative costs of the transmissions 
used in the NPRM analysis including learning and retail price 
equivalent.
[GRAPHIC] [TIFF OMITTED] TR30AP20.197

BILLING CODE 4910-59-C
(1) Automatic Transmissions
    Several comments were received on technology costs, or cost 
effectiveness. Meszler Engineering Services noted that ``AT10L2 (level 
2 ten-speed automatic) transmission technology is another example of an 
end-of-path technology with very poor cost effectiveness relative to 
other transmission options.'' \1032\ A cost analysis by ICCT also 
showed relative costs of transmission technologies may not be in line 
with the modeled effectiveness.\1033\
---------------------------------------------------------------------------

    \1032\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723, at 33.
    \1033\ Comments from International Council on Clean 
Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741, 
at I-64.
---------------------------------------------------------------------------

    The agencies conducted a review of transmission costs in response 
to the comments. For the final rule analysis, adjustments were made to 
costs of the AT6L2, AT7L2, AT9L2, AT10L2, and the AT10L3. The costs 
were adjusted based on reviewing the recommended relative costs 
discussed in the NAS 2015 report. Table VI-81 shows the cost for the 
automatic transmissions in the final rule analysis.
    The direct manufacturing cost (DMC) estimate for the AT6 is drawn 
from Table 5.7 of the NAS report. The DMC estimate for the AT6L2 is 
based on the cost of the AT6 with HEG level 2 technology costs applied. 
This cost change is applied in accordance with the effectiveness 
adjustment made for the AT6L2.
    A DMC estimate for the AT7 was drawn from Table 5.9 of the NAS 
report and was based on the cost of a system already equipped with HEG 
technology. The DMC estimate was given in 2007 dollars and relative to 
an AT5/AT4. The new DMC replaces the DMC from the NPRM, which did not 
account for the HEG technology.
    The DMC for the AT9 technology was drawn from Table 8A.2a of the 
NAS (2015) report and per the NPRM description of the technology made 
relative to the AT8L2. The AT9 is assumed to have at least the level 2 
HEG technology applied. The NPRM analysis assumed the AT9 cost was only 
relative

[[Page 24467]]

to the AT8 and did not account for the cost of the HEG technology.
    The DMC for the AT10 technologies was drawn from Table 8A.2a of the 
NAS report and per the NPRM description of the technology made relative 
to the AT8L2. The AT10L2 is assumed to have at least the level 2 HEG 
technology applied. The AT10L3 has the HEG3 technology applied. The 
NPRM analysis assumed the AT10 costs were only relative to the AT8 and 
did not account for the cost of the HEG technology.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.198

(2) Continuously Variable Transmissions
    No adjustments were made to the NPRM costs of the CVT technologies 
for the final rule analysis. Table VI-82 shows the cost for the CVTs in 
the final rule analysis.

[[Page 24468]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.199

(3) Dual Clutch Transmissions
    The agencies received one comment on cost learning over time for 
DCT technologies. Roush Industries ``believes that the [actual] 
learning factors for such systems are significantly better than those 
estimated by either the 2018 PRIA or the 2016 Draft TAR.'' Roush stated 
that ``eight-speed DCTs (DCT8) are currently in production (MY2018), 
with quantities increasing significantly,''\1034\ but provided no 
specific supporting data.
---------------------------------------------------------------------------

    \1034\ Comments from Roush Industries, Attachment 1, NPRM Docket 
No. NHTSA-2018-0067-11984, at 14-15.
---------------------------------------------------------------------------

    The current learning curve for the DCT technologies was established 
based on recommendations from the NAS 2015 report and on CBI data 
collected from manufacturers and suppliers. Since Roush did not supply 
any data to support its comment, the agencies decided it was reasonable 
to make no change to the DCT learning curve for the final rule 
analysis. Table VI-83 shows the cost for the DCTs in the final rule 
analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.200

(4) Manual Transmissions
    No adjustments were made to the NPRM costs of the manual 
transmission technologies for the final rule analysis. Table VI-84 
shows the cost for the MTs in the final rule analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.201


[[Page 24469]]


BILLING CODE 4910-59-C
3. Electric Paths
    The electric paths include a large set of technologies that share 
the common element of using electrical power for certain vehicle 
functions that were traditionally powered mechanically by engine power. 
Electrification technologies thus can range from electrification of 
specific accessories (for example, electric power steering to reduce 
engine loads by eliminating parasitic loss) to electrification of the 
entire powertrain (as in the case of a battery electric vehicle).
    Electrified vehicles are considered, for this analysis, to mean 
vehicles with a fully or partly electrified powertrain. These include 
several electrified vehicle categories, including: Battery electric 
vehicles (BEVs), which have an all-electric powertrain and use only 
batteries for propulsion energy; plug-in hybrid electric vehicles 
(PHEVs), which have a primarily electric powertrain and use a 
combination of batteries and an engine for propulsion energy; and 
hybrid electric vehicles (HEVs), which use electrical components and a 
battery to manage power flows and assist the engine for improved 
efficiency and/or performance. HEVs are further divided into strong 
hybrids (including P2 and power-split hybrids) that provide strong 
electrical assist and in many cases, can support a limited amount of 
all-electric propulsion, and mild hybrids (such as belt integrated 
starter generator (BISG) hybrids, crankshaft integrated starter 
generator (CISG) hybrids, and 48V mild hybrids) that typically provide 
only engine on/off with minimum electrical assist.
    Fuel cell electric vehicles (FCEVs) are also another form of 
electrified vehicle having a fully electric powertrain, and are 
distinguished by the use of a fuel cell system rather than grid power 
as the primary energy source.
    The factors that influence the cost and effectiveness of 
electrification technologies are their components. These include: 
Energy storage components such as battery packs; propulsion components 
such as electric motors; and power electronics components, such as 
inverters and controllers, that process and route electric power 
between the energy storage and propulsion components. For the purpose 
of this analysis, these components are divided into battery components 
and non-battery components.
    Battery components strongly influence the cost of electrified 
vehicles.\1035\ Because developments in battery technology may apply to 
more than one category of electrified vehicles, they are discussed 
collectively in Section VI.C.3.e). That section details battery-related 
topics that directly affect the specification and costing of batteries 
for all types of electrified vehicles considered in this analysis.
---------------------------------------------------------------------------

    \1035\ Battery costs are not necessarily a strong influence on 
fuel Cell Electric Vehicles, where the cost of the fuel cell 
technology has a larger influence.
---------------------------------------------------------------------------

    Non-battery components also have an influence on both the cost and 
effectiveness of electrified vehicles. The selection and configuration 
of non-battery technologies distinguish the different architecture 
among electrified vehicles. Non-battery components largely consist of 
propulsion components and power electronics.
    Propulsion components typically include one or more electric 
machines (an umbrella term that includes what are commonly known as 
motors, generators, and motor/generators). Depending on how they are 
employed in the design of a vehicle, electric machines commonly act as 
motors to provide propulsion, and/or act as generators to enable 
regenerative braking and conversion of mechanical energy to electrical 
energy for storage in the battery.
    ``Power electronics'' refers to the various components that control 
or route power between the battery system and the propulsion 
components, and includes components such as: Motor controllers, which 
issue complex commands to control torque and speed of the propulsion 
components precisely; inverters and rectifiers, which convert and 
manage DC and AC power flows between the battery and the propulsion 
components; onboard battery chargers, for charging the BEV or PHEV 
battery from AC line power; and DC-to-DC converters that are sometimes 
needed to allow DC components of different voltages to work together.
    Onboard chargers are charging devices permanently installed in 
electrified vehicles to allow charging from grid electrical power. 
Onboard chargers travel with the vehicle and are distinct from 
stationary charging equipment. Level 1 charging refers to charging 
powered by a standard household 110-120V AC power outlet. Level 2 
charging refers to charging at 220-240V AC power.
    The agencies included a more extensive overview of charging 
technology and the state of charging infrastructure in the NPRM and 
PRIA, however, this was purely qualitative because charging was not 
accounted for in any respect in the NPRM analysis. The Alliance 
commented that ``[w]hile the costs of installing chargers and charger 
convenience were not taken into account within the Volpe model . . . 
these factors will continue to have an impact on the overall 
penetration of electrification technologies that the market will be 
willing to accept.'' \1036\ In contrast, the National Coalition for 
Advanced Transportation (NCAT) commented that the qualitative 
discussion overstated the risks and understated the benefits of 
electric vehicle charging.\1037\ Specifically, NCAT took issue with the 
characterization of potential risks of charging to the electric grid, 
stating that ``the PRIA's focus on worst case hypotheticals does not 
reflect the current capabilities of the grid, nor the dynamic nature of 
EV charging to mitigate any potential negative impacts. In both in the 
short-term and long-term, the impact of EVs with respect to the 
electric grid would have a net-positive impact to society, including 
the EV owners and utility customers broadly.'' NCAT also commented that 
``[w]hile substantial investments in EV infrastructure have and will be 
made, the costs and benefits to consumers must be put into the 
appropriate context.'' NCAT cited two studies for the proposition that 
the average lifetime distribution electric vehicle infrastructure 
impact is about $80-$90 per electric vehicle sold, with the adoption of 
time of use rates and assuming a diversity of charging rates. NCAT also 
cited the California Public Utilities Commission 2016-2017 Electric 
Vehicle Load Research Report in support of their statement that the 
additional service and distribution system upgrades due to additional 
plug-in electric vehicle load is minimal, as ``of the approximately 
275,000 [electric] vehicles estimated to be on the road as of October 
2017 in the service areas of California's three investor-owned 
utilities, only 460, or 0.16 percent required a service line or 
distribution system upgrade solely to support the plug-in electric 
vehicle load at their residential charging location.''\1038\
---------------------------------------------------------------------------

    \1036\ NHTSA-2018-0067-12073.
    \1037\ NHTSA-2018-0067-11969.
    \1038\ Citing Joint IOU Electric Vehicle Load Research Report 
(December 29, 2017), pp. 1-2, 12, available at http://www.cpuc.ca.gov/zev/ (2016-2017 Load Research Report).
---------------------------------------------------------------------------

    The agencies agree that adding electric vehicle infrastructure will 
require additional costs, and information about what that cost is and 
how it can or should be accounted for

[[Page 24470]]

in the analysis is helpful for commenters to submit in order to put 
those considerations in the appropriate context. For this final rule, 
the agencies did not incorporate any costs related to electric vehicle 
charging infrastructure in the technology compliance analysis because 
those costs are separate from the costs that manufacturers and 
consumers would directly incur from a manufacturer transitioning part 
of their fleet to plug-in electric vehicles and consumers paying for 
those vehicles, even though local electric ratepayers will in all 
likelihood pay higher rates to upgrade local power grids to accommodate 
any widespread adoption of electrified vehicles. Accordingly, this 
means that the actual costs associated with electrified vehicles have 
been underestimated for the final rule analysis. The agencies did 
refine the estimates for the value of refueling time for electric 
vehicles, and that topic is discussed in Section VI.D.1.b)(11)(b). The 
agencies will continue to explore whether and how charging 
infrastructure should be incorporated into the analysis for future 
actions.
    The following sections discuss vehicle electrification issues that 
were accounted for in the analysis, including the agencies' 
characterizations of electric vehicle technology, additional electric 
vehicle configurations added for the final rule analysis per 
commenters' requests, and the sources and methods used to develop 
battery and non-battery components, which were also refined for this 
final rule.
a) Electrification Modeling in the CAFE Model
    A set of technologies was chosen to represent the spectrum of 
electrification methods observed in the baseline fleet and that the 
agencies believed could be applied to vehicles in the rulemaking 
timeframe. Each technology was placed in a specific electrification 
pathway, grouping and defining the progression of related technologies. 
In the NPRM analysis, a total of eleven electrification technologies 
were contained in four electrification pathways. In consideration of 
comments received, the electrification technologies and associated 
pathways were modified for the final rule analysis, resulting in a 
total of eighteen variants of electrification technologies. Each of 
these NPRM and final rule technologies, and the electrification 
pathways they belong to, are detailed below. Operational modes of 
electrified vehicles are further described in the Argonne Model 
Documentation for the final rule.
(1) Electrification Technologies
(a) Electric Improvements
    The electrification of power steering (EPS) and other accessories 
(IACC) have the potential of reducing fuel consumption by facilitating 
power-saving control strategies that avoid parasitic loss of engine 
power. These accessories traditionally are directly coupled to and 
driven by the conventional combustion engine; any time the engine is 
running some energy is continuously consumed by each accessory, even 
when it is not needed. By decoupling these accessories from the engine 
and instead driving them ``on-demand'' with electric motors, a more 
energy-efficient control strategy can be employed to reduce fuel 
consumption. EPS and IACC are discussed in detail in Section VI.C.7, 
Other Vehicle Technologies.
(b) Micro Hybrid
    12-volt stop-start (SS12V), sometimes referred to as start-stop, 
idle-stop or 12-volt micro hybrid, is the most basic hybrid system that 
facilitates idle-stop capability. In this system, the integrated 
starter generator is coupled to the internal combustion (IC) engine. 
When the vehicle comes to an idle-stop the IC engine completely shuts 
off and, with the help of 12-volt battery, the engine cranks and starts 
again in response to throttle to move the vehicle, or release of the 
brake pedal. The 12-volt battery used for the start-stop system is an 
improved unit capable of higher power, increased life cycle, and 
capable of minimizing voltage drop on restart. This technology is 
beneficial to reduce fuel consumption and emissions when the vehicle 
frequently stops, such as in city driving conditions or in stop and go 
traffic, and can be applied to all vehicle technology classes.
    (c) Mild Hybrids
    The belt integrated starter generator (BISG) and crank integrated 
starter generator (CISG), sometimes referred to as mild hybrid systems, 
provide idle-stop capability and use a higher voltage battery with 
increased energy capacity over typical automotive batteries. The higher 
voltage allows the use of a smaller, more powerful and efficient 
electric motor/generator, which replaces the standard alternator. In 
BISG systems, the motor/generator is coupled to the engine via belt 
(similar to a standard alternator), while the CISG integrates it to the 
crankshaft between the engine and transmission; both of these systems 
allow the engine to be automatically turned off as soon as the vehicle 
comes to a full stop. In addition, these motor/generators can recover 
braking energy while the vehicle slows down (regenerative braking) and 
in turn can propel the vehicle at the beginning of launch, allowing the 
engine to be restarted later. Some limited electric assist is also 
provided during acceleration to improve engine efficiency. The CISG 
system has a higher efficiency, but also higher cost than the BISG.
    The agencies received limited high-level comments on CISG systems, 
with CARB stating that CISG systems are generally considered more 
capable and more efficient relative to BISG systems because they do not 
have the same belt-related constraints including maximum torque 
limitations, load restrictions on the front crank to avoid uneven 
crankshaft bearing wear, and mechanical energy transfer losses.\1039\ 
CARB also noted that the decision to implement a CISG system is 
typically made early in the design process because doing so often 
requires an engine block casting change. CARB stated that the current 
high costs and larger dimensions, compared to BISGs, will likely delay 
major market penetration of CISG systems until beyond the MY 2025 
timeframe.
---------------------------------------------------------------------------

    \1039\ Roush Industries on behalf of California Air Resources 
Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
0067-11984, at 15.
---------------------------------------------------------------------------

    For the final rule analysis, the agencies did not include CISG 
systems. The effectiveness of CISG systems were similar to the BISG, 
and the high cost of the CISG caused it to be applied infrequently. 
Other packaging and integration issues make it difficult for most 
vehicles to adopt CISG technology. Typically, a manufacturer would have 
to modify the flywheel housing to allow the installation of an electric 
motor, which must also fit where the system is mounted between the 
transmission and the engine block. Space in that part of the vehicle 
also comes at a premium because other components such as exhaust 
systems and piping systems must also be housed in the same area. In the 
final rule analysis, all vehicles previously considered to possess CISG 
technology were instead assigned a BISG system.
    (d) Strong Hybrids
    A hybrid vehicle is a vehicle that combines two or more sources of 
propulsion energy, where one uses a consumable fuel (like gasoline), 
and one is rechargeable (during operation, or by another energy 
source). Hybrids reduce fuel consumption through three major 
mechanisms, including (1) potential engine downsizing, (2) optimizing 
the performance of the engine to operate at

[[Page 24471]]

the most efficient operating point and under some conditions storing 
excess energy such as by charging the battery, and (3) capturing energy 
during braking and some decelerations that might otherwise be lost to 
the braking system and using the stored energy to provide launch 
assist, coasting, and propulsion during stop and go traffic conditions. 
The effectiveness of the hybrid systems depends on how the above 
factors are balanced, taking into account complementary equipment and 
vehicle application. For some performance vehicles, the hybrid 
technologies are used for performance improvement without any engine 
downsizing.
    The NPRM analysis evaluated the following strong hybrid vehicles: 
Hybrids with ``P2'' parallel drivetrain architecture (SHEVP2),\1040\ 
and hybrids with power-split architecture (SHEVPS). The parallel hybrid 
drivetrain, although enhanced by the electric portion, remains 
fundamentally similar to a conventional powertrain. In contrast, the 
power-split hybrid drivetrain is novel and considerably different than 
a conventional powertrain. Although these hybrid architectures are 
quite different, both types provide start-stop or idle-stop 
functionality, regenerative braking capability, and vehicle launch 
assist. A SHEVPS has a higher potential for fuel economy improvement 
than a SHEVP2, although its cost is also higher.
---------------------------------------------------------------------------

    \1040\ Depending on the location of electric machine (motor with 
or without inverter), the parallel hybrid technologies are 
classified as P0-motor located at the primary side of the engine, 
P1-motor located at the flywheel side of the engine, P2-motor 
located between engine and transmission, P3-motor located at the 
transmission output, and P4-motor located on the axle.
---------------------------------------------------------------------------

    Power-split hybrid (SHEVPS) is a hybrid electric drive system that 
replaces the traditional transmission with a single planetary gear set 
(the power-split device) and a motor/generator. This motor/generator 
uses the engine either to charge the battery or to supply additional 
power to the drive motor. A second, more powerful motor/generator is 
permanently connected to the vehicle's final drive and always turns 
with the wheels. The planetary gear splits engine power between the 
first motor/generator and the drive motor either to charge the battery 
or to supply power to the wheels. During vehicle launch, or when the 
battery state of charge (SOC) is high, the engine, which is not as 
efficient as the electric drive, is turned off and the electric machine 
propels the vehicle. During normal driving, the engine output is used 
both to propel the vehicle and to generate electricity. The electricity 
generated can be stored in the battery and/or used to drive the 
electric machine. During heavy acceleration, both the engine and 
electric machine (by consuming battery energy) work together to propel 
the vehicle. When braking, the electric machine acts as a generator to 
convert the kinetic energy of the vehicle into electricity to charge 
the battery.
    The Autonomie simulations assumed all SHEVPS' used an Atkinson 
cycle engine (Eng26). Therefore, all vehicles equipped with SHEVPS 
technology in the CAFE model simulations were assumed to have Atkinson 
cycle engines. This Atkinson cycle engine with high compression ratio 
is optimized for efficiency, rather than performance. Accordingly, 
SHEVPS technology as modeled in this analysis was not suitable for 
large vehicles that must handle high loads.\1041\ Further discussion of 
Atkinson engines and their capabilities is discussed in Section VI.C.1 
Engine Paths.
---------------------------------------------------------------------------

    \1041\ Kapadia, J., Kok, D., Jennings, M., Kuang, M. et al., 
``Powersplit or Parallel--Selecting the Right Hybrid Architecture,'' 
SAE Int. J. Alt. Power. 6(1):68-76, 2017, https://doi.org/10.4271/2017-01-1154.
---------------------------------------------------------------------------

    P2 parallel hybrids (SHEVP2) are a type of hybrid vehicle that uses 
a transmission-integrated electric motor placed between the engine and 
a gearbox or CVT, with a clutch that allows decoupling of the motor/
transmission from the engine. Although similar to the configuration of 
the CISG system discussed previously, a P2 hybrid would typically be 
equipped with a larger electric machine and battery in comparison to 
the CISG. Disengaging the clutch allows all-electric operation and more 
efficient brake-energy recovery. Engaging the clutch allows efficient 
coupling of the engine and electric motor and, when combined with a 
transmission, reduces gear-train losses relative to power-split or 2-
mode hybrid systems. P2 hybrid systems typically rely on the internal 
combustion engine to deliver high, sustained power levels. Only low and 
medium power demands are allowed for electric-only mode.
    In the NPRM CAFE modeling, the SHEVP2 system represented a hybrid 
system paired with an existing engine on a given vehicle, while the 
SHEVPS removed and replaced the previous engine with an Atkinson cycle 
engine. The agencies explained that while many vehicles may use HCR1 
engines as part of a hybrid powertrain, HCR1 engines may not be 
suitable for some vehicles, such as high performance vehicles or 
vehicles designed to carry or tow large loads (this is further 
discussed in Section VI.C.1, Engine Paths). Many manufacturers may 
prefer turbocharged engines (with high specific power output) for P2 
hybrid systems, in order to maintain performance. Accordingly, in the 
NPRM analysis, to satisfy power demands, many SHEVP2 systems were 
paired with non-HCR powertrains.
    ICCT and Meszler Engineering Services commented that as a result of 
NPRM CAFE model constraints, low-cost, HCR engines were too 
infrequently paired with SHEVP2 technology. These commenters claimed 
that frequent pairing of SHEVP2 with downsized turbocharged engines 
resulted in higher cost and lower effectiveness for these strong 
hybrids.1042 1043
---------------------------------------------------------------------------

    \1042\ Meszler Engineering Services, Attachment 2, Docket No. 
NHTSA-2018-0067-11723, at 15.
    \1043\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-25.
---------------------------------------------------------------------------

    In consideration of these comments, the final rule analysis 
includes additional strong hybrids (P2HCR0, P2HCR1, and P2HCR2\1044\) 
that use HCR engines in a P2 parallel hybrid system. The SHEVP2 
technology allows the engine type to be inherited from the outgoing 
engine; this is unchanged from the NPRM and provides a good solution 
for vehicles that need to undergo hybridization but require other 
engine technologies (such as turbocharging) to meet performance 
requirements. In addition, this final rule analysis allows any 
conventional engine technology to go to P2HCR strong hybrid technology 
within the set performance requirements. This is further discussed in 
the Section VI.C.3.c), Electrification Adoption Features.
---------------------------------------------------------------------------

    \1044\ P2HCR2 was included in simulations used for sensitivity 
studies, but was excluded in the central analysis simulations for 
reasons surrounding the HCR2 engine, as discussed in Section VI.C.1.
---------------------------------------------------------------------------

(e) Plug-In Hybrids
    Plug-in hybrid electric vehicles (PHEV) are hybrid electric 
vehicles with the means to charge their battery packs from an outside 
source of electricity (usually the electric grid). These vehicles have 
larger battery packs with more energy storage and a greater capability 
to be discharged than other non-plug-in hybrid electric vehicles. PHEVs 
also generally use a control system that allows the battery pack to be 
substantially depleted under electric-only or blended mechanical/
electric operation and batteries that can be cycled in charge-
sustaining operation at a lower state of charge than is typical of 
other hybrid electric vehicles. These vehicles generally have a greater 
all-electric range than the typical SHEVs discussed above. In the NPRM 
analysis,

[[Page 24472]]

PHEVs with two all-electric ranges--a 30 mile and a 50 mile all-
electric range (AER)--were included as technologies that vehicles could 
adopt. The PHEV30 represented a ``blended-type'' plug-in hybrid, which 
can operate in all-electric (engine off) mode only at light loads and 
low speeds, and must blend electric machine and engine power together 
to propel the vehicle at medium or high loads and speeds. The PHEV50 
represented an extended range electric vehicle (EREV), which is capable 
of travelling in all-electric mode even at higher speeds and loads.
    Unlike other alternative fuel systems that require specific 
infrastructure for refueling or recharging (e.g., hydrogen vehicles or 
rapidly charged battery electric vehicles), PHEV batteries can be 
charged using existing infrastructure, although widespread adoption may 
require upgrades to electrical power distribution systems.\1045\ PHEVs 
are considerably more expensive than conventional vehicles and more 
expensive than SHEVPS technologies because of larger battery packs and 
charging systems capable of connecting to the electric grid.
---------------------------------------------------------------------------

    \1045\ See above for a discussion of electrical vehicle 
infrastructure.
---------------------------------------------------------------------------

    Commenters, such as CARB, stated that in the NPRM analysis the PHEV 
motors were oversized and overpowered, and that model-built PHEV30s 
have excessive battery pack size and electric range when compared to 
actual production vehicles.\1046\ In response to such comments, the 
agencies, in collaboration with Argonne, conducted further market study 
to confirm CARB's observations and determined that replacing PHEV30 
(with a nominal 30 mile AER) with PHEV20 (with a nominal 20 mile AER) 
would more closely characterize the PHEVs actually in production.\1047\ 
The agencies therefore elected to replace PHEV30 with PHEV20 in the 
final rule.
---------------------------------------------------------------------------

    \1046\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 150, 153.
    \1047\ ``ANL response on NPRM comments (PHEV sizing)- 
181112.pptx,'' available in Docket No. NHTSA-2018-0067.
---------------------------------------------------------------------------

    The final rule also includes four additional types of plug-in 
hybrids; two additional plug-in hybrids were added to allow the use of 
turbocharged engines (PHEV20T, PHEV50T), and two additional plug-in 
hybrids were added to provide maximum efficiency by utilizing an 
Atkinson cycle engine (PHEV20H, PHEV50H).
    In practice, many PHEVs recently introduced in the marketplace use 
turbocharged engines in the PHEV system, and this is particularly true 
for PHEVs produced by European manufacturers and for other PHEV 
performance vehicle applications. However, the NPRM Autonomie 
simulations (and thus all the CAFE model simulations) assumed all PHEVs 
used a naturally aspirated, Atkinson cycle engine. The agencies 
determined through continued marketplace observation that PHEV vehicles 
should indeed be allowed to adopt or retain turbocharged engines. Also, 
BorgWarner commented that modeling of PHEVs should include turbocharged 
engines, since these engines can be downsized to reduce vehicle mass 
and fit into smaller engine compartments, and offer efficiency and 
performance advantages especially when paired with a higher expansion 
ratio.\1048\ Thus, in addition to the PHEV20 and PHEV30, the final rule 
analysis included PHEV20T and PHEV50T variations which are, 
respectively, 20 and 50 mile all electric range PHEVs with turbocharged 
engines.
---------------------------------------------------------------------------

    \1048\ BorgWarner, Attachment 2, Docket No. NHTSA-2018-0067-
11873, at 150,153.
---------------------------------------------------------------------------

    This final rule also added PHEV20H and PHEV50H, although 
effectively these are not used by the model simulations. These plug-in 
types represent 20 and 50 mile all electric range plug-in hybrids that 
use particularly efficient high-compression, Atkinson cycle engines. 
These were added with the intent to provide PHEVs with a maximum level 
of fuel economy at a lower cost. However, they proved to be too similar 
to existing plug-in technology choices and were thus assigned identical 
characteristics as the PHEV20 and PHEV50. In this final rule analysis, 
PHEV20 and PHEV50 sizing were updated and so the similarities in 
performance between different engines converged. For further discussion 
on PHEV sizing, see Section VI.C.3.d), Electrification Effectiveness 
Modeling and resulting Effectiveness values.\1049\ The PHEV20H and 
PHEV50H technologies are still considered by the CAFE model but they 
remain as ``placeholders'' for potential incorporation in future 
analyses.
---------------------------------------------------------------------------

    \1049\ This final rule analysis used Atkinson Engine for PHEVPS 
electrified vehicles. The components such as electric motor and 
engine power in these hybrid systems were sized in ways to meet 
vehicle class performance characteristics and efficiency. And after 
these vehicle components were sized, the Atkinson engines in these 
vehicles were operating in similar efficiency as HCR engines as the 
full vehicle modeling and simulation. As discussed in PO 06 C.1.c.1 
Non-HEV Atkinson Engine Modes, power-split hybrid-based Atkinson 
engines attempt to operate in the most efficient regions while using 
electric motors to meet deficiencies in performance. And so, PHEV20H 
and PHEV50H HCR engines compared to PHEV20 and PHEV50 Atkinsons 
engines would have be sized to operate in the most efficiency 
regions and the thermal efficiency between these two set of 
combinations would have had similar efficiency for this analysis.
---------------------------------------------------------------------------

(f) Battery Electric Vehicles
    Electric vehicles (EVs), or battery electric vehicles (BEVs) are 
equipped with all-electric drive and with systems powered by energy-
optimized batteries charged primarily from grid electricity. The range 
of a battery electric vehicle depends on the vehicle's class and the 
battery pack size. The NPRM analysis included BEVs with a range of 200 
miles.
    Following the NPRM, the agencies conducted continued market 
analysis of production BEVs, and observed a growing number of vehicles 
with nominal ranges above 200 miles. CARB also commented that certain 
BEVs modeled as BEV200 in the NPRM in fact had ``well over 200 miles of 
range.'' \1050\ The agencies thus concluded that a 300-mile-range 
BEV300 should be included in the final rule to represent better these 
higher-range electric vehicles as well as a potential future range 
alternative more comparable to IC engines. The agencies still believe 
that, in the rulemaking timeframe, BEV300 will be the most cost 
effective extended range BEVs that could be available for adoption. 
Longer-range electric vehicles could have been modeled in the analysis, 
but the compliance simulation would likely not have selected the 
longer-range vehicle if lower-range vehicles were still available. This 
is because the CAFE model only applies technologies until a 
manufacturer meets its CAFE or CO2 standard, and the BEV200 
and BEV300 vehicles operate functionally the same in helping a 
manufacturer towards meeting its compliance obligations. The only 
difference between these vehicles is cost. As discussed further in 
Section VI.C.3.c), the agencies used phase-in caps to control expected 
BEV200 and BEV300 penetration based on the current trend and future 
assumption that consumers will transition towards longer-range electric 
vehicles.
---------------------------------------------------------------------------

    \1050\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 147.
---------------------------------------------------------------------------

(g) Fuel Cell Vehicles
    Fuel cell electric vehicles (FCEVs or FCVs) utilize a full electric 
drive platform but consume hydrogen fuel to generate electricity in an 
onboard fuel cell. Fuel cells are electrochemical devices that directly 
convert reactants (hydrogen and oxygen via air) into electricity, with 
the potential of achieving more than twice the efficiency of 
conventional internal combustion engines. High pressure gaseous 
hydrogen storage tanks are used by most

[[Page 24473]]

automakers for FCEVs. These high-pressure tanks are similar to those 
used for compressed gas storage in more than 10 million CNG vehicles 
worldwide, except that they are designed to operate at a higher 
pressure (350 bar or 700 bar vs. 250 bar for CNG), and to contain the 
very small, and very flammable, gaseous hydrogen molecule. FCEVs are 
currently produced in limited numbers and are available in limited 
geographic areas.
(2) Electrification Pathways
    The electrification technologies described above were applied in 
the CAFE model through a number of technological pathways. Three main 
electrification technology pathways were modeled: The Electric 
Improvements Path, the Electrification Path, and the Hybrid/Electric 
Path. These three electrification pathways are evaluated in parallel by 
the CAFE model; the model can consider any of the three right away, and 
does not need to go ``through'' one pathway in order to begin 
evaluating another. Any superseded technology is also disabled whenever 
a succeeding technology is applied to a vehicle, even if a specific 
superseded technology was not previously utilized on that vehicle. As 
previously explained, this requirement exists so that the modeling 
system does not downgrade technologies during analysis.
    The Electrics Improvements Path defined in the NPRM and final rule 
is shown in Figure VI-29 below, which starts with EPS and progresses to 
IACC. While these two electrified-accessory technologies are mutually 
exclusive, either one can be modularly paired with any other 
technology, including those in the other electrification pathways.
[GRAPHIC] [TIFF OMITTED] TR30AP20.203

    The Electrification Path shown in Figure VI-29 allows a 
conventional powertrain to become a micro-hybrid with SS12V, or a mild 
hybrid with BISG, or CISG (which is no longer available for the final 
rule analysis, as discussed previously) technologies. All three of the 
Electrification Path technologies are mutually exclusive with respect 
to all conventional powertrain technologies, as well as technologies 
contained in the Hybrid/Electric path discussed below. The model first 
evaluates SS12V, and then progresses to BISG or CISG (NPRM-only). The 
conventional engine technology CONV is grayed out to indicate that the 
model uses information about the previous conventional (non-
electrified) powertrain to map properly to simulation results found in 
the vehicle simulation database. Although the adoption of these 
technologies will classify a vehicle as a micro/mild hybrid (MHEV) and 
no longer a conventional (CONV), the vehicle is allowed to retain the 
engine and transmission technologies possessed before entering the 
Electrification Path.
[GRAPHIC] [TIFF OMITTED] TR30AP20.204

    The Hybrid/Electric Pathways are shown in Figure VI-30. Both the 
NPRM and final rule Hybrid/Electric paths begin at the ``strong 
hybrid'' technology types, each of which is mutually exclusive of the 
others; once one is chosen, the other is eliminated from future 
selection for that vehicle. The paths then progress into plug-in 
hybrids and then culminate with the mutually exclusive battery electric 
vehicles or fuel cell vehicles. The additional final rule technologies 
described above can be found in the final rule Hybrid/Electric pathway 
on the right side of Figure VI-31, in comparison to the NPRM 
technologies shown on the left

[[Page 24474]]

side of the figure.\1051\ The hybrid/electric pathways contains 
multiple ``roots,'' or starting points, which force a vehicle to remain 
within the branches of a chosen root. For example, the final rule 
hybrid/electric pathway has three roots: SHEVP2, SHEVPS, and P2HCR0. If 
a vehicle uses SHEVPS, then SHEVP2 technology and the entire P2HCR0 
through PHEV50H branch will be disabled from further consideration. In 
other words, from one technology in the pathway, a vehicle can only 
move forward along any of the indicated arrows, and never in the 
reverse direction. Also, when using any technology in the Hybrid/
Electric pathway, with the exception of SHEVP2, all engine and 
transmission technologies as well as the Electrification Path 
technologies shown in Figure VI-31 are prohibited. SHEVP2 is an 
exception because it allows engine technologies previously held by the 
vehicle to be inherited into the parallel hybrid system.
---------------------------------------------------------------------------

    \1051\ Note that the NPRM Hybrid/Electric Path (left side of 
Figure I-3) refers to a portion of the path containing plug-in 
hybrids and electric vehicles as the ``Advanced Hybrid/Electric 
Path.'' For this discussion, we will simply refer to the entire 
collection of these technologies, including the ``Advanced'' 
technologies, as the ``Hybrid/Electric Path.''
[GRAPHIC] [TIFF OMITTED] TR30AP20.205

b) Electrification Analysis Fleet Assignments
    Since the 2012 rulemaking, manufacturers have implemented a number 
of powertrain electrification technologies, including 48V mild hybrid, 
strong HEV, PHEV, and BEV powertrains.1052 1053 For the NPRM 
analysis, the agencies identified the specific electrification 
technologies in each vehicle model in the MY 2016 analysis fleet, and 
used those technology levels as the starting point for the regulatory 
analysis. The agencies assigned electrification technology levels based 
on manufacturer-submitted CAFE compliance information, vehicle 
technical specifications released publicly by manufacturers, agency-
sponsored vehicle benchmarking studies, technical publications, and 
manufacturer CBI.\1054\ For the final rule analysis, the agencies used 
a similar process and data sources to identify the electrification 
technologies in the MY 2017 analysis fleet.\1055\
---------------------------------------------------------------------------

    \1052\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
    \1053\ FOTW #1108, Nov 18, 2019: Fuel Economy Guide Shows the 
Number of Conventional Gasoline Vehicle Models Achieving 45 miles 
per gallon or Greater is Increasing. DOE VTO. Available at https://www.energy.gov/eere/vehicles/articles/fotw-1108-november-18-2019-fuel-economy-guide-shows-number-conventional. Last accessed Nov 18, 
2019.
    \1054\ NPRM Market Data central analysis input file.
    \1055\ FRM Market Data central analysis input file.
---------------------------------------------------------------------------

    The agencies received comments regarding the application of 
electrification technologies in the MY 2016 analysis fleet. Commenters, 
such as the California Air Resources Board, stated the agencies 
mischaracterized some hybrid technologies, such as power-split and P2 
hybrid architectures.\1056\ Specifically CARB was concerned about the 
``misclassification of the 2016 Chevrolet Malibu Hybrid as having a P2 
hybrid,'' noting the Malibu shared many of its drivetrain components 
with the 2016 Chevy Volt, a vehicle classified as a power-split HEV.
---------------------------------------------------------------------------

    \1056\ Comments from CARB, Attachment 2, NHTSA Docket No. NHTSA-
2018-0067-11873, at 136.
---------------------------------------------------------------------------

    BorgWarner stated that the ``modeling should be inclusive of all 
approaches of PHEV and HEV and not be limited only to Atkinson Cycle 
engines,'' suggesting that it was appropriate for the NPRM analysis to 
include turbocharged engines in combination with PHEV and HEV 
technologies.\1057\
---------------------------------------------------------------------------

    \1057\ Comments from BorgWarner, Attachment 1, Appendix, NHTSA 
Docket No. NHTSA-2018-0067-11895, at 10.
---------------------------------------------------------------------------

    The agencies agree with the underlying issue identified by both 
CARB and BorgWarner's comments. In

[[Page 24475]]

both cases a limitation of modeling classification, and not a lack of 
academic understanding of HEV systems, is the crux of the issue. In the 
specific case of the 2016 Chevy Malibu, the electrical architecture is 
a power split, however, the vehicle uses a non-Atkinson, basic direct 
injection engine. These characteristics put the Malibu HEV in an 
overlap with the powertrain models used to represent HEV systems in the 
agencies' analysis. If the system had been classified as a PS HEV 
system in the analysis fleet, the engine would have incorrectly been 
modeled as an Atkinson engine, resulting in overestimation of the 
baseline system's level of efficiency and technology applied. The 
overestimation of the baseline fleet model would have limited the 
potential for the baseline system to improve over the timeframe of the 
analysis. With the system classified as the P2 HEV, the engine can be 
accurately modeled while still accounting for the benefits of an HEV 
system. This allowed the platform the full potential for technology and 
efficiency improvement in the analysis.
    The agencies considered the issues identified in comments and 
reviewed the MY 2017 analysis fleet information to determine what 
changes could improve the final rule analysis. The agencies determined 
that expanding the number of electrification technologies would address 
the CARB and BorgWarner comments, as well as the comments from others 
that are discussed in Section VI.C.3.a)(1) Electrification 
Technologies. The agencies increased the number of unique 
electrification technologies from twelve in the NPRM to eighteen for 
the final rule analysis. The expanded list enabled greater precision in 
the assignment of technologies to the MY 2017 analysis fleet, and 
enabled the agencies to characterize the electrification technologies 
found in the fleet accurately and realistically. The expanded list also 
provided more granularity for the application of technologies for the 
rulemaking analysis. Table VI-85 shows the full list of electrification 
technologies for the final rule analysis.
    This collection of technologies represents the best available 
information the agencies have, at the time of this action, regarding 
both currently available electrification technologies and 
electrification technologies that could be feasible for application to 
the U.S. fleet during the rulemaking timeframe. The agencies believe 
this effort has yielded the most accurate analysis fleet utilized for 
rulemakings to date.
    As discussed in the previous section and shown in Figure VI-29, 
Figure VI-30, and Figure VI-31, electrification may be added to 
vehicles as shown on the decision tree pathways. Further application of 
electrification technologies to vehicle platforms was dependent on 
electrification technology already present on vehicles in the MY 2017 
analysis fleet. Electrification may also be predicated on whether a 
vehicle has a dedicated platform that accommodates battery electric 
capability or whether a platform is designed (``package protected'') 
\1058\ to enable the addition of some form(s) of hybridization. The 
agencies' assessment of each existing platform's capability to adopt 
electrification technologies is identified in the CAFE model market 
data input file.\1059\
---------------------------------------------------------------------------

    \1058\ `Package Protected' is an automotive industry term used 
to describe the purposeful design of a vehicle to include space and 
weight allowances for future technology additions.
    \1059\ FRM Market Data central analysis input file.
---------------------------------------------------------------------------

c) Electrification Adoption Features
    In the NPRM and final rule analysis, electrification adoption 
features were applied in multiple ways. First, when an electrification 
technology is selected, a path logic is applied that dictates what 
other technologies are either superseded or mutually exclusive to the 
applied technology. For a detailed discussion of path logic for the 
final rule analysis, including technology supersession logic and 
technology mutual exclusivity logic, please see CAFE model 
documentation section. Second, application of the more advanced 
electrification technologies, such as the strong hybrids, plug-in 
hybrids, and full BEVs, result in major changes to the whole 
powertrain. The changes to the powertrain include substitution of 
transmission and engine technologies, and accordingly these 
technologies can only be applied at a vehicle redesign, as shown in 
Table VI-85 below. Finally, some of electrification technologies are 
restricted from application to certain vehicle classifications. These 
restrictions will be discussed under the specific technology sections.
    The fully-electric technologies, BEV technology and FCV technology, 
qualify as alternative fuel technologies. As a result, these 
technologies are not considered during portions of the agencies' 
analysis. Specifically, the exclusion of dedicated alternative fuel 
technology from NHTSA's analysis of potential fuel economy standards is 
a result of statutory obligations prescribed under EPCA/EISA.\1060\ 
However, NHTSA performed two fuel economy analyses, a standard-setting 
analysis that constrained the use of the technologies, and an 
unconstrained analysis that did not exclude the technologies, which 
provides an estimation of real-world environmental impacts used as 
inputs for the Environmental Impact Statement (EIS). The unconstrained 
analysis included the alternative fuel technologies, and used the 
adoption features for BEVs and FCVs discussed below. Further, for 
purposes of analyzing EPA's tailpipe CO2 emissions 
rulemaking pursuant to the Clean Air Act, consideration of these 
technologies is likewise unconstrained. For a detailed discussion of 
the analysis versions and statutory obligations please refer to Section 
VI.A Analytical Approach as Applied to Regulatory Alternatives, 
Overview of Methods and Section VI.A.4 Compliance Simulation.
---------------------------------------------------------------------------

    \1060\ 49 U.S.C. 32902(b)(1). A ``dedicated automobile'' is 
defined in 49 U.S.C. 32901 as ``an automobile that only operates on 
alternative fuel.''
---------------------------------------------------------------------------

    The exclusion of the BEV and FCV technology from the standard-
setting analysis resulted in a comment from ICCT. ICCT stated, ``the 
agencies prevented their fleet compliance model from allowing battery 
electric vehicles from being applied in their analysis of the Augural 
standards.'' \1061\ The agencies believe this reflects a 
misunderstanding of NHTSA's statutory obligation under EPCA/EISA and 
how the agencies ran the analysis. NHTSA did consider alternative 
fueled vehicles in the unconstrained analysis--but as discussed further 
in Section VIII, is prohibited from considering the availability of 
such technologies when setting maximum feasible standards.
---------------------------------------------------------------------------

    \1061\ Comments from ICCT, Attachment 3, Appendix, NPRM Docket 
No. NHTSA-2018-0067-11741, at 182.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24476]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.206

BILLING CODE 4910-59-C
(1) Micro and Mild Hybrid
    For the NPRM and final rule analysis, the only adoption features 
for the SS12V and BISG technologies were functions of path logic. The 
SS12V and BISG technologies were allowed for consideration in any 
existing vehicle configuration that did not already have a more 
advanced electrification technology applied. Per Table VI-85 above, the 
BISG technology was considered more advanced than the SS12V technology.
    Meszler Engineering commented that 48V batteries used in 
conjunction with 12 volt systems (what are referred to in the analysis 
as BISG systems) are one example of a ``bolt-on'' technology that can 
be added to a vehicle during a product refresh without causing

[[Page 24477]]

production problems or significantly increasing costs.\1062\ Meszler 
Engineering stated that 48V systems do not require reengineering of the 
engine and can be added at any time during a model's lifespan, as shown 
by key suppliers that are expanding production capacity to meet 
customer demand for the technology.\1063\ Meszler Engineering also 
pointed to examples of vehicles that utilize 48V systems, including 
high-volume non-luxury vehicles like the Ram pickup truck, Jeep 
Wrangler, and Ford F-150.\1064\
---------------------------------------------------------------------------

    \1062\ Comments by Meszler Engineering, Attachment 4 CAF[Eacute] 
Model Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-
11723, at 2-4. (citing A.K. Kumawat and A.K. Thakur, A Comprehensive 
Study of Automotive 48V Technology, SSRG International Journal of 
Mechanical Engineering (SSRG-IJME), Vol. 4 (5) (May 2017), available 
at: https://jalopnik.com/everything-you-need-to-know-about-the-upcoming-48-volt-1790364465 (last viewed 10/23/2018)).
    \1063\ Comments by Meszler Engineering, Attachment 4 CAFE Model 
Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-11723, 
at 2-4.
    \1064\ Comments by Meszler Engineering, Attachment 4 CAFE Model 
Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-11723, 
at 2-4.
---------------------------------------------------------------------------

    The agencies disagree with Meszler Engineering's assessment of 48V 
technology as a ``bolt-on'' technology. Although BISG systems represent 
a first step in vehicle electrification, and the number of components 
involved is fewer than most other types of hybrid systems, a BISG 
system still requires engineering and packaging of motors, cooling 
systems, additional wiring harnesses from the 48V battery pack to the 
motors, control systems, and other components incorporated into the 
front engine compartment. Further, the addition of a BISG system 
requires recalibration and validation of numerous engine performance 
parameters, including emissions controls, balancing torque supply to 
the transmission between the BISG system and engine, and noise-
vibration-harshness controls. In addition, the examples Meszler 
Engineering provided support the agencies' designation of SS12V and 
BISG systems as redesign technologies; the BISG system in the MY 2019 
Ram pickup and in the MY 2018 Jeep Wrangler were introduced during a 
product redesign and not during a mid-cycle product 
refresh.1065 1066 Although Ford has indicated that the F-150 
will include hybrid variants,\1067\ the agencies do not have 
information about specific plans for a 48V system on the F-150. In 
consideration of this information, the agencies maintained the redesign 
schedule for mild hybrids for the final rule analysis.
---------------------------------------------------------------------------

    \1065\ See, e.g., K.C. Colwell, The 2019 Ram 1500 eTorque Brings 
Some Hybrid Tech, If Little Performance Gain, to Pickups, Car and 
Driver (Mar. 14, 2019), available at: https://www.caranddriver.com/reviews/a22815325/2019-ram-1500-etorque-hybrid-pickup-drive/ (``Any 
2019 Ram 1500--the all-new one, not the Ram Classic that is just a 
continuation of the previous generation--can be equipped with a 
motor/generator attached to its engine's crankshaft via a belt that 
is capable of adding torque, cranking the engine in a stop/start 
event, or making electricity with regenerative braking.'').
    \1066\ See, e.g., Tony Quiroga, The 2018 Jeep Wrangler Hybrid 
Provides Effortless Thrust, Much Improved Fuel Economy, Car and 
Driver (Oct. 15, 2018), available at: https://www.caranddriver.com/reviews/a23746585/2018-jeep-wrangler-unlimited-suv-turbo-four-cylinder-hybrid/ (``Completely redesigned for 2018, the Wrangler is 
even more like a Power Wheels now that it's available with an 
electric motor.'').
    \1067\ ``Ford to Invest more than $1.45 Billion, Add 3,000 Jobs 
in SE Mich. Plants to Deliver New Pickups, SUVs, EVS, and AVS,'' 
Ford Media Center, 17 Dec 2019. https://media.ford.com/content/fordmedia/fna/us/en/news/2019/12/17/ford-invests-adds-jobs-southeast-michigan-plants.html.
---------------------------------------------------------------------------

(2) Strong Hybrids--SHEVP2, SHEVPS, P2HCR0, P2HCR1, P2HCR2
    NPRM adoption features applied to strong hybrid technologies 
included path logic, powertrain substitution, and vehicle class 
restrictions. For the NPRM analysis technologies on the Hybrid/Electric 
path (SHEVP2 and SHEVPS) were defined as stand-alone and mutually 
exclusive. When the modeling system applies one of those technologies, 
the other one is immediately disabled from future application. Once a 
strong hybrid technology is applied it also supersedes lower 
technologies on the electrification path, allowing future application 
of technology to consider only more advanced forms of electrification.
    In the NPRM when the SHEVP2 technology or the SHEVPS technology 
were applied, the transmission technology was superseded. Regardless of 
the transmission technology present when the technology was applied, 
the transmission technology was replaced by either the AT6 or DCT6. The 
specific transmission technology selected was based on choosing the 
best cost versus effectiveness.
    During the NPRM analysis when the SHEVP2 technology was selected 
the engine technology for the platform was maintained. However, the 
engine technology was locked at the current level and could not be 
changed. For the SHEVPS technology the existing engine was replaced 
with an Atkinson cycle engine (Eng26).
    The SHEVPS was also constrained from application to particular 
vehicle technology classes or vehicles with specific performance 
characteristics in the NPRM. Application of the power-split 
architecture was restricted from high performance vehicles and vehicles 
with a high towing capability requirements.\1068\ These constraints 
prevented application to the pick-up and performance pick-up class of 
vehicles. The constraints also prevented application to any platform 
with a base horsepower rating greater than 400 HP. Additional platforms 
determined to be purpose built as performance platforms were also 
restricted from receiving SHEVPS technology.
---------------------------------------------------------------------------

    \1068\ Kapadia, J., Kok, D., Jennings, M., Kuang, M. et al., 
``Powersplit or Parallel--Selecting the Right Hybrid Architecture,'' 
SAE Int. J. Alt. Power. 6(1):68-76, 2017, https://doi.org/10.4271/2017-01-1154.
---------------------------------------------------------------------------

    Comments from ICCT criticized the manner in which SHEVP2 technology 
was applied to a platform. ICCT stated ``the benefits of level-2 
transmission efficiency and TURBO2 over TURBO1 are removed when P2 
strong hybrid systems (SHEVP2) are selected on the electrification 
pathway.'' \1069\
---------------------------------------------------------------------------

    \1069\ Comments from ICCT, Attachment 3, 15 page summary and 
full comments appendix, NPRM Docket No. NHTSA-2018-0067-11741, at 
I25.
---------------------------------------------------------------------------

    Additional comments regarding the adoption features of the SHEVP2 
technology were received from Meszler Engineering and ICCT. Meszler 
argued that the locking of engine technologies when a manufacturer 
selects the SHEVP2 technology may preclude the selection of a more 
cost-effective engine technology.\1070\ This concern was echoed by 
ICCT, who also felt the engine technology lock-in artificially 
increased cost for effectiveness on the overall SHEVP2 technology 
packages.\1071\ Both commenters specifically wanted an option for a 
high compression ratio engine technology to be considered in place of 
any advanced engine technology carried into the SHEVP2 technology 
pathway.
---------------------------------------------------------------------------

    \1070\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723, at 15-16.
    \1071\ Comments from ICCT, Attachment 3, 15 page summary and 
full comments appendix, NPRM Docket No. NHTSA-2018-0067-11741, at 
I25-I26.
---------------------------------------------------------------------------

    The agencies agreed with the need for maintaining the benefits of a 
higher transmission technology, and for the final rule analysis a AT8L2 
transmission technology replaced the AT6 or DCT6 transmissions for all 
hybrid-electric technologies. The AT8L2 was selected as the optimal 
transmission technology point for HEV systems. The transmission 
technology point was selected based on observed diminishing returns for 
applying advanced transmission technologies to advanced engine/
powertrains.\1072\
---------------------------------------------------------------------------

    \1072\ 2015 NAS Report--The National Academy of Science, in 
their 2015 report, noted that ``as engines incorporate new 
technologies to improve fuel consumption, the benefits of increasing 
transmission ratios or switching to a CVT diminish.''

---------------------------------------------------------------------------

[[Page 24478]]

    The agencies also reconsidered engine options for SHEVP2 
technology, and other strong hybrid-electric technologies. The agencies 
agreed with Meszler and ICCT's observation and instituted new P2 engine 
technology options, as discussed above. For the final rule analysis, 
when a platform considered the SHEVP2 option, the platform also 
compared maintaining the current engine technology, or selecting an HCR 
technology. If the SHEVP2 system chooses to apply a HCR engine, the 
system diverts to the new electrification sub-path of technologies that 
includes the P2HCR0, P2HCR1, and P2HCR2.
    The P2HCR path introduced in the final rule analysis had similar 
constraints as the SHEVPS. Performance vehicles and vehicles with a 
high towing requirement were restricted from selection of the P2HCR 
technology. Restrictions that were applied used the same criteria 
described for the SHEVPS.
(3) Plug-In Hybrids--PHEV20/30, PHEV50, PHEV20T, PHEV50T, PHEV20H, 
PHEV50H
    The plug-in hybrid options in the NPRM included PHEV30 and PHEV50 
technologies. The plug-in technologies superseded the micro, mild, and 
strong hybrid electrification technologies and could only be replaced 
by full electric technologies. The path logic also allowed a PHEV30 to 
progress to a PHEV50.
    In the NPRM, when a platform progressed to the plug-in hybrid 
technologies the powertrain was automatically modified. The engine 
technology was replaced by a high compression ratio engine (Eng26) and 
the transmission was replaced by the AT6 or DCT6 technology.
    PHEV30 and PHEV50 were also constrained from application to 
vehicles with the potential for high towing demands.\1073\ This 
constraint was applied by restricting access to the pickup truck 
vehicle technology class. Additional specific vehicle platforms were 
restricted based on engineering judgment.
---------------------------------------------------------------------------

    \1073\ Power split or Parallel-selecting the Right Hybrid 
Architecture: SAE 2017-01-1154. = Kapadia, J., Kok, D., Jennings, 
M., Kuang, M. et al., ``Powersplit or Parallel--Selecting the Right 
Hybrid Architecture,'' SAE Int. J. Alt. Power. 6(1):68-76, 2017, 
https://doi.org/10.4271/2017-0-1154.
---------------------------------------------------------------------------

    Comments were received regarding the options for PHEV battery-
electric technology. The comments are presented and discussed in 
Section VI.C.3.e) Electrification Technologies above, and resulted in 
the creation of additional technology options for plug-in hybrids, as 
well as a modification of available ranges. Comments were also received 
regarding the engine and transmission options used in the 
electrification technologies, these comments are also presented and 
discussed above in Section VI.C.3.e) Electrification Technologies.
    For the final rule analysis, the plug-in hybrid options included 
PHEV20, PHEV50, PHEV20T, PHEV50T, PHEV20H, and PHEV50H. As with the 
NPRM, the plug-in technologies superseded the micro, mild, and strong 
hybrid technologies. For the final rule analysis, plug-in hybrid 
technologies were also mutually exclusive, and the PHEV20 technologies 
can progress to the PHEV50 technologies.
    When a platform applied plug-in hybrid technologies in the final 
rule analysis, the engine and transmission technologies are superseded. 
For all plug-in technologies, an AT8L2 transmission is used. For the 
PHEV20/50 and PHEV20/50H, the engine is replaced by an Atkinson cycle 
based engine (Eng26). For the PHEV20/50T, the engine is replaced by the 
TURBO1 technology engine (Eng12).
    The PHEV20/30 and PHEV20/50H path also had similar constraints as 
the SHEVPS in the final rule analysis. Performance vehicles and 
vehicles with a high towing requirement were restricted from selection 
of the PHEV20/30 and PHEV20/50H technologies. Restrictions that were 
applied used the same criteria described for the SHEVPS.
(4) Battery Electric Vehicles
    For the NPRM analysis, the BEV200 technology was applied as an end-
of-path technology. The BEV200 technology was the only battery electric 
vehicle option. For the final rule analysis, the BEV300 was added as a 
technology option beyond the BEV200, as discussed in Section 
VI.C.3.a)(1)(f) Battery Electric Vehicles. BEV200 and BEV300 technology 
was applied in place of all engine and transmission technologies, and 
was an end of path technology.
    For the final rule analysis, both the BEV 200 and BEV300 had phase-
in cap limitations applied based on an analysis of the market 
availability and cost of batteries.\1074\ The BEV200 was limited to a 
greater extent than the BEV300, accounting for expected limits in 
market demand for the shorter-range BEV.\1075\ The phase-in capacity 
numbers were determined based on the results of the analysis of the 
National Energy Model System (NEMS) discussed in Section 
VI.D.1.b)(1)(b) Macroeconomic assumptions used to analyze economic 
consequences of the final rule.
---------------------------------------------------------------------------

    \1074\ John Elkin, MIT finds that it might take a long time for 
EVs to be as affordable as you want, Digital Trends (November 23, 
2019), https://www.digitaltrends.com/cars/mit-study-finds-ev-market-will-stall-in-the-2020s/.
    \1075\ MIT Energy Initiative. 2019. Insights into Future 
Mobility. Cambridge, MA: MIT Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
---------------------------------------------------------------------------

(5) Fuel Cell Vehicle
    For the NPRM analysis, FCV technology was also applied as an end of 
path technology. The FCV technology was also applied as end of path 
technology in the final rule analysis.
    For the final rule analysis, a phase-in cap was assigned to FCV 
technology. The phase-in cap was assigned based on existing market 
share as well as an analysis of expected infrastructure availability 
during the time frame of regulation.1076 1092
---------------------------------------------------------------------------

    \1076\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends. Last accessed Aug 23, 2019.
---------------------------------------------------------------------------

d) Electrification Effectiveness Modeling and Resulting Effectiveness 
Values
    For this analysis, the agencies considered a range of 
electrification technologies which, when modeled, resulted in varying 
levels of effectiveness at reducing fuel consumption. Each technology 
consists of many different complex sub-systems with unique component 
efficiencies and operational modes. As discussed further below, the 
systems that contribute to the effectiveness of an electrified 
powertrain in the analysis include the vehicle's battery, electric 
motors, power electronics, and accessory load. Procedures for modeling 
each of these sub-systems are discussed below, and also in Section 
VI.B.3 Technology Effectiveness Values and in the FRM Argonne Model 
Documentation.
    The modeled electrification technologies included micro hybrids, 
mild hybrids, strong hybrids, plug-in hybrids, and full electric 
vehicles. This section discusses how Autonomie was used to model these 
technologies' effectiveness. The models for the micro hybrids included 
a SS12V system model; mild hybrid models included BISG system models 
and CISG system models; strong hybrid models included SHEVP2 system 
models and SHEVPS system models; and finally, electric vehicle models 
included BEV system models and FCV system models.

[[Page 24479]]

(1) Electric Motors, Power Electronics and Accessory Load
    Each electrified powertrain type possesses a unique effectiveness 
for reducing fuel consumption. Autonomie determines the effectiveness 
of each electrified powertrain type by modeling the basic components, 
or building blocks, found in each powertrain, and then combining the 
components modularly to determine the overall efficiency of the entire 
powertrain. The basic building blocks that comprise an electrified 
powertrain in the analysis included the battery, electric motors, power 
electronics, and accessory loads. Autonomie identified which components 
comprise each electrified powertrain type, and how these components are 
interlinked within each unique electrified powertrain architecture. 
This creates a model for each electrified powertrain architecture that 
simulates how efficiently energy is transferred through each system. 
For example, Autonomie determines a BEV's overall efficiency by 
considering the efficiencies of the battery, the electric traction 
drive system (the electric machine and power electronics) and 
mechanical power transmission devices. Or, for a SHEVP2, Autonomie 
combines a very similar set of components to model the electric portion 
of the hybrid powertrain, and then also includes the combustion engine 
and related power transmission components.
    For the NPRM and this final rule analysis, Autonomie employed a set 
of electric motor efficiency maps, which originated from two Oak Ridge 
National Laboratory (ORNL) studies: one for a traction motor and an 
inverter, the other for a motor/generator and 
inverter.1077 1078 Autonomie also used test data validations 
from technical publications to determine the efficiency of certain 
electric motors. The electric motor efficiency maps are visual 
measurements of percent efficiency of power as a function of torque and 
motor RPM, and were based on representative production vehicles, 
especially for base and maximum speeds as well as maximum torque curve. 
The maps were used to determine the efficiency characteristics of the 
motors, but were scaled such that their peak efficiency value 
corresponded to the latest state of the art technologies for different 
electrified powertrains. The maps also included some of the losses due 
to power transfer through the electric machine.\1079\ Table VI-86 
details the electric machine efficiency map sources for the different 
powertrain configurations used for the NPRM.
---------------------------------------------------------------------------

    \1077\ See PRIA, at 374.
    \1078\ Oak Ridge National Laboratory (2008). Evaluation of the 
2007 Toyota Camry Hybrid Synergy Drive System. Submitted to the U.S. 
Department of Energy; Oak Ridge National Laboratory (2011). Annual 
Progress Report for the Power Electronics and Electric Machinery 
Program.
    \1079\ See Chapters 4.7 and 5.5 in the FRM ANL Model 
Documentation.
[GRAPHIC] [TIFF OMITTED] TR30AP20.207

    For the final rule, the agencies used the same efficiency maps as 
the NPRM, except for BEVs. The agencies updated the BEV electric motor 
efficiency for the final rule analysis using data from a more recent 
technical publication.\1081\ The agencies also scaled the maps to have 
peak efficiencies ranging from 96-98 percent depending on the 
powertrain type.\1082\ Table VI-87 below shows powertrain types and the 
source of data used for the final rule.
---------------------------------------------------------------------------

    \1080\ Burak Ozpineci, Oak Ridge National Laboratory Annual 
Progress Report for the Power Electronics and Electronic Motors 
Program, ORNL/SPR-2014/532, https://info.ornl.gov/sites/publications/Files/Pubs3253422.pdf, November 2014. (Nissan Leaf data 
was used for FCV powertrain type).
    \1081\ Faizul Momen, Electric Motor Design of General Motors' 
Chevrolet Bolt Electric Vehicle, 2016-01-1228, SAE International, 
April 5, 2016.
    \1082\ See. Chapter 5.5 in FRM ANL Model Documentation.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24480]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.208

    Battery performance data (e.g., internal resistance, open circuit 
voltage) were measured using individual cell testing on a bench using 
standard test procedures, and BatPaC was used to design battery packs 
of different capacities and cell counts. The battery utilization (e.g. 
SOC range) were developed based on numerous vehicle test data.\1083\ In 
addition, as discussed further below, for the NPRM analysis, the 
agencies resized the battery pack only with the addition of incremental 
mass reduction technology levels. For this final rule, the agencies 
updated the modeling to consider battery resizing with the application 
of all road load reduction technologies. Accordingly, a more 
appropriately-sized battery pack could result in lower vehicle mass, 
resulting in potentially improved effectiveness.
---------------------------------------------------------------------------

    \1083\ Kim, N., & Jeong, J. (2017). Control Analysis and Model 
Validation for BMW i3 Range Extender. SAE Technical Paper 2017-01-
1152. doi:10.4271/2017-01-1152. Jeong, J. K. (2019). Analysis and 
Model Validation of the Toyota Prius Prime. SAE World Congress. SAE. 
Namdoo Kim, A. R. (2017). Vehicle Level Control Analysis for Voltec 
Powertrain. Presented at the 30th International Electric Vehicle 
Symposium and Exhibition (EVS30). Stuttgart, Germany. Hanho Son, N. 
K. (2015). Development of Performance Simulation for a HEV with CVT 
and Validation with Dynamometer Test Data. Presented at the 28th 
International Electric Vehicle Symposium (EVS28). Kintex, Korea.
---------------------------------------------------------------------------

    Beyond the powertrain components, Autonomie also considered on-
board accessory devices that consume energy and affect overall vehicle 
effectiveness. Some electrical power is consumed by electrical 
accessories such as headlights, radiator fans, wiper motors, engine 
control units (ECU), transmission control unit (TCU), cooling systems, 
and safety systems, in addition to driving the motor and the wheels. In 
real-world driving, the electrical accessory load on the powertrain 
varies depending on the how features are used and the condition the 
vehicle is operating in, such as for night driving or hot weather 
driving. However, for regulatory test cycles related to fuel economy, 
the electrical load is repeatable because the fuel economy and 
CO2 regulations control for these factors, as discussed in 
Section VI.B.3 Technology Effectiveness Values.\1084\ Accessory loads 
during test cycles do vary by powertrain type and vehicle technology 
class, since distinctly different powertrain components and vehicle 
masses will consume different amounts of energy.
---------------------------------------------------------------------------

    \1084\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford 
F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812 
520.
---------------------------------------------------------------------------

    The baseline fleet consists of hundreds of different vehicle types 
that vary in the amount of accessory electrical power that they 
consume. For example, vehicles with different motor and battery sizes 
will require different capacities of electric cooling pumps and fans to 
manage component temperatures. Autonomie has built-in models that can 
simulate these varying sub-system electrical loads. However, for the 
NPRM and this final rule analysis, the agencies used a fixed (by 
vehicle technology class and powertrain type), constant power draw to 
represent the effect of these accessory loads on the powertrain. The 
agencies intended and expected that fixed accessory load values would, 
on average, have similar impacts on effectiveness as found on actual 
manufacturers' systems. This process was in line with the past 
analyses, such as in the Draft TAR and the EPA Proposed 
Determination.\1085\ \1086\ For assumptions regarding accessory load 
modeling for the rulemaking timeframe, the agencies relied on research 
and development data from DOE's Vehicle Technologies Office and Argonne 
Advanced Mobility Technology Laboratory, as well as input from 
automotive manufacturers.\1087\ \1088\ \1089\
---------------------------------------------------------------------------

    \1085\ Draft Technical Assessment Report (July 2016), Chapter 5.
    \1086\ EPA Proposed Determination TSD (November 2016), at p.2-
270.
    \1087\ DOE VTO Power Electronics Research and Development. 
https://www.energy.gov/eere/vehicles/vehicle-technologies-office-
electric-drive-systems. Last Accessed Jan 2, 2020.
    \1088\ ANL Advanced Mobility Technology Laboratory (AMTL). 
https://www.anl.gov/es/advanced-mobility-technology-laboratory. Last 
Accessed Jan 2, 2020.
    \1089\ DOE's lab years are ten years ahead of manufacturers 
potential production intent (i.e 2020 Lab Year is MY 2030).

---------------------------------------------------------------------------

[[Page 24481]]

    Table VI-88 below shows the NPRM assumptions for all the vehicle 
classes and powertrain types for accessory loads.\1090\ Data from AMTL 
D \3\ testing were used to designate electric loads for different types 
of powertrains.\1091\
---------------------------------------------------------------------------

    \1090\ See NPRM ANL Assumptions Summary.
    \1091\ ANL Energy Systems Division Downloadable Dynamometer 
Database: https://www.anl.gov/es/downloadable-dynamometer-database.
[GRAPHIC] [TIFF OMITTED] TR30AP20.209

BILLING CODE 4910-59-C
    For the final rule analysis, the agencies updated the electrical 
load assumptions for many of the powertrain types and classes,\1092\ 
based on further consideration of comments from the Alliance on the 
2016 Draft TAR and EPA Proposed Determination.\1093\ \1094\ These 
assumptions are provided below, in Table VI-89.
---------------------------------------------------------------------------

    \1092\ See ANL Assumptions Summary, ANL--All 
Assumptions_Summary_FRM_06172019_FINAL.
    \1093\ Alliance of Automobile Manufacturers Comments on Draft 
TAR at p. 30. September 26, 2016.
    \1094\ EPA Proposed Determination TSD (November 2016), at p.2-
270.
[GRAPHIC] [TIFF OMITTED] TR30AP20.210


[[Page 24482]]


    CARB commented on NPRM non-battery component efficiency assumptions 
in two respects; first by claiming that the agencies relied on outdated 
data for electric machines and inverter efficiencies across all 
electrification applications,\1095\ and second by claiming that the 
agencies did not project any efficiency gains in those components over 
time.\1096\ CARB stated that the three vehicles benchmarked in the ORNL 
studies (MY 2007 Toyota Camry Hybrid, a MY 2011 Hyundai Sonata Hybrid, 
and MY 2012 Nissan Leaf) were inappropriate for the agencies to use to 
assess the costs and efficiencies for the same components in MY 2020-
2030 vehicles, given the rapid development in the past ten years in 
automotive electrification. CARB cited the MY 2016 Chevrolet Volt and 
Bolt, and the MY 2016 Toyota Prius, as examples of vehicles that had 
undergone electric machine efficiency improvements from one generation 
to the next; those vehicles generally employed efficiency improvements 
including reduced electric motor volume and mass, reduced power 
inverter volume, increased electric motor peak power density, and 
reduced mechanical losses through friction reduction, among other 
improvements.
---------------------------------------------------------------------------

    \1095\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 127.
    \1096\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 128.
---------------------------------------------------------------------------

    In support of their comments that the agencies did not project any 
efficiency gains in non-battery components over time, CARB faulted the 
agencies for not including data from the October 2015 ORNL progress 
report for electric drive technologies, stating that benchmarking data 
for a MY 2014 Honda Accord Hybrid inverter and traction motor 
components could have been used to compare against and update the data 
from the MY 2007 Toyota Camry Hybrid and MY 2011 Hyundai Sonata Hybrid 
efficiency maps benchmarked in the older ORNL report. CARB stated that 
the lack of consideration of this newer data was evidence that the 
agencies' data selection was biased to support weakening fuel economy 
standards.
    CARB also cited 2017 research from Argonne's Autonomie group as a 
source of updated data that showed efficiency gains over time for 
electrification technologies not considered in the agencies' analysis, 
including increases in high voltage system peak efficiency, increases 
in high voltage specific power, and decreases in costs.\1097\ CARB 
stated that had the agencies included newer data in the analysis, 
including from the same data sources from which prior data came, the 
analysis would have not supported the agencies' proposal.
---------------------------------------------------------------------------

    \1097\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 131. Note that comments on non-battery 
component costs are addressed in Section VI.C.3.e)(2) Non-Battery 
Electrification Component Costs.
---------------------------------------------------------------------------

    The agencies agree that there have been improvements in non-battery 
component efficiency over the past few years, however CARB's 
characterization of the process used to employ the ORNL benchmarking 
data in the analysis was incorrect. Autonomie used high-level electric 
machine characteristics such as base and max motor speed from 
production vehicles along with generic efficiency map curves for each 
technology type, with peak efficiencies matching the current state of 
the art technologies discussed in ORNL reports. Although the source 
data for the electric machines were from older production vehicles, the 
peak electric motor and controller efficiencies were updated to reflect 
the latest available data. Specifically, the NPRM analysis modeled a 92 
percent peak efficiency for motors and controllers.\1098\
---------------------------------------------------------------------------

    \1098\ See the Non_Vehicle_Attribute tab in the NPRM ANL 
Assumptions_Summary.
---------------------------------------------------------------------------

    That said, the agencies also agreed that the analysis could use 
updated peak electric and controller efficiencies, and updated those 
for the final rule. For the final rule analysis, the agencies used 96 
percent efficiency for HEVs and PHEVS, and 98 percent peak efficiency 
for BEVS and FCEVs.\1099\ The agencies believe the final rule 
efficiencies are appropriate for the rulemaking timeframe.
---------------------------------------------------------------------------

    \1099\ See the Non_Vehicle_Attribute tab in the FRM ANL 
Assumptions_Summary.
---------------------------------------------------------------------------

    In addition, as discussed above, other changes for the final rule 
analysis include updating the electric motor sizing as a function of 
electric power to account for lower electric machine mass, updating the 
BEV electric machine map to use a newer efficiency map from the Chevy 
Bolt, updating baseline and reference vehicle mass assumptions to 
reflect latest machine weight technology development, and updating the 
electrical accessory loads for vehicle modeling to reflect data from 
vehicle benchmarking. Changes and updates to the Autonomie analysis are 
discussed throughout this electrification section and in the FRM 
Argonne Model Documentation. In addition, for this final rule analysis, 
the agencies used the latest Argonne BatPaC model to determine the 
battery pack mass and manufacturing costs for electric vehicle 
batteries. Updates to non-battery component efficiency were small in 
comparison to the impact of using updated battery modeling for the 
final rule analysis. Further discussion on battery modeling can be 
found in Section VI.C.3.e)(1) Battery Pack Modeling.
(2) Modeling and Simulating Vehicles With Electrified Powertrains in 
Autonomie
    Data from Argonne's AMTL was used to develop the electrified 
powertrain models in Autonomie. The modeled electrification components 
were sized based on performance neutrality needs, as discussed further 
below, and the control algorithms were based on Argonne -collected 
data.\1100\ Detailed discussion about the development of the HEV 
drivetrains can be found in the Autonomie modeling documentation.\1101\ 
The modeled powertrains are not intended to represent any specific 
manufacturer's architecture, but are intended to act as surrogates 
predicting representative levels of effectiveness for each 
electrification technology.
---------------------------------------------------------------------------

    \1100\ See FRM ANL Model Documentation.
    \1101\ See NPRM ANL Model Documentation at p.92.
---------------------------------------------------------------------------

    The agencies also broadly discussed in Section VI.B.3 Technology 
Effectiveness Values that certain technologies' effectiveness for 
reducing fuel consumption requires optimization through the appropriate 
sizing of the powertrain. This analysis iteratively minimizes the size 
of the powertrain components to maximize efficiency while at the same 
time enabling the vehicle to meet multiple performance criteria. The 
Autonomie simulations use a series of resizing algorithms which contain 
``loops,'' such as an ``Acceleration Performance Loop (0-60 mph),'' 
which automatically adjust the size of certain powertrain components 
until a criterion, for example 0-60 acceleration time, converges to a 
target value. As the algorithms examine different performance or 
operational criteria that must be met, no single criterion is allowed 
to degrade; once a resizing algorithm completes, all criteria will be 
met, and some may be exceeded as a necessary consequence of meeting 
others.
    Autonomie applies different powertrain sizing algorithms depending 
on the type of vehicle considered because different types of vehicles 
not only contain different powertrain components to be optimized, but 
they must also operate in different driving modes. While the 
conventional powertrain sizing algorithm must consider only the power 
of the engine, the more complex algorithm for

[[Page 24483]]

electrified powertrains must simultaneously consider multiple factors, 
which could include the engine power, electric machine power, battery 
power and battery capacity. Also, while the resizing algorithm for all 
vehicles must satisfy the same performance criteria, the algorithm for 
some electric powertrains must also allow those electrified vehicles to 
operate in certain driving cycles without assistance of the combustion 
engine, and ensure the electric motor/generator and battery can handle 
the vehicle's regenerative braking power, all-electric mode operation 
and intended range of travel.
    To establish the effectiveness of the technology packages, 
Autonomie simulated the vehicles performing compliance test cycles, as 
discussed in Section VI.B.3 Technology Effectiveness 
Values.1102 1103 1104 For vehicles with conventional 
powertrains and micro hybrids, Autonomie simulated the vehicles using 
the 2-cycle test procedures and guidelines.\1105\ For mild HEVs, strong 
HEVs, and FCVs, Autonomie simulated the 2-cycle test, with the addition 
of repeating the drive cycles until the final state of charge was 
approximately the same as the initial state of charge, a process 
described in SAE J1711. For PHEVs and BEVs, Autonomie simulated 
vehicles performing the test cycles per guidance provided in SAE 
J1711.\1106\ For BEVs, Autonomie simulated vehicles performing the test 
cycles per guidance provided in SAE J1634.\1107\
---------------------------------------------------------------------------

    \1102\ EPA, ``How Vehicles are Tested.'' https://www.fueleconomy.gov/feg/how_tested.shtml. Last accessed Nov 14, 
2019.
    \1103\ See FRM ANL Model Documentation at Chapter 6: Test 
Procedures and Energy Consumption Calculations.
    \1104\ EPA Guidance Letter. ``EPA Test Procedures for Electric 
Vehicles and Plug-in Hybrids.'' Nov. 14, 2017. https://www.fueleconomy.gov/feg/pdfs/EPA%20test%20procedure%20for%20EVs-PHEVs-11-14-2017.pdf. Last accessed Nov. 7, 2019.
    \1105\ 40 CFR part 600.
    \1106\ PHEV testing is broken into several phases based on SAE 
J1711. Charge-Sustaining on the City cycle, Charge-Sustaining on the 
HWFET cycle, Charge-Depleting on the City and HWFET cycles.
    \1107\ SAE J1634. ``Battery Electric Vehicle Energy Consumption 
and Range Test Procedure.'' July 12, 2017.
---------------------------------------------------------------------------

    A survey of comments about the modeled effectiveness of 
electrification technologies showed most comments could be sorted in 
three major categories. The first, and largest category of comments, 
were concerned with effectiveness values used for the technologies. 
Specifically, commenters were concerned the values for the modeled 
effectiveness of the technologies were too low, particularly when 
compared to past analysis efforts. The second major category of 
comments were concerned with the size of the electrification components 
selected in the Autonomie tool, and used to simulate the system 
performance. Commenters were concerned because oversized components can 
lead to the system violating performance neutrality constraints and 
artificially increasing the cost of the technology. The third major 
category of comments were concerned not enough variety of technologies 
were represented in the electrification technology models. 
Specifically, commenters wanted additional engine technologies allowed 
to couple with electrification technologies.
    Each of the comments from the first category will be referenced and 
addressed under the specific technology sections, below. However, 
broadly, two factors have led to the comments raised by stakeholders. 
First, as discussed throughout this document, the agencies avoided 
using performance values in the analysis that can be traced to specific 
implementation of a technology type. Thus, when comparing simulated 
performance to any specific real world vehicle, there will be a 
deviation. The modeled inputs are meant to represent the typical range 
of values for a technology--reasonable and realistic values--but are 
not likely to result in performance outputs that would equal any 
specific existing vehicle. Second, the modeling approach implemented in 
the NPRM and final rule analysis succeeds in isolating the effects of 
individual technologies to a higher degree then previous analysis. Due 
to the greater use of parametric modeling of full vehicle systems, the 
specific effects of technologies could be isolated to a higher degree 
from the amplifying or muting effects of other technologies. This 
isolation of effect often results in lower predicted effectiveness 
values for individual technologies than has been observed in previous 
analysis, where the isolation of effect was not as precise, and often 
attributed efficiency gains from a combination of technological changes 
to a single technology.
    For the second major group of comments, the agencies mostly agreed 
with the stakeholder observations. The issues identified were 
investigated by the agencies and resulted in changes to the sizing 
algorithms used by the agencies for the final rule analysis. The 
agencies further investigated the constraints of performance neutrality 
and ensured those constraints were followed for sizing of 
electrification components. Further discussion of the changes made, as 
well as specific answers to comments under each technology section, can 
be found in the following technology subsections and in Performance 
Neutrality, Section VI.B.3.a)(6).
    The third major group of comments from stakeholders were concerned 
with allowing more engine technologies to be incorporated in 
electrification systems. The agencies agreed with these comments and 
increased the number of technology combinations available. The new 
combinations are discussed in Section VI.C.3.a)(1) Electrification 
Technologies, as well as under each technology section below.
(a) Micro and Mild Hybrid Vehicles
    The micro and mild hybrid systems modeled in Autonomie represented 
SS12V and BISG technology (and CISG technology for the NPRM). SS12V and 
BISG were modeled using a similar approach because both systems have 
low peak power, low energy storage, and allow stop/start engine idle 
reduction. The effectiveness improvement from both technologies is 
attributable to the amount of fuel saved during engine idling period on 
the 2-cycle test. However, only the BISG system model allowed limited 
assist to propel the vehicle and limited regenerative braking. For 
further discussion of these system models, see the FRM Argonne Model 
Documentation.\1108\
---------------------------------------------------------------------------

    \1108\ See FRM ANL Model Documentation at chapters 4.6, 4.7 and 
4.13.
---------------------------------------------------------------------------

    Powertrain resizing was not employed for micro or mild hybrid 
system application, in either the NPRM or this final rule analysis. 
These systems have little to no impact on the vehicle performance 
metrics that would be adjusted by powertrain resizing, and in turn 
there would be limited or no benefit in attempting to resize upon 
application of these systems. For example, the micro hybrid SS12V 
system allows the engine to be turned off when the vehicle is fully 
stopped to reduce idle-stop fuel consumption, but the combustion engine 
size must be retained to maintain performance metrics such as 
acceleration. The main focus of mild hybrid vehicles is to provide 
idle-stop and capture some regenerative braking energy, and although 
they also can provide some assistance to the engine during the initial 
propelling of the vehicle, this is done to improve efficiency and does 
not significantly improve the acceleration performance of the vehicle. 
With BISG mild hybrids, the electric machine is linked to the engine 
through a belt, and thus the potential power assistance is usually 
limited. In the NPRM, the BISG system used an 806 Wh capacity battery

[[Page 24484]]

pack and a 10 kW motor/generator. For the final rule analysis, the 10 
kW motor/generator was paired with a 403 Wh battery pack to align with 
BISG systems emerging in the marketplace.
    ICCT commented that the agencies unjustifiably reduced the 
CO2 and fuel consumption benefits of SS12V from the Draft 
TAR, including a reduction in the overall effectiveness benefit when 
the SS12V system was applied in combination with other 
technologies.\1109\ ICCT stated that the agencies should know the 
precise effectiveness improvement for SS12V technology based on EPA 
compliance data, and the agencies should report a full listing of all 
the baseline 2016 vehicle models with stop-start technology, with their 
test-cycle, and off-cycle improvement in g/mile and percent 
effectiveness. ICCT claimed that the agencies either intentionally 
ignored the full compliance benefits of SS12V technology or ``ignored 
the knowledge and expertise of the EPA engineering and compliance 
staff,'' and argued that not reporting the requested data would be 
``hiding relevant data the agencies have readily available to more 
rigorously assess existing stop-start technologies and their impact for 
the rulemaking.'' ICCT also stated that the agencies did not 
appropriately include the full regulatory benefit (i.e., inclusion of 
the additional off-cycle ``credit'' under EPA's program or fuel 
consumption improvement value under NHTSA's program) of SS12V 
technologies due to their off-cycle improvements.\1126\
---------------------------------------------------------------------------

    \1109\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-22.
---------------------------------------------------------------------------

    HDS made a similar observation, noting that the SS12V benefit from 
the NRPM was similar to the 2012 TSD projection, but lower than the 
benefit quoted by stakeholders in the Draft TAR.\1110\ HDS cited the 
difference in fuel economy between two vehicles that were produced with 
and without a SS12V option (the 2015 Ford Fusion 1.5L TGDI and the 2015 
Mazda 3 i-ELOOP) which suggested effectiveness values for SS12V of 
about 3.3 percent for both vehicles. HDS also cited a Bosch 
presentation that claimed newer SS12V systems could provide 
effectiveness of up to 6 percent. HDS argued that this actual data and 
supplier data supported a benefit of at least 3.3 percent, which they 
stated was double the benefit in the NRPM analysis.
---------------------------------------------------------------------------

    \1110\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
11985, at 44.
---------------------------------------------------------------------------

    The agencies disagree with ICCT and HDS' comments regarding the 
effectiveness of the SS12V technology modeled in the NPRM analysis. The 
implementation of the full vehicle simulation approach used in the 
NPRM, and carried forward to the final rule analysis, clearly defines 
the contribution of individual technologies and separates those 
contributions from other technologies. The modeling approach also shows 
when technologies have amplifying or muting interactions. In some 
cases, this may appear as a reduction in performance compared to 
previous analysis. The agencies modeled the SS12V system in conjunction 
with all the IC engine and transmission combinations. The results of 
this parametric modeling accounted for each engine and transmission 
combination's unique fuel consumption rate at idle.\1111\ The range of 
effectiveness for the technology in the NPRM analysis is a result of 
these differences. This range of values will result in some modeled 
effectiveness values being close to real-world measured values, and 
some modeled values that will depart from measured values, depending on 
the level of similarity between the modeled hardware configuration and 
the real-world hardware configuration. This modeling approach comports 
with the National Academy of Science 2015 recommendation to use full 
vehicle modeling supported by application of lumped improvements at the 
sub-model level.\1112\ The approach allows the isolation of technology 
effects in the analysis supporting an accurate assessment.
---------------------------------------------------------------------------

    \1111\ For example, when idling, a larger eight-cylinder engine 
has more friction and pumping losses than a smaller four-cylinder 
engine, and therefore will save more fuel when the engine is shut-
off at rest.
    \1112\ National Research Council. 2015. Cost, Effectiveness, and 
Deployment of Fuel Economy Technologies for Light-Duty Vehicles. 
Washington, DC--The National Academies Press. https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles, at 292.
---------------------------------------------------------------------------

    For both the NPRM and final rule analysis, the agencies assigned 
SS12V technology to vehicles in the analysis fleet using compliance 
data, and used compliance data to assign a vehicle's baseline fuel 
economy value. The market data file indicated the presence of SS12V on 
a vehicle, and accordingly, the vehicles reported to include SS12V 
technology in the analysis fleet were modeled with the technology. For 
more discussion on how technologies were assigned to the vehicle 
platforms in the analysis fleet, please see Section VI.B.1 Analysis 
Fleet. The agencies accounted for the contribution of the SS12V 
technology in the analysis fleet by using the reported compliance fuel 
economy values as the baseline fuel economy values for vehicles that 
included the technology. The analysis fleet fuel economy values were 
the reported final compliance values for the given vehicle platform and 
should include the benefits from all technologies on the vehicle 
platform.\1113\ The agencies also captured the off-cycle credits 
provided to a manufacturer for the existence of the technology in the 
manufacturer's fleet. For the NPRM and final rule analysis, the 
manufacturers' fleets are modeled with baseline year compliance-
reported off-cycle credits. Further, for the final rule analysis, the 
agencies increased the application of off-cycle credits in the 
analysis, as discussed in Section VI.B.2.a) Off-cycle and A/C 
Efficiency Adjustments to CAFE and Average CO2 Levels.
---------------------------------------------------------------------------

    \1113\ Sec.  32904. Calculation of average fuel economy, https://uscode.house.gov/browse/prelim@title49/subtitle6/partC/chapter329&edition=prelim.
---------------------------------------------------------------------------

    Commenters similarly disagreed with the BISG effectiveness 
presented in the NPRM analysis, suggesting the resulting effectiveness 
improvement should be at a range of 4 percent to 6 percent.\1114\ Such 
commenters claimed that it was unclear why effectiveness values were so 
much lower than previous effectiveness estimates. More specifically, 
comments centered on (1) arguing that the agencies' modeling of BISG 
and CISG systems in Autonomie likely underestimated the resulting 
effectiveness values; (2) suggesting that the values in prior documents 
like the Draft TAR and the 2015 NAS report were more accurate; and (3) 
comparing modeled effectiveness values to claimed values achieved by 
actual on-road vehicles and mild hybrid systems.
---------------------------------------------------------------------------

    \1114\ ICCT, Attachment 3, Docket No. NHTSA-2018-0067-11741; 
California Air Resources Board, Attachment 2, Docket No. NHTSA-2018-
0067-11873; Roush Industries, Attachment 1, NPRM Docket No. NHTSA-
2018-0067-11984; H-D Systems, ``HDS final report,'' Docket No. 
NHTSA-2018-0067-11985; Union of Concerned Scientists, Attachment 2, 
Docket No. NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    CARB claimed that the agencies failed to disclose the necessary 
details to conclude why mild hybrid systems were projected to have 
lower efficiency values than past estimates. CARB also concluded the 
lack of engine downsizing when adding a BISG/CISG system and the lack 
of adjusting transmission drive ratios and shift logic were reasons why 
BISG/CISG effectiveness was underpredicted.\1115\ CARB claimed not 
resizing the engines resulted in a ``less than optimized system that 
does not take full advantage

[[Page 24485]]

of the mild hybrid system.'' \1116\ CARB argued that the agencies' 
assumption that manufacturers ``would not optimize the engine and 
transmission when installing a CISG is not realistic and results in 
improper pairing of advanced gasoline engines and transmissions in the 
modeling and leads to underestimation of the efficiency benefits.'' As 
mentioned above, CARB stated that manufacturers ``often are required to 
make a[n] engine casting change to accommodate the system,'' and when 
doing so, ``no manufacturer would fail to pair the system with an 
optimally sized engine and configured transmission to take full 
advantage of the system's capabilities.'' \1117\
---------------------------------------------------------------------------

    \1115\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 163.
    \1116\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 185.
    \1117\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 186.
---------------------------------------------------------------------------

    CARB also inquired into whether the Argonne modeling ``took full 
advantage'' of the system, using Daimler's EQ Boost system, that 
provides temporary boosts for acceleration and enables engine shut-off 
during coasting events, as an example.\1118\ Similarly, CARB noted that 
CISG systems' ability to provide low end torque makes it an ``ideal 
technology to pair with an engine technology that may have poor low end 
torque but improved efficiency under other conditions; examples could 
include an HCR engine sized with minimal low end torque to maximize 
efficiency improvements in other operating conditions or a turbocharged 
downsized engine equipped with a larger turbine to reduce backpressure 
but provide improved efficiency over a larger portion of the engine 
map.'' \1119\ CARB stated that manufacturers are using such systems to 
boost engine torque at higher operating speeds so they can keep the 
engine operating in a more efficient region.
---------------------------------------------------------------------------

    \1118\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 163.
    \1119\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 163.
---------------------------------------------------------------------------

    Commenters also cited data from suppliers that produce 48V BISG 
systems, including data from TULA that showed a 11 percent fuel economy 
benefit from a 48V system,\1120\ data from a Delphi 48V system 
prototype installed on a Honda Civic that showed a 10 percent reduction 
in CO2 emissions levels,\1121\ and data from Continental 
showing a 13 percent fuel savings improvement from its BISG 
system.\1122\ ICCT also cited its supplier and technology report on 
hybrids that estimated the benefit of mild hybrid technology at 12.5 
percent, which it characterized as ``remarkably similar'' to that 
achieved by the 2019 RAM pickup truck.\1123\ HDS noted that even if the 
effectiveness values from TULA are regarded as optimistic because they 
are the developers of the technology, EPA's previous modeling results 
of 8-9 percent effectiveness ``appear reasonable in light of what is 
observed from certification data.'' \1124\ ICCT ultimately recommended 
the agencies revise the effectiveness value for mild hybrid systems to 
include a CO2 effectiveness value of 12.5 percent.\1125\
---------------------------------------------------------------------------

    \1120\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
11985, at 45.
    \1121\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 160.
    \1122\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 160.
    \1123\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-24.
    \1124\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
11985, at 45.
    \1125\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-25.
---------------------------------------------------------------------------

    Commenters also stated that the effectiveness estimates for CISG 
systems were significantly understated, \1126\ with UCS characterizing 
CISG systems as showing ``virtually no benefit whatsoever for CISG over 
BISG, and in many cases actually show[ing] an increase in fuel 
consumption.'' \1127\ UCS stated this was a dramatic departure from 
previous Autonomie results, and with ``no explanation whatsoever'' 
given for the decrease in technology effectiveness.
---------------------------------------------------------------------------

    \1126\ Union of Concerned Scientists, Attachment 2, Docket No. 
NHTSA-2018-0067-12039; Roush-Industries, Attachment 1, Docket No. 
NHTSA-2018-0067-11984; California Air Resources Board, Attachment 2, 
Docket No. NHTSA-2018-0067-11873.
    \1127\ Union of Concerned Scientists, Attachment 2, Docket No. 
NHTSA-2018-0067-12039, at 3.
---------------------------------------------------------------------------

    The agencies agree with commenters that the NPRM analysis of mild 
hybrid technologies could be more representative of production vehicles 
and vehicles likely to be produced during the rulemaking time period. 
The agencies further conclude that the NPRM analysis overestimated the 
costs of such technologies. Thus, for the final rule analysis, the 
agencies only considered one 48V BISG system in the mild hybrid 
technology category. The 48V mild hybrid BISG system used the same 10 
kW electric motor as the one used in the NPRM analysis, and the 48V 
BISG battery pack was also reduced in size to 403 W-hr from 806 W-hr to 
reflect more accurately the size of battery packs available in the 
market. In addition, the Autonomie model increased the usable battery 
capacity, increasing the duration of electric motor use by the vehicle 
before starting the engine. The specifications and assumptions for the 
48V BISG system are further discussed in the FRM Argonne Model 
Documentation and FRM Argonne Assumptions Summary.1128 1129 
The discontinued use of the CISG technology is discussed in Section 
VI.C.3.a)(1)(c) Electrification Technologies, Mild Hybrids.
---------------------------------------------------------------------------

    \1128\ See FRM ANL Model Documentation, at 4.6, 4.13, and 5.7.
    \1129\ FRM ANL Assumptions Summary (see Model Documentation 
tables in Section VI.A.7 Structure of Model Inputs and Outputs).
---------------------------------------------------------------------------

    The agencies disagree with comments stating incremental 
effectiveness estimated by Autonomie modeling was incorrect because the 
effectiveness values deviated from past effectiveness values estimated 
in the agencies' rulemakings or from real-world values measured on 
specific vehicles. As discussed in previous sections, the 
implementation of the full vehicle simulation approach used in the NPRM 
analysis and carried forward to the final rule analysis clearly defines 
the contribution of individual technologies through the application of 
parametric modeling. This approach clearly separated the contributions 
of each technology. The modeling approach also showed the amplifying or 
muting interactions between technologies. In some cases, this may 
appear as reduced performance in comparison to previous analysis. The 
agencies also strongly disagree that they should use the performance 
values for any specific vehicle as representative of all mild hybrid 
systems.
    CARB also commented that the agencies' decision to use a fixed 
final drive ratio and fixed shift logic based on the selected 
transmission did not allow for efficiency improvements when mated with 
electrified powertrains, with specific regards to mild hybrid BISG and 
CISG systems.\1130\ CARB stated that based on the information disclosed 
in the NPRM, ``it appears that Argonne did not utilize the system in 
these manners nor did they allow for changes in gear ratios, final 
drive ratio, or transmission shift logic to optimize for efficiency 
improvements when mated with different electrified powertrains.'' 
\1131\ Roush Industries similarly stated that the analysis under-
predicted the potential improvements of employing a BISG system because 
the engine could operate at a lower RPM with the help of the torque 
assist of the electric motor/generator, with a change to the final

[[Page 24486]]

drive ratio and transmission shift logic, but the analysis did not do 
so.\1132\
---------------------------------------------------------------------------

    \1130\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 185.
    \1131\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 185.
    \1132\ Roush Industries, Attachment 1, Docket No. NHTSA-2018-
0067-11984, at 16.
---------------------------------------------------------------------------

    The agencies disagree with CARB and Roush Industries' claims about 
the gear ratio and shift logic used for the NPRM. As discussed in 
Section VI.C.2.d) Transmission Effectiveness Modeling and Resulting 
Effectiveness Values, manufacturers commonly maintain the same gear 
hardware across vehicle platforms and applications, relying on controls 
and shift strategy to achieve optimization. Autonomie maintained gear 
hardware but customized the shifting strategy for each unique vehicle 
system modeled \1133\ to reflect real-world manufacturing strategies 
more accurately.
---------------------------------------------------------------------------

    \1133\ FRM ANL Model Documentation, at 4.4.5.
---------------------------------------------------------------------------

    CARB also commented that the performance modeled by the Autonomie 
tool in the NPRM analysis failed to remain neutral for over 80 percent 
of the modeled systems with mild hybrids. CARB felt the over-
performance was ``indicating some portion of the system capability was 
improperly modeled to improve performance rather than reduce 
CO2 emissions.'' \1134\
---------------------------------------------------------------------------

    \1134\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 163.
---------------------------------------------------------------------------

    The agencies agree with CARB's observations about the performance 
of mild hybrid combinations. The mild hybrid configuration exhibited 
higher performance in comparison to non-mild hybrid configurations in 
the NPRM analysis. For the final rule analysis, the agencies updated 
sizing and control of the mild hybrid systems to minimize performance 
changes and maintain neutrality. As discussed earlier in this chapter, 
updates include using smaller battery systems, updated algorithms, and 
updated component weights. For further discussion of performance 
neutrality for the final rule, see the Performance Neutrality Section 
VI.B.3.a)(6).
    Finally, ICCT commented that the agencies should include off-cycle 
and ``game-changing'' pickup truck credits in the effectiveness 
estimates for hybrid pickup trucks, as ``[i]t is the responsibility of 
the agencies to include all applicable credits with their technology 
packages calculations and their projections, including any additional 
credits that will automatically accrue.'' \1135\
---------------------------------------------------------------------------

    \1135\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-25.
---------------------------------------------------------------------------

    While the agencies included many compliance flexibilities in the 
modeling for the final rule analysis, hybrid pickup truck credits were 
not modeled. The referenced pickup truck credit is set to expire for 
all pickup trucks after MY 2021, so in analyzing this comment the 
agencies considered what technologies manufacturers could apply to 
pickup trucks through that model year to meet the requirements 
specified in the regulation. To receive credit in a model year, 
manufacturers must produce a quantity of improved full size pickup 
trucks--improvement characterized by including either hybrid technology 
or improved emissions performance--such that the proportion of 
production of such vehicles, when compared to the manufacturer's total 
production of full size pickup trucks, is not less than an amount 
specified in that model year. The agencies determined that, based on 
manufacturers' MY 2017 pickup truck offerings characterized in the 
analysis fleet and with the technology considered in this rule, no 
pickup truck manufacturer could meet the criteria set by EPA to qualify 
for the mild credit before the credit is set to expire. For the strong 
HEV credit, the agencies considered that forcing the application of 
strong HEV pickups to meet the minimum threshold of 10 percent of the 
fleet in order to earn the incentive credits would significantly 
increase the cost for compliance and be less cost-effective than other 
technology pathways. As the analysis seeks the most cost-effective 
pathway for compliance, the agencies disagree the analysis should force 
the application of strong HEV technology to at least 10 percent of full 
size pickup trucks. However, the agencies did allow and simulated 
maximum off-cycle and A/C off-cycle FCIVs for all manufacturers in the 
CAFE model for both the CAFE and CO2 programs during the 
rulemaking time frame. So, while the agencies did not model pickup 
truck credits specifically, the final rule analysis allowed 
manufacturers to reach the maximum off-cycle credit cap during the 
rulemaking timeframe.
(b) Strong Hybrid Vehicles
    The power-split hybrid (SHEVPS) model in Autonomie included a 
power-split device, two electric machines and an engine, and allowed 
various interactions between these components. The SHEVP2 model in 
Autonomie is based on the pre-transmission (P2) configuration where the 
electric motor is placed between the engine and transmission for direct 
flow of power to the wheels. The vehicle can be propelled either by the 
combustion engine, electric motor, or both simultaneously, but the 
speed/efficiency region of operation for SHEVP2s under any engine/motor 
combination is ultimately dictated by the transmission gearing and 
speed. Detailed discussion of SHEVPS and SHEVP2 modeling and validation 
are provided in the Argonne Model Documentations.\1136\ Autonomie full 
vehicle models representing strong hybrids were based on vehicle test 
data from vehicle benchmarking.
---------------------------------------------------------------------------

    \1136\ FRM ANL Model Documentation, at Chapters 4.13, 4.16 and 
6.0.
---------------------------------------------------------------------------

    As discussed previously in this section, power-split hybrids 
utilize a combustion engine, two electric machines and a planetary gear 
set along with a battery pack to propel the vehicle. The smaller motor/
generator (EM1) is used to control the engine speed and uses the engine 
to either charge the battery or to supply additional electric power to 
the second ``drive'' motor. The more powerful drive motor/generator 
(EM2) is permanently connected to the vehicle's final drive and always 
turns with the wheels. The SHEVPS resizing algorithm makes an initial 
estimate of the size of the engine, battery, and electric motors. The 
initial estimates for the combustion engine and EM2 sizes are based on 
the peak power required for acceleration performance and the continuous 
power required for gradeability performance. The initial estimates for 
the battery and EM1 powers are based the maximum regenerative braking 
power. With these initial size estimates, the algorithm computes the 
vehicle mass, and simulations are run to determine if 0-60 and 50-80 
mph acceleration performance is acceptable. If acceleration is not 
satisfactory (too fast or too slow), the algorithm iteratively adjusts 
the sizes of the engine, motors, and battery, and runs simulations 
until a minimum powertrain size is found that meets all requirements. 
With each iteration, the engine, battery, and motor characteristics 
were also updated for gradeability performance and regeneration, if 
necessary. Figure VI-32 below shows the general steps of the SHEVPS 
sizing algorithm. Detailed descriptions are available in section 8.3 of 
the FRM Argonne Model Documentation.
BILLING CODE 4910-59-P

[[Page 24487]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.211

    A parallel hybrid (SHEVP2) uses a combustion engine and a multi-
speed transmission-integrated electric motor (EM1), as discussed 
previously in this section. As is done with SHEVPS, the SHEVP2 resizing 
algorithm creates a starting point by making an initial estimate of the 
size of the engine, battery, and electric motor based on performance 
criteria or an estimated regenerative braking power, in turn 
calculating the associated vehicle mass. The algorithm then uses a 
simulation loop to find a more precise value of regenerative braking 
power generated in the UDDS ``city driving'' cycle, and adjusts the 
electric motor size and vehicle mass accordingly. Next, the algorithm 
uses simulation loops to optimize the engine, motor, and battery sizes 
in relation to acceleration performance criteria. In the event that the 
acceleration criteria requires downsizing the powertrain, the electric 
motor size is not reduced as this would not be suitable for the 
handling of regenerative braking power. If the acceleration criteria 
cause the electric motor to increase in size, the algorithm then 
returns to the regenerative braking loop and subsequently all other 
loops until all components are optimized. Figure VI-33 below shows a 
simplified sizing algorithm for SHEVP2s.
BILLING CODE 4910-59-C
    In the NPRM, the acceleration optimization loops in the SHEVP2 
algorithm did not resize the powertrain if the resulting acceleration 
time was less than the target. This strategy was intended to avoid 
reducing the engine size compared to the conventional vehicle, 
mimicking an observed marketplace trend in which parallel hybrid models 
tend to retain similar engine sizes as the non-hybrid models bearing 
the same nameplate. However, in some cases this resulted in overly 
aggressive SHEVP2 acceleration times; to further maintain performance 
neutrality, the final rule sizing algorithm for standard (non-
performance) SHEVP2 vehicle powertrains was changed to allow engine 
downsizing such that acceleration performance could converge toward the 
target value. This algorithm update is also detailed in Section 
VI.B.3.a)(6), Performance Neutrality.
    CARB, ICCT, Meszler and ACEEE commented that some combinations of 
advanced engines mated with strong hybrids were illogical and 
inefficient.\1137\ \1138\ \1139\ \1140\ The commenters specifically 
discussed combinations of SHEVP2 with TURBO2 and CEGR1 technologies 
that stated the incremental effectiveness resulted in near zero to 
negative value, but also clarified that not all combinations showed 
inappropriate effectiveness. CARB further expanded that ``[t]hese are 
not likely combinations utilized by manufacturers as they unnecessarily 
add both gasoline technology and hybrid technology that negates many of 
the benefits of the advanced gasoline technology. This error in the 
Agencies' modeling leads to inflated technology costs on vehicles that 
are converted into P2HEVs.'' \1141\
---------------------------------------------------------------------------

    \1137\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 155.
    \1138\ American Council for an Energy-Efficient Economy, ACEEE 
SAFE NPRM comments, Docket No. NHTSA-2018-0067-12122-22, at 8.
    \1139\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-25.
    \1140\ Comments from Meszler Engineering Services, Attachment 2, 
NPRM Docket No. NHTSA-2018-0067-11723, at 14.
    \1141\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 186.
---------------------------------------------------------------------------

    The agencies now conclude that the NPRM included certain engine and 
strong hybrid pairings that resulted in incremental effectiveness that 
exceeded a reasonable level of performance neutrality. The agencies 
also agree that Autonomie should model strong hybrid technology 
combinations with other engine technologies. In response to these 
comments, for the final rule analysis the agencies updated the CAFE 
model to allow the use of HCR engine technologies with strong hybrids, 
as discussed in Section VI.C.1.c)(4) Engine Maps, HEV Atkinson Cycle 
Engines, and improved full vehicle modeling of turbocharged engine 
combinations. These changes were discussed in Section VI.B.3.a)(1) 
Full-Vehicle Modeling, Simulation Inputs and Data Assumptions and 
Section VI.C.2.d)(1)(a) Shifting Controller.
    In addition, the agencies limited adoption of advanced engine 
technologies with strong hybrids in cases where the electrification 
technology would have little effectiveness benefit beyond the benefit 
of the advanced engine system, but

[[Page 24488]]

would substantially increase costs. Specifically, the agencies did not 
model strong hybrid technologies with VCR engines (eng26a) and eBoost 
engines (eng23c). The agencies believe that manufacturers would not 
consider these combinations because the combination of electrification 
and advanced engine technologies are not as cost-effective as other 
technologies.
c) Plug-In Hybrid Vehicles
    The effectiveness of the PHEV systems in the analysis was dependent 
on both the vehicle's battery pack size and range, in addition to the 
other fuel economy-improving technologies on the vehicle (e.g., 
aerodynamic and mass reduction technologies). For the NPRM analysis, 
the electrification components were sized to achieve the specified all-
electric range (AER) on the combined cycle (UDDS + HWFET) on the basis 
of adjusted energy values. As mentioned above, the PHEV would provide 
propulsion energy for a limited range in addition to start-stop or 
idle-stop. The NPRM analysis classified PHEVs into two levels: (1) 
PHEV30 indicating a vehicle with an AER of 30 miles; and (2) PHEV50 
indicating a vehicle with AER of 50 miles.
    The resizing algorithm for plug-in hybrid (PHEV) vehicles, 
similarly as for SHEVs, considered the power needed for acceleration 
performance and all-electric mode operation (compared to regenerative 
braking for SHEVs); the PHEV resizing algorithms used those metrics for 
an initial estimation of engine, motor(s) and battery powers, and 
battery capacity. The initial mass of the vehicle was then computed, 
including weight for a larger battery pack and charging 
components.\1142\ However, since PHEVs offer expanded electric driving 
capacity, their resizing algorithm must also yield a powertrain with 
the ability to achieve certain driving cycles and range in electric 
mode, in which the engine remains off all or the majority of the 
operation. The analysis sized the PHEV electric motors and battery 
powers to be capable of completing either the City Cycle (UDDS) or US06 
(aggressive, high speed) driving cycle in electric mode, and the 
battery energy storage capacity to achieve the specified all-electric 
range on the 2-cycle tests on the basis of adjusted energy 
values.1143 1144
---------------------------------------------------------------------------

    \1142\ FRM ANL Model Documentation, at 8.3 Vehicle Powertrain 
Sizing Algorithms.
    \1143\ Battery sizing and definition of combined 2-cycle tests 
all-electric range is discussed in detail in ANL Autonomie Model 
Documentation Chapter 6 Test Procedure and Energy Consumption 
Calculation.
    \1144\ ANL has incorporated SAE J1711 standard into Autonomie 
Modeling. J1711: Society of Automotive Engineers Recommend Practice 
for Measuring Exhaust Emissions and Fuel Economy of Hybrid-Electric 
Vehicles, Including Plug-In Hybrid Vehicles.
---------------------------------------------------------------------------

    The final rule analysis classified PHEVs into four technology 
levels, as discussed previously: (1) PHEV20 indicating a vehicle with 
an AER of 20 miles and powertrain system based on SHEVPS hybrid 
architecture; (2) PHEV50 indicating a vehicle with an AER of 50 miles 
and powertrain system based on SHEVPS hybrid architecture; (3) PHEV20T 
indicating a vehicle with an AER of 20 miles and powertrain system 
based on SHEVP2 hybrid architecture; and (4) PHEV50T indicating a 
vehicle with AER of 50 miles and powertrain system based on SHEVP2 
hybrid architecture.\1145\ The PHEV20, PHEV20T, PHEV50, and PHVE50T 
resizing algorithms were functionally equal, and differed only in the 
type of electric mode driving cycle simulated in each one (UDDS for 
PHEV20/20T, or US06 for PHEV50/50T). These algorithms simulated the 
driving cycles in an iterative loop to determine the size of the 
electric motors and the battery required to complete the cycles. In the 
case of PHEV20 and PHEV20T, the power of the electric motors and 
battery must be sized to propel the vehicle through the UDDS cycle in 
``charge-depleting (CD) mode;'' in this mode, the electric machine 
alone propels the vehicle except during high power demands, at which 
point the engine may turn on and provide propulsion assistance. The 
PHEV50 and PHEV50T motor(s) and battery must be sized to power the 
vehicle through the US06 cycle in ``electric vehicle (EV) mode,'' where 
the engine is off at all times. Then, all PHEV algorithms adjusted the 
battery capacity, or vehicle range, by ensuring the battery energy 
content was sufficient to complete a simulated UDDS+HWFET combined 
driving cycle, based on EPA-adjusted energy consumption. Finally, the 
engine, electric motor(s), and battery powers were then sized 
accordingly to meet 0-60 and 50-80 mph acceleration targets. All loops 
were repeated until the acceleration targets were met without needing 
to resize the electric motors, at which point the resizing algorithm 
finished. Figure VI-34 below shows the general steps of the PHEV sizing 
algorithm. Detailed steps can be seen in section 8.3 of the FRM Argonne 
Model Documentation.
---------------------------------------------------------------------------

    \1145\ As discussed previously, the NPRM analysis included 
PHEV30 instead of PHEV20. However, the related resizing algorithm is 
applicable to either.

---------------------------------------------------------------------------

[[Page 24489]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.213

    Meszler, CARB, and BorgWarner provided comments on the 
effectiveness of the PHEV models. The commenters were concerned with 
underperformance of the technology, sizing of the components, and the 
variety of PHEV technologies available.
    Meszler commented that PHEVs in the 2016 analysis fleet were 
inappropriately constrained in their future fuel economy potential by 
the ratio of baseline electric-only fuel economy to baseline engine-on 
fuel economy; and those vehicles should be allowed to improve that 
ratio over time, identically to vehicles that adopt PHEV technology 
during the analysis period.\1146\
---------------------------------------------------------------------------

    \1146\ Meszler Engineering Services, Attachment 2, NPRM Docket 
No. NHTSA-2018-0067-11723 at 32.
---------------------------------------------------------------------------

    The agencies must use the SAE J1711 method for determining the fuel 
economy for the PHEV systems. The use of SAE J1711 and the underlying 
duel fuel vehicle fuel economy calculations are defined by 
statute.\1147\ However, it is important to note that PHEVs are not 
excluded from applying greater range technologies within the PHEV 
technology paths; that is, a PHEV with a lower AER can progress to 
become a PHEV with a longer AER.
---------------------------------------------------------------------------

    \1147\ 49 U.S.C. 32901(b)(1).
---------------------------------------------------------------------------

    CARB commented that several aspects of the agencies' PHEV modeling 
contributed to increased PHEV costs. CARB stated that the electric 
motors were oversized, that all-electric vehicle efficiencies were low, 
and that the lack of battery resizing for road load reductions other 
than mass reduction resulted in battery energy capacities much higher 
than production vehicles.\1148\ CARB stated the modeled battery 
capacity to achieve a given range (kWh/mi) was larger than what exists 
on several representative production vehicles.
---------------------------------------------------------------------------

    \1148\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 149. Specific comments related to costs 
are discussed in Section VI.C.3.e) Overview of Electrification 
Costs, below.
---------------------------------------------------------------------------

    The agencies agreed with CARB's comments that electric motors and 
batteries may be oversized. As a result, the agencies reviewed the 
sizing algorithms and methods used in the NPRM analysis and updated the 
model for the final rule analysis. The updates resulted in smaller 
motor sizes and battery pack sizes for electrified powertrains, as 
discussed above. In addition, the review also resulted in a change to 
the range categories used for the PHEVs in the final rule analysis; the 
final rule analysis classified PHEVs into two levels: (1) PHEV20 
indicating a vehicle with an AER of 20 miles; and (2) PHEV50 indicating 
a vehicle with AER of 50 miles. For more discussion on the change in 
classifications see Section VI.C.3.a)(1)(e) Electrification 
Technologies, Plug-in Hybrids.
    BorgWarner commented that ``PHEVs and HEVs are complex systems and 
should be modeled in detail,'' and further provided, ``[t]herefore, 
modeling should be inclusive of all approaches of PHEV and HEV and not 
be limited only to Atkinson Cycle engines.'' \1149\ In response, the 
agencies created additional powertrain options for PHEV technologies 
for the final rule analysis. The additional PHEV technologies included 
a plug-in HEV using a turbocharged engine. The additional PHEV paths 
used in the final rule analysis are described in Section 
VI.C.3.a)(1)(e) Electrification Technologies, Plug-in Hybrids.
---------------------------------------------------------------------------

    \1149\ BorgWarner, BorgWarner NPRM public comments 10-26-2018 
Final, Docket No. NHTSA-2018-0067-11895, at 10.
---------------------------------------------------------------------------

d) Battery Electric Vehicles
    Battery electric vehicles (BEVs) are vehicles with all-electric 
drive and with vehicle systems powered by energy-optimized batteries 
charged primarily from grid electricity. The effectiveness

[[Page 24490]]

of BEV powertrains is dependent on the efficiency of the components 
that transfer power from the battery to the driven wheels. These 
components include the battery, electric machine, power electronics, 
and mechanical gearing. For the analysis, electric machine efficiency 
was based on efficiency maps derived from actual electrified vehicles, 
and was scaled such that the peak efficiency value corresponded to the 
latest state-of-the-art technologies. The range of the battery electric 
vehicles depends on the vehicle's class and the battery pack size. For 
the NPRM analysis, manufacturers could apply BEV technology with an AER 
of 200 miles. As discussed previously, the final rule analysis added a 
BEV 300 to reflect vehicles in the market for the MY 2017 analysis 
fleet. For further detailed discussion of how BEV sub-models are 
simulated in Autonomie see the FRM Argonne model documentation.\1150\
---------------------------------------------------------------------------

    \1150\ FRM ANL Model Documentation, at 4.6, 4.7, 4.13, 4.14, and 
5.8.
---------------------------------------------------------------------------

    The resizing algorithm for BEVs is functionally the same as the 
PHEV algorithm; the difference is that BEVs do not use a combustion 
engine, and thus this component was not included in the BEV algorithm. 
To begin, initial estimates of motor and battery powers were calculated 
based on the criteria of acceleration performance, gradeability 
performance, and vehicle range. Then, the algorithm successively ran 
four simulation loops to fine tune the powertrain size to ensure that 
all performance and operational criteria were maintained. First, the 
BEV motor and battery were sized to power the vehicle through the US06 
cycle. Next, the battery capacity was adjusted to ensure the energy 
content is sufficient to complete a simulated UDDS+HWFET combined 
driving cycle, based on EPA adjustment factors to represent sticker 
values, and meet the vehicle range requirement. Finally, the electric 
motor and battery powers were sized accordingly to meet 0-60 and 50-80 
mph acceleration targets. If either acceleration simulation loop 
resulted in a change to the electric motor size, the algorithm repeated 
all simulation loops. Once the acceleration targets were met without 
any resizing of the electric motors, the algorithm finished. Figure VI-
35 below shows a simplified sizing algorithm for BEVs.
[GRAPHIC] [TIFF OMITTED] TR30AP20.214

    Meszler Engineering Services, commenting on behalf of NRDC, argued 
that the fuel economy for a vehicle adopting BEV technology was 
inappropriately dependent on the petroleum-based fuel economy of the 
transforming vehicle.\1151\ Meszler reiterated that the fuel economy of 
the internal combustion engine that BEV technology replaces does not 
have any impact on the efficiency of the resulting BEV, and the 
electric machine ``should not care'' whether it replaces a high or low 
efficiency engine, and should be modeled accordingly.
---------------------------------------------------------------------------

    \1151\ Meszler Engineering Services, Attachment 2, NPRM Docket 
No. NHTSA-2018-0067-11723 at 33.
---------------------------------------------------------------------------

    The agencies agree with Meszler that BEV effectiveness should be 
independent of the vehicle powertrain it will replace in production. 
This is, in fact, how the vehicle model and simulation was performed in 
Autonomie. Autonomie models the capabilities of each unique full 
vehicle system independently, including BEVs. As BEV technology is 
adopted by vehicles, the CAFE model uses the Autonomie databases to 
determine the added incremental efficiency that will

[[Page 24491]]

bring a specific vehicle up to the appropriate level. Since the CAFE 
model considers a variety of vehicle types with differing powertrain 
types, vehicle technology classes, performance criteria, and physical 
properties (curb weight, etc.), each with a different overall 
effectiveness, the observed efficiency increment needed to achieve BEV 
effectiveness will vary with each case. While these increments may 
differ, the final effectiveness of a BEV is independent of the 
powertrain it replaced. The effectiveness used in the CAFE model 
represents the difference between the performance of the full vehicle 
models--the full vehicle model representing the baseline vehicle and 
the full vehicle model representing the end-state with all additional 
fuel economy improving technology applied, as discussed in Section 
VI.B.3 Technology Effectiveness Values.
    ICCT alleged that the agencies did not assess BEV efficiency 
improvements from road load reductions (i.e., from mass reduction, tire 
rolling resistance, or aerodynamic improvements) to reduce the battery 
and power electronic component sizing costs.\1152\ CARB similarly 
commented that battery packs were improperly sized, resulting in 
underestimation of electrified vehicle effectiveness. CARB stated that 
the NPRM constraints on battery sizing caused electrified vehicles to 
end up with oversized, less cost-effective battery packs. CARB further 
stated that battery designs are more scalable than engines and could 
thus be adjusted by manufacturers even at incremental technology 
steps.\1153\
---------------------------------------------------------------------------

    \1152\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-82.
    \1153\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 145.
---------------------------------------------------------------------------

    For reference, battery resizing in the NPRM was constrained in the 
same manner as other powertrain components, such as the combustion 
engine. Resizing would typically be associated with a major vehicle or 
engine redesign, which in turn would justify the high costs of changing 
the powertrain. In the NPRM, the battery pack and other powertrain 
components were not resized for other improvements in incremental 
technologies such as AERO and ROLL. The agencies agree that battery 
packs, due to their modularity, should be capable of being resized at 
relatively lower cost and complexity, and thus should not be subject to 
the same resizing restrictions applied to other powertrain components 
such as conventional combustion engines. In consideration of CARB and 
ICCT's comments on battery pack resizing, for the final rule, the 
agencies allowed SHEV, PHEV, and BEV battery packs to be resized at all 
incremental technology steps, including for road load reduction 
technology improvements (aerodynamics, rolling resistance reduction, 
and low levels of mass reduction). This avoided the additional cost and 
range associated with oversized battery packs on BEVs and other 
electrified vehicles.
    CARB commented that the NPRM analysis oversized battery packs that 
targeted 200-mile label range, resulting in exaggerated battery pack 
costs. CARB also stated that some MY 2016-2018 BEVs exist that have a 
higher efficiency than simulated for BEV200s in Autonomie. They further 
argued that although these vehicles were assigned BEV200s, their actual 
range was greater than 200 miles.\1154\
---------------------------------------------------------------------------

    \1154\ California Air Resources Board, Attachment 2, Docket No. 
NHTSA-2018-0067-11873, at 147.
---------------------------------------------------------------------------

    We agree with CARB that the NPRM modeled and simulated battery 
packs were oversized and that the AERs for BEVs did not match the 
current and expected future vehicle AERs. In response to these 
comments, for the final rule analysis, the agencies removed certain 
constraints from the Autonomie battery sizing algorithm, allowing 
batteries to be sized as function of all road load reduction 
technologies. As discussed earlier, this additional battery sizing is 
feasible due to the modularity of battery pack construction. This 
update allowed the battery pack cost and mass to better reflect the 
actual required energy capacity and power, and improved the efficiency 
of modeled BEVs. The agencies also updated the modeling of electric 
machines used in BEVs to reflect improvements in efficiency. 
Furthermore, the agencies added the BEV300 (with an AER of 300 miles) 
to the final rule analysis, providing a better representation of 
production BEVs with more than 200 miles of range. For more discussion 
on BEV300 and electrification efficiency improvements, see Sections 
VI.C.3.a)(1) Electrification technologies and VI.C.3.d)(1) Electric 
Motors, Power Electronics and Accessory Load.
e) Fuel Cell Vehicles
    The fuel-cell system in the analysis was modeled to represent 
hydrogen consumption as a function of the produced power, assuming 
normal-temperature operating conditions with a peak system efficiency 
of 60 percent, including the balance of plant.\1155\ The system's 
specific power is 650 W/kg. The hydrogen storage technology selected 
was a high-pressure tank with a specific weight of 0.04 kg H2/kg, sized 
to provide a 320-mile range on the 2-cycle tests on the basis of 
adjusted energy values.
---------------------------------------------------------------------------

    \1155\ Power needed for supporting components and auxiliary 
systems. The balance of plant in a fuel cell system is the auxiliary 
equipment required to ensure the fuel cell operates as a reliable 
power source. This may include fuel reformers and pumps, for 
example.
---------------------------------------------------------------------------

    The sizing algorithm for FCVs was similar to PHEVs and BEVs, but 
adapted to size the specific components of a FCV powertrain: the 
electric motor, fuel-cell, hydrogen (H2) fuel tank, and 
battery pack. The electric motor drives the wheels needed to propel the 
vehicle. During very low power operation, the battery pack alone powers 
the motor/wheels, depleting the battery charge. At moderate driving 
loads, the fuel-cell provides electrical power (generated by consuming 
stored H2) to the motor and also to charge the battery. 
Under heavy loads, both the fuel cell and battery deliver electric 
power to the motor. To begin, initial estimates of motor, fuel cell, 
and battery powers are calculated based on criteria for acceleration 
performance, gradeability performance, and vehicle range. Then, the 
algorithm successively runs four simulation loops to finetune 
powertrain size, ensuring that all performance and operational criteria 
are maintained. First, the FCV motor and battery are sized to power the 
vehicle through the US06 cycle. Next, the on-board mass of H2 fuel, as 
well as the fuel tank mass are adjusted to ensure the vehicle can 
complete a simulated 2-cycle test and meet the range requirement. 
Finally, the electric motor and fuel cell powers are sized accordingly 
to meet 0-60 and 50-80 mph acceleration targets. If either acceleration 
simulation loop results in a change to the electric motor size, the 
algorithm repeats all simulation loops. Once the acceleration targets 
can be met without any resizing of the electric motor, the algorithm 
completes. Figure VI-36 below shows a simplified sizing algorithm for 
FCVs.

[[Page 24492]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.215

    The agencies did not receive comments on FCV modeling in Autonomie. 
For the final rule analysis, the agencies used the same FCV model and 
simulations to estimated effectiveness values.
e) Electrification Costs
    The primary factors that influence the cost and effectiveness of 
hybrid or battery electric vehicles are the cost and efficiency of the 
energy storage components and electric machines. Energy storage 
components include battery cells, battery management systems, and 
thermal management systems. The electric machine components include 
electric motors, power electronics, controllers, and other devices that 
support thermal management.
    Charging infrastructure is an essential component for PHEVs and 
BEVs, and may add to the total cost of ownership of the vehicle. 
However, most households are equipped with a 110-volt outlet for level 
1 charging, for which no additional cost is incurred. Installing a 
level 2 charging outlet (220-volt) will add cost to the total ownership 
of the vehicle but decreases charging time. The price of level 2 
residential charging equipment varies, but typically ranges from $500 
to $2,000 before installation and state or utility incentives.\1156\
---------------------------------------------------------------------------

    \1156\ U.S. Department of Energy Office of Energy Efficiency and 
Renewable Energy, Charging at Home, https://www.energy.gov/eere/electricvehicles/charging-home (last visited March 20, 2020).
---------------------------------------------------------------------------

    For this final rule analysis, the agencies used Argonne's BatPaC 
modeling tool to develop battery pack manufacturing costs as well as 
weight.\1157\ Battery packs were sized in terms of the vehicle's energy 
and power requirement and costs were estimate for each of the simulated 
technology combinations. The Argonne team used BatPaC to create a 
``lookup table'' with battery pack size (energy and power) and cost as 
well as weight data for the full vehicle simulations to ``reference,'' 
to avoid the need for conducting a full BatPaC simulation for each 
unique vehicle modeled in the analysis. The table included cost data 
for each technology key and vehicle technology classes. As discussed 
below, Autonomie runs linearly interpolate between points in the lookup 
tables when deriving final values from BatPaC, the differences between 
using BatPaC for each configuration and the interpolation using the 
lookup table was insignificant.
---------------------------------------------------------------------------

    \1157\ The agencies used BatPaC version 3.0 (released in 2015) 
for the NPRM and BatPaC version 3.1 (June 2018) for the final rule.
---------------------------------------------------------------------------

    The agencies used the cost of electric machines from U.S. DRIVE's 
October 2017 report, ``Electrical and Electronics Technical Team 
Roadmap.'' In industry, manufacturers use different types of electric 
machines resulting in a range of actual costs for the systems. To 
capture this range, the agencies considered a single type of high 
efficiency electric machine, representative of the range of technology 
available in the rulemaking timeframe, uniquely sized for each of the 
simulated combinations. For the final rule analysis, the cost of the 
electric machine was determined using a dollar-per-kilowatt metric. The 
agencies sized the electric machines using the method discussed in 
Section VI.C.3.d) Electric Effectiveness Modeling and Resulting 
Effectiveness Values.
    The following sections discuss the method used for modeling battery 
and non-battery component costs, the learning curves applied to those 
costs, and the total costs for each type of electrification technology 
considered in this final rule analysis.
(l) Battery Pack Modeling
    BatPaC is a software designed for policymakers and researchers 
interested in estimating the manufacturing cost of lithium-ion 
batteries for electric drive

[[Page 24493]]

vehicles.\1158\ BatPaC is used to estimate the cost of manufacturing 
lithium-ion batteries and examine trade-offs that result from different 
battery performance specifications such as power and energy capacity. 
BatPaC includes a library of lithium ion electrode combinations and 
inputs for all the parameters associated with materials and 
manufacturing operations in a factory.
---------------------------------------------------------------------------

    \1158\ BatPaC: Battery Manufacturing Cost Estimation, Argonne 
National Laboratory, https://www.anl.gov/tcp/batpac-battery-manufacturing-cost-estimation.
---------------------------------------------------------------------------

    Specifically, BatPaC models stiff-pouch, laminated prismatic format 
cells, placed in double-seamed, rigid modules. The model supports 
liquid- and air-cooling, accounting for the resultant structure, 
volume, cost, and heat rejection capacity. The model considers cost of 
capital equipment, plant area and labor for each step in the 
manufacturing process. The model places relevant limits on electrode 
coating thickness, and considers limits applicable to current and near-
term manufacturing processes. The model also considers annual pack 
production volumes and economies of scale for high-volume production.
    BatPaC calculations are based on a generic pack designs that 
reasonably represents the weight and manufacturing cost of batteries 
deployed commercially. The advantage of using this approach is the 
ability to model wide range of commercial design specifications for the 
various classes of vehicles. This modeling approach is particularly 
advantageous because the data from commercially available battery packs 
is limited and varies widely with respect to the underlying 
specifications (power and energy) and constraints (mass, volume, 
dimensions, durability) set by the manufacturer.
    BatPaC is a Microsoft Office Excel spreadsheets-based model. The 
data needed to design and build a battery pack, such as dimensions of 
the cell, estimate of materials, and manufacturing cost, are provided 
in the model, with the manufacturing costs for the designed battery 
based on a ``baseline plant'' designed for a battery of intermediate 
size and production scale so as to establish a center-point for other 
designs. BatPaC can be configured with alternative chemistries, 
charging constraints, battery configurations, production volumes, and 
cost factors for other battery designs by customizing these parameters 
in the modeling tool.
    For this analysis, running individual BatPaC simulations for each 
full vehicle simulation requiring an electrified powertrain would have 
been computationally intensive and impractical, given that 
approximately 750,000 simulated vehicles out of the 1.2 million total 
simulated vehicles had an electrified powertrain. Accordingly, staff at 
Argonne built ``lookup tables'' with BatPaC to provide battery pack 
manufacturing costs, battery pack weights, and battery pack cell 
capacities for vehicles modeled in the large-scale simulation runs.
    To build the lookup tables, Argonne staff selected a range of 
minimum and maximum values for battery pack power (kW) and battery pack 
energy (kWh) for each vehicle powertrain based on a combination of 
market analysis and analysis of the Autonomie simulations that were run 
for the NPRM and final rule. The performance requirements (vehicle 
acceleration times, EV range, etc.) were defined from set assumptions 
and validated from existing vehicles.\1159\ The range, as well as the 
number of power and energy points considered to generate each lookup 
table, varies across powertrains. The minimum and maximum power and 
energy values have been selected to encompass current designs. For 
example, one end of the spectrum is representative of the MY 2016-2017 
Tesla Model S 100D (100 kWh total battery energy, 335-mile range), 
while the other end of the spectrum is representative of the 2017 
Mitsubishi iMiEV (16 kWh total battery energy, 62-mile range). The 
components were then sized in Autonomie across all vehicle classes to 
define the minimum and maximum values to be considered, as shown in 
Table VI-90.
---------------------------------------------------------------------------

    \1159\ See Final Rule Argonne Model Documentation Section 5.9, 
Battery Performance and Cost Model (BatPaC).
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.216

    Figure VI-37 illustrates the inputs generated in Autonomie to 
create the BatPaC-based lookup tables, and the outputs characterized in 
the BatPaC-based lookup tables that are used to provide estimates 
referenced in the agencies' analysis. A linear interpolation was then 
performed in MATLAB to determine the associated values for battery pack 
manufacturing cost, weight, and cell capacity.

[[Page 24494]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.217

    Figure VI-38 shows the linear relationship between cost, power, and 
weight used to generate the compact passenger car BEV200 technology 
class lookup table presented in Figure VI-39. As seen from the figures 
below, the energy values produced by BatPaC consist of a fairly linear 
relationship with respect to power and energy for a vehicle class. 
Since Autonomie runs would linearly interpolate between the points in 
the lookup tables when deriving the final values from BatPaC, the 
differences between using BatPaC for each configuration and the 
interpolation using the lookup table were insignificant.
[GRAPHIC] [TIFF OMITTED] TR30AP20.218

BILLING CODE 4910-59-C
    Figure VI-39 details the estimates of $ per kWh at the pack level 
generated from the lookup table for BEV200 compact cars used in the 
final rule analysis. As discussed further below, the specific battery 
costs for each simulated vehicle were presented for the NPRM (and now 
for the final rule) in the docketed Argonne assumptions files and in 
the vehicle simulation database included in the CAFE model.

[[Page 24495]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.219

    During the Autonomie large-scale simulation runs, calling the 
BatPaC model for each individual simulation would have been 
computationally intensive. Using the MATLAB lookup tables reduced the 
time to run the approximately 750,000 simulations significantly, which 
in turn reduced the total simulation run time for all of the technology 
combinations by several days with insignificant impact on the 
analytical results.
(a) BatPaC Inputs and Assumptions
    The Argonne documentation describing the analysis performed for the 
NPRM, ``A Detailed Vehicle Simulation Process To Support CAFE 
Standards,'' detailed the specific assumptions that Argonne's experts 
used to simulate batteries and their associated costs for the full 
vehicle simulation modeling.\1160\ In addition, detail on the NPRM 
electrification analysis was presented in the PRIA.\1161\ While the 
Argonne Summary of Main Component Assumptions Excel file correctly 
identified the chemistry used in the NPRM analysis as NMC333,\1162\ the 
PRIA inadvertently described that NMC441 was used. The agencies 
presented selected lookup table battery cost values in the Argonne 
Summary of Main Component Assumptions Excel file,\1163\ as shown above, 
and the specific battery costs for each simulated vehicle were 
presented for the NPRM and final rule in the vehicle simulation 
database included in the CAFE model.
---------------------------------------------------------------------------

    \1160\ Islam S. Ehsan. Moawad, Ayman. Kim, Namdoo. Rousseau, 
Aymeric. ``A Detailed Vehicle Simulation Process to Support CAFE 
Standards.'' ANL/ESD-18/6. Energy Systems Division, Argonne National 
Laboratory (2018).
    \1161\ PRIA at 362-384.
    \1162\ ANL--All Assumptions Summary, NHTSA-2018-0067-0005.
    \1163\ ANL--Summary of Main Component Performance Assumptions 
NPRM, NHTSA-2018-0067-0003.
---------------------------------------------------------------------------

    Several commenters claimed that costs for electrification 
technologies were too high, especially regarding battery costs (note 
that comments on non-battery component costs are addressed separately 
in Section VI.C.3.e)(2) Non-battery Electrification Component Costs, 
below).\1164\ Several commenters pointed to text in interagency review 
documents that stated the NPRM battery modeling costs were higher than 
what EPA recommended,\1165\ and higher than what EPA had obtained from 
the most recent version of the BatPaC model.\1166\
---------------------------------------------------------------------------

    \1164\ Meszler Engineering Services, NHTSA-2018-0067-11723 
Attachment 2; National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969; Workhorse Group Inc., NHTSA-2018-0067-12215; 
International Council on Clean Transportation, NHTSA-2018-0067-
11741; California Air Resources Board, NHTSA-2018-0067-11873.
    \1165\ California Air Resources Board, NHTSA-2018-0067-11873.
    \1166\ Boulder County Public Health et al., NHTSA-2018-0067-
11975; International Council on Clean Transportation, NHTSA-2018-
0067-11741.
---------------------------------------------------------------------------

    CARB commented that the agencies incorrectly identified and 
assessed existing technologies, improperly oversized components and 
batteries for the modeled vehicle classes, and underestimated 
technology efficiency through improper modeling.\1167\ CARB also 
submitted supplemental comments (discussed further, below) stating that 
the PRIA and the underlying modeling were inconsistent regarding which 
exact battery chemistries were modeled for every electrified model in 
the fleet, which CARB argued was crucial for understanding the battery 
compositions and thus their production costs.\1168\
---------------------------------------------------------------------------

    \1167\ California Air Resources Board, NHTSA-2018-0067-11873.
    \1168\ California Air Resources Board, NHTSA-2018-0067-4166.
---------------------------------------------------------------------------

    ICCT stated that the agencies misrepresented the leading research 
on both battery and electric vehicle costs, with the result being that 
electric vehicles were so costly that they were modeled to remain at 
approximately the same penetration in 2025 with the Augural 2025 fuel 
economy and adopted 2025 CO2 standards, as they were in mid-
2018 (i.e., between 1.5 percent and 2 percent of new vehicle 
sales).\1169\ ICCT stated that the agencies' inputs failed to reflect 
the latest industry data on future potential electric vehicle cost 
parity with combustion vehicles. ICCT commented that through a 
combination of incorrectly high electric vehicle prices (which, they 
argue, do not reflect Argonne or other leading battery research groups' 
work), and modeling restrictions on electric vehicles, the agencies 
unduly inflated technology costs of electric vehicles to comply with 
the standards. ICCT argued that although the agencies purported to use 
state-of-the-art tools like the BatPaC model for battery costs, the 
cost calculations erroneously pushed up electric vehicles' incremental 
costs above $10,000 per vehicle. ICCT claimed that the agencies 
introduced errors that artificially pushed up the battery costs higher 
than indicated by BatPaC and other experts in the field.
---------------------------------------------------------------------------

    \1169\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
---------------------------------------------------------------------------

    NCAT noted that the PRIA described some ways in which the modeling 
increased battery costs, namely, that the battery pack costs were 
adjusted

[[Page 24496]]

upwards, the cost of the battery management system increased, and a 
cost for a battery automatic and manual disconnect unit was 
added.\1170\ Regardless, NCAT stated that the agencies analysis was not 
sufficiently transparent, and argued that the battery costs were 
significantly overestimated in the modeling supporting the NPRM. 
Boulder County Public Health and other Colorado municipal organizations 
claimed that overstated battery costs had the effect of 
mischaracterizing and downplaying the benefits of increased numbers of 
electric vehicles as part of the vehicle fleet.\1171\ Commenters also 
argued that discrepancies existed between battery costs used in the 
rulemaking documents and battery costs found in the Argonne database, 
referring specifically to BISG and CISG costs (discussed further 
below).\1172\
---------------------------------------------------------------------------

    \1170\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing PRIA at 366-67.
    \1171\ Boulder County Public Health et al., NHTSA-2018-0067-
11975.
    \1172\ Meszler Engineering Services, NHTSA-2018-0067-11723 
Attachment 2; International Council on Clean Transportation, NHTSA-
2018-0067-11741.
---------------------------------------------------------------------------

    In addition to comments claiming that the agencies' battery cost 
projections were incorrect or difficult to interpret, many commenters 
submitted general information about the state of battery technology and 
cost advances now and as projected into the future. For example, NCAT 
stated that battery technology has improved and battery costs have 
fallen dramatically, due in part to reduced material costs, 
manufacturing improvements, and higher manufacturing volumes.\1173\ In 
compliment, NCAT asserted that the demand for EVs is growing 
``dramatically.''
---------------------------------------------------------------------------

    \1173\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969. NCAT also stated that the increase in mass 
manufacturing of lithium-ion storage is expected to continue to 
reduce battery prices.
---------------------------------------------------------------------------

    ICCT stated that the agencies' analysis of electric vehicle costs 
and the resulting extremely low penetration levels was not in line with 
automakers' announcements, which included statements that they would 
produce far greater numbers of electric vehicles to comply with 
standards around the world.
    ICCT summarized projections of electric vehicle battery costs for 
2020-2030, and stated that the agencies did not analyze the studies and 
automaker announcements they cited to understand the potential for 
cost-effective electric drive technology.\1174\ ICCT stated the data 
they reviewed included a variety of different technologies, production 
volumes, and cost elements, and although there were differences in 
methods for each, ``they generally include in some variation of 
material, process, overhead, depreciation, warranty, and profit 
costs.'' ICCT summarized the results of their review, projecting that 
battery pack costs will decline to $150/kWh by 2020-2023 and then to 
about $120-$135/kWh by 2025, with the exception of Tesla, which reports 
costs of $150 kWh in 2018 and projected costs of $100/kWh by 2022. ICCT 
stated that the results of this review were corroborated in the 
aforementioned EPA interagency comments on battery costs used in the 
proposal.
---------------------------------------------------------------------------

    \1174\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
---------------------------------------------------------------------------

    NCAT stated that the average price of a battery pack dropped from 
$1,000/kWh in 2010 to $209/kWh in 2017, demonstrating a decrease of 79 
percent in seven years.\1175\ NCAT stated Tesla is on track to achieve 
$100/kWh by the end of 2018, and Audi has been buying batteries at 
$114/kWh, according to trade press reports.\1176\ NCAT also cited BNEF 
analyses showing that battery costs are projected to continue to 
decline substantially,\1177\ specifically projecting a decrease in 
battery cost of 77 percent between 2016 and 2030. Accordingly, NCAT 
stated that EVs will be less expensive to buy than conventional 
gasoline vehicles by 2025 in the United States.\1178\ Workhorse 
similarly echoed the assertion that EV costs will reach parity with 
conventional vehicle costs before 2025.\1179\
---------------------------------------------------------------------------

    \1175\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Bloomberg New Energy Finance, ``Electric 
Vehicle Outlook: 2018,'' https://bnef.turtl.co/story/evo2018?teaser=true.
    \1176\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Fred Lambert, ``Tesla to achieve leading 
$100/kWh battery cell cost this year, says investor after 
Gigafactory 1 tour'' (Sept. 11, 2018), https://electrek.co/2018/09/11/tesla-100-kwh-battery-cost-investor-gigafactory-1-tour/.
    \1177\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Bloomberg New Energy Finance, ``Electric 
Vehicle Outlook: 2018,'' https://bnef.turtl.co/story/evo2018?teaser=true.
    \1178\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Jess Shankleman, ``Pretty Soon Electric Cars 
Will Cost Less Than Gasoline'' (May 26, 2017), https://www.bloomberg.com/news/articles/2017-05-26/electric-cars-seen-cheaper-than-gasoline-models-within-a-decade; Jess Shankleman, ``The 
Electric Car Revolution Is Accelerating'' (July 6, 2017), https://www.bloomberg.com/news/articles/2017-07-06/the-electric-car-revolution-is-accelerating. NCAT also noted that the up-front cost 
parity does not take into consideration the fuel savings and 
maintenance savings over the lifetime of EV use as compared to 
gasoline vehicle use.
    \1179\ Workhorse Group Inc., NHTSA-2018-0067-12215.
---------------------------------------------------------------------------

    NCAT also cited the ICCT Efficiency Technology and Cost Assessment, 
which concluded that, primarily because of rapid developments in 
battery pack technologies, EV costs will be reduced by $4,300-$5,300 
per vehicle by 2025 compared to EPA's prior estimates in support of the 
MY 2017-2025 standards.\1180\ In that report, ICCT concluded that 
battery costs of $140/kWh is a realistic estimated value by 2025, as 
compared with EPA estimates in the 2016 Mid-Term Evaluation (MTE) 
analysis of $180-200/kWh.\1181\
---------------------------------------------------------------------------

    \1180\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing ICCT, ``Efficiency Technology and Cost 
Assessment for U.S. 2025-2030 Light-duty Vehicles'' (Mar. 2017) at 
11, 15, available at http://www.theicct.org/US-2030-technology-cost-assessment.
    \1181\ Id.
---------------------------------------------------------------------------

    NCAT also cited improvements in manufacturing techniques, 
specifically by Tesla, as an example of how batteries are being 
manufactured in large volumes with high quality at low cost.\1182\ NCAT 
stated that in mid-2018, Tesla was producing batteries at its 
Gigafactory 1 facility at an annualized rate of roughly 20 GWh, making 
it the highest-volume battery plant in the world.\1183\ NCAT and other 
commenters also cited Bloomberg's New Energy Finance research stating 
that the average energy density of EV batteries is improving at around 
5-7 percent per year.
---------------------------------------------------------------------------

    \1182\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Tesla, Inc., S.E.C. Form 10-K (Feb. 22, 
2018) at 3-4, available at https://www.sec.gov/Archives/edgar/data/1318605/000156459018002956/tsla-10k-20171231.htm.
    \1183\ National Coalition for Advanced Transportation, NHTSA-
2018-0067-11969, citing Tesla, ``Tesla Gigafactory,'' https://www.tesla.com/gigafactory (last visited Oct. 25, 2018).
---------------------------------------------------------------------------

    Finally, Workhorse commented that they have more than ten years of 
experience in the field of designing and assembling battery packs, and 
their business plans are predicated on battery costs much lower than 
assumed by the agencies.\1184\
---------------------------------------------------------------------------

    \1184\ Workhorse Group Inc., NHTSA-2018-0067-12215.
---------------------------------------------------------------------------

    As explained above, the agencies consulted with and relied on 
Argonne battery experts to develop inputs to the BatPaC model and 
generate the battery cost lookup tables used as references for the 
Autonomie full-vehicle simulations, as detailed in Argonne's 
documentation supporting the NPRM analysis.\1185\ As explained further 
below, the agencies also directed CARB to information about the NPRM 
battery cost analysis available

[[Page 24497]]

in the public docket in response to their FOIA request.
---------------------------------------------------------------------------

    \1185\ Islam S. Ehsan. Moawad, Ayman. Kim, Namdoo. Rousseau, 
Aymeric. ``A Detailed Vehicle Simulation Process to Support CAFE 
Standards.'' ANL/ESD-18/6. Energy Systems Division, Argonne National 
Laboratory (2018).
---------------------------------------------------------------------------

    Commenters are correct that the EPA Draft TAR and Proposed 
Determination estimates for battery sizing and cost were different than 
the NPRM analysis. For the Draft TAR and in the Proposed Determination, 
a separate battery and motor sizing spreadsheet was built to determine 
the energy and power requirements for PHEVs and BEVs at different curb 
weights, and then BatPaC was used to determine specific energy (kWh/kg) 
and the battery pack cost estimate.\1186\ For this NPRM and final rule, 
the energy requirements for PHEVs and BEVs were determined using 
Autonomie simulations with the integrated BatPaC lookup table to select 
the appropriate battery pack size, cost, and weight. As discussed in 
Sections VI.B.3.a)(4) How Autonomie Sizes Powertrains for Full Vehicle 
Simulation and VI.B.3.a)(6) Performance Neutrality, the Autonomie full-
vehicle simulation modeling assessed metrics to ensure performance 
requirements were met for every modeled vehicle. Appropriately 
accounting for vehicle metrics and individual vehicle power and weight 
requirements resulted in some of the differences observed between the 
Draft TAR and Proposed Determination estimates and the estimates 
presented in the NPRM and this final rule.
---------------------------------------------------------------------------

    \1186\ Draft TAR at 5-315.
---------------------------------------------------------------------------

    For the final rule, the agencies considered these public comments, 
market observations, literature, industry reports, and additional 
research. In addition, as described further below and in the Argonne 
documentation accompanying this final rule, Argonne consulted the 
A2Mac1 database for additional data points on batteries that were used 
to inform the final rule battery cost modeling.
    As discussed above, BatPaC version 3.0 was used for the NPRM 
analysis because that was the most up-to-date version of BatPaC 
available at the time the NPRM analysis was being conducted. BatPaC 
version 3.1, released after the NPRM analysis was completed, was used 
for this final rule because that was the most up-to-date version of 
BatPaC available at the time the final rule analysis was being 
conducted.
    The agencies note that BatPaC version 4.0 has been released since 
the analysis was completed for this final rule. Specifically, that 
version was released on January 14, 2020, after the rule had been 
submitted for interagency review. The default battery chemistry in 
BatPaC version 4.0 continues to be NMC622, which as discussed further 
in Section (i) below, reflects the reasonable assumption this chemistry 
will likely continue to be used in the rulemaking timeframe based on 
its commercial application and market trends towards higher-nickel, 
lower-cobalt content chemistries.\1187\ As explained in this section, 
and further in Section (c) below, the agencies' modeled costs for 
battery packs aligns with current industry estimates and closely tracks 
future projections of battery pack costs from the Department of 
Energy's Vehicle Technology Office (DOE VTO) lab 
targets.1188 1189
---------------------------------------------------------------------------

    \1187\ The agencies note that BatPaC version 4.0 provides a new 
option to build battery packs with NMC811.
    \1188\ Freyermuth, Vincent. Rousseau, Aymeric. ``Impact of 
Vehicle Technologies Office Targets on Battery Requirements.'' ANL/
ESD-16/22. Energy Systems Division, Argonne National Laboratory 
(2016).
    \1189\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
---------------------------------------------------------------------------

    In addition to using BatPaC version 3.1 for this final rule, BatPaC 
assumptions were updated to reflect what the Argonne battery experts 
and the agencies believed would be representative and attainable of 
battery manufacturing trends in the rulemaking timeframe. Section (ii) 
provides additional information on BatPaC inputs and assumptions that 
were updated for the final rule based on public comments and the 
agencies own market observations and additional research. In addition, 
as discussed further below, for the final rule, the calculated battery 
pack weight and manufacturing cost was compared with the battery pack 
cost and weight data obtained through various benchmarking studies. The 
agencies believe that the Argonne methodology for producing the 
hundreds of thousands of battery pack cost estimates required for the 
full-vehicle modeling and simulation resulted in reasonable estimates 
of battery pack costs. The following sections provide additional 
context and response to comments on specific BatPaC inputs and 
assumptions used in the NPRM and final rule.
(i) Chemistry
    The choice of chemistry for battery cells depends on the 
application and consideration of cost, energy density, and safety, 
among other factors. The PRIA described the battery pack cell chemistry 
used for different powertrain types modeled in the NPRM analysis.\1190\ 
For Micro HEVs, BISG HEVs, CISG HEVs, and Full HEVs, the agencies used 
LFP-G, rather than LMO-G, because the latter has a limited lifespan 
which is expected to degrade functionality over a vehicle's lifetime, 
and has greater limitations on available ranges of battery charge and 
discharge rates. As described above, for PHEVs and BEVs, the Argonne 
``Summary of Main Component Performance Assumptions'' file correctly 
stated that NMC333 was used, however the PRIA misstated that NMC441 was 
used.
---------------------------------------------------------------------------

    \1190\ PRIA at 373.
---------------------------------------------------------------------------

    Both UCS and CARB commented on the agencies' choice of battery 
chemistry, with UCS noting that this choice can have a large impact on 
performance and materials costs, and therefore on the modeled cost of 
drivetrain electrification.
    First, both commenters stated that the NPRM documentation was 
inconsistent and unclear. UCS noted the discrepancy between the PRIA 
and Argonne model documentation, and also that the rulemaking documents 
stated the most recent version of Argonne's BatPaC model was used to 
estimate battery costs, but the default lithium ion chemistry in the 
current BatPaC model is NMC622. UCS stated the choice of NMC variant 
effects battery costs, as NMC622 replaces more expensive cobalt with 
nickel. UCS further stated it was not possible to determine the 
magnitude of the cost error in the PHEV and BEV battery pack costs, 
only that the costs were likely higher than current battery cost data 
supported.
    CARB stated that the agencies' selected battery chemistries 
represented a step backward from previous analysis done for the Draft 
TAR. CARB claimed that the biggest lithium-ion production companies 
have indicated that they will use NMC811 for BEVs, and therefore NMC441 
or NMC333 would not represent current technology going into BEVs or 
near-future BEV battery technology. CARB stated that NMC811 technology 
was expected to come to market in 2019, which is far sooner than 
anticipated, even in the agencies' prior analyses.
    Commenters also noted that the chemistry chosen for mild and strong 
hybrids differed from what is used in current and announced HEVs. UCS 
stated that all non-plug-in hybrids in the proposed rule analysis used 
lithium iron phosphate (LFP) chemistry, but in practice, most hybrids 
on the road did not use this chemistry. UCS referenced the Toyota Prius 
and the new RAM 1500 pickup as examples of vehicles that do not use LFP 
chemistry. CARB similarly stated that the NPRM battery chemistry 
selection for PHEV and strong hybrid batteries does not represent many 
of the batteries that are being deployed in the market, nor have been, 
for several years now, but did not provide an alternative chemistry 
they believed to be better

[[Page 24498]]

represented in the market. CARB stated that this resulted in a 
``misappropriation of higher costs for electrification technologies in 
the Agencies' analysis, and further highlights the Agencies' sudden 
lack of knowledge about electrification, despite the far more 
directionally correct projections in previous analysis for the 2016 
Draft TAR and EPA's Proposed Determination.''
    Similarly, UCS pointed to a discrepancy in strong hybrid battery 
costs between the proposed rule estimates (greater than $1,200, even 
for the small car classes) and an estimate from Argonne in 2017 ($614), 
to argue that the lack of detailed information made it impossible to 
determine if the choice of battery chemistry was responsible for the 
discrepancy.
    The agencies carefully considered these comments. As stated above, 
the agencies disagree that the discrepancy in the Argonne Summary of 
Main Component Performance Assumptions file and the PRIA over the use 
of NMC333 for the NPRM analysis limited commenters ability to comment 
on battery chemistry, as both UCS and CARB communicated a belief that 
the agencies choice of battery chemistry contributed to the overstated 
battery costs in the NPRM. The agencies understand how the choice of 
chemistry impacts battery costs, and many of the commenters' concerns 
intertwined the NPRM choice of battery chemistry with the NPRM battery 
costs. Here, the agencies respond to comments on the choice of 
chemistries. The agencies will also discuss costs below.
    As stated earlier, although manufacturers use different battery 
chemistries in various HEV, PHEV, and BEV applications, the choice of 
chemistry for a given application depends on several factors including 
safety, stability, and functional requirements (high power or high 
energy requirements for performance) of the battery pack. In 
determining whether to select one battery chemistry over another, the 
agencies concluded that using commercially proven technologies that 
represented the current cost of production was more reasonable than 
assuming additional technologies would come to fruition during the 
rulemaking timeframe, and attempting to project the cost and 
effectiveness of such technologies. While there is ongoing research and 
development in battery chemistry and in other battery related 
technologies that have the potential to reduce costs and increase 
battery capacity, these technologies have yet to be proven viable for 
commercial use.\1191\
---------------------------------------------------------------------------

    \1191\ Recent Advances in Energy Chemical Engineering of Next-
Generation Lithium Batteries, Engineering, Volume 4, Issue 6 
(December 2018), at 831-847. Available at https://www.sciencedirect.com/science/article/pii/S2095809918312177. Some 
examples include lithium-sulfur battery cell chemistry and solid-
state electrolyte battery cells.
---------------------------------------------------------------------------

    In addition, as discussed throughout this document, the agencies 
considered technologies that manufacturers could use to comply with 
standards in the rulemaking timeframe that reasonably represented the 
state of technology across the industry. While the battery chemistries 
used in commercial vehicles are largely confidential business 
information, proprietary teardown reports are one source of information 
used to learn more about the chemistries actually employed in the 
market. For both the NPRM and final rule, the agencies consulted 
Argonne's battery experts to determine the chemistries that should be 
modeled in the BatPaC analysis. Argonne consulted A2Mac1 battery pack 
teardown reports, which confirmed that indeed, manufacturers use a 
range of chemistries across the electrified vehicle types. Selecting 
battery chemistries that can reasonably represent the range employed in 
the market ensured that the analysis better captured the average of 
costs across the industry.
    For example, in addition to the reasons listed in the NPRM, LFP has 
been proven in commercial use, as identified in literature and battery 
teardown reports.\1192\ This presented a basis for using LFP, as the 
chemistry was reasonably representative of chemistries used in mild and 
strong hybrids at the time of the analysis. The agencies also 
considered that LFP's lower cost compared to other potential HEV 
battery chemistries (contrary to commenters' statements) made it more 
attractive for vehicles with tight cost constraints, even with the 
associated lower energy density.
---------------------------------------------------------------------------

    \1192\ Details of cell chemistry and battery cooling system are 
described in Nelson, Paul A., Gallagher, Kevin G., Bloom, Ira D., 
and Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles--SECOND EDITION (2012), 
available at https://publications.anl.gov/anlpubs/2015/05/75574.pdf.
---------------------------------------------------------------------------

    Similarly, although EPA selected NMC622 as the modeled battery 
chemistry for the Draft TAR, manufacturers were also using other NMC 
chemistries in hybrid and BEV applications in that timeframe depending 
on the required application. The chemistry selected for the NPRM, 
NMC333, was selected based on proprietary teardown reports that 
demonstrated the chemistry's commercial use: a survey of twelve MY 2013 
to MY 2018 HEVs, PHEV, and BEVs showed that NMC333 was used in eleven 
of those vehicles, and NMC622 was only used in one.\1193\
---------------------------------------------------------------------------

    \1193\ A Detailed Vehicle Simulation Process To Support CAFE and 
CO2 Standards for the MY 2021--2025 Final Rule Analysis, 
Section 5.9 Battery Performance and Cost Model (BatPaC), referencing 
A2Mac1 Automotive Benchmarking, https://a2mac1.com.
---------------------------------------------------------------------------

    Accordingly, the agencies believe that assuming LFP-G as the 
modeled cell chemistry for HEVs and NMC333 as the modeled PHEV and BEV 
chemistry for the NPRM analysis of battery costs was not unreasonable, 
based on their demonstrated commercial use in a range of electric 
vehicle applications. However, employing BatPaC version 3.1 for the 
final rule analysis also presented the opportunity to update the 
modeled battery chemistry used to assess battery costs.
    The agencies similarly consulted Argonne battery experts on battery 
chemistry and trends to inform the final rule analysis. Argonne staff 
used the A2Mac1 database to determine real-world battery chemistry and 
configurations in different electric vehicle applications. As shown in 
the Argonne Full Vehicle Modeling documentation for the final rule, the 
A2Mac1 battery pack teardown analysis provided an array of data points 
on battery chemistries for different electric vehicle applications, 
among other relevant battery pack data, that informed the final rule 
battery analysis.\1194\
---------------------------------------------------------------------------

    \1194\ Id.
---------------------------------------------------------------------------

    In determining which of these chemistries would best represent the 
range of chemistries demonstrated in the market, the agencies 
considered several issues. Due to the increasing manufacturing volume 
of battery packs with NMC, it is expected that NMC battery cells will 
continue to be used in battery packs across different electric vehicle 
applications in the future. The agencies considered concerns about NMC 
formulations with varying cobalt content, and issues including the 
current and future cost of cobalt,\1195\

[[Page 24499]]

and the cobalt supply chain.\1196\ These concerns, among others, have 
led to the market shift towards cathode active materials with a higher 
fraction of nickel and less cobalt.\1197\ Manufacturers have 
demonstrated the use of NMC622, which contains more nickel and less 
cobalt than NMC333, in different electric vehicle applications. In 
addition, as CARB noted and has been reported in the news for some 
time, the expected next step in battery chemistries using even less 
cobalt is NMC811. However, the shift to higher-nickel-content 
chemistries is not without challenges; increasing nickel content 
results in lower thermal stability, leading to safety concerns.\1198\
---------------------------------------------------------------------------

    \1195\ See, e.g., MIT Energy Initiative. 2019. Insights into 
Future Mobility, at 78. Cambridge, MA: MIT Energy Initiative (``. . 
. significant uncertainty remains about the steady-state price of 
cobalt in the future as demand and supply continues to increase 
[internal citation omitted]. Under our base case scenario, global 
demand for cobalt in 2030 from new EV sales (even if all EVs use 
batteries with the high nickel content of NMC811) would reach 
approximately 80% of the world's total cobalt output in 2016. 
Considering that only 15% of the worldwide demand for cobalt in 2017 
was used in EV batteries (Jackson 2019), an increase in demand of 
this magnitude might result in higher prices for cobalt. Thus, 
automakers may need to move to different battery chemistries that 
are less reliant on cobalt to avoid raw materials shortages and 
price volatility.'').
    \1196\ See, e.g., Todd C. Frankel, The Cobalt Pipeline: Tracing 
the path from deadly hand-dug mines in Congo to consumers' phones 
and laptops, Washington Post (Sept. 30, 2016), https://www.washingtonpost.com/graphics/business/batteries/congo-cobalt-mining-for-lithium-ion-battery/?itid=lk_inline_manual_9&tid=lk_inline_manual_9; Peter Whoriskey and 
Todd C. Frankel, Tech giants pledge to keep children out of cobalt 
mines that supply smartphone and electric-car batteries, Washington 
Post (Dec. 20, 2016), https://www.washingtonpost.com/news/the-switch/wp/2016/12/20/tech-giants-pledge-to-keep-children-out-of-cobalt-mines-that-supply-smartphone-and-electric-car-batteries/.
    \1197\ See, e.g., Gohlke, David, and Zhou, Yan. Assessment of 
Light-Duty Plug-In Electric Vehicles in the United States, 2010-
2018. United States: N. p., 2019. Web. doi:10.2172/1506474 (citing 
Berman, Kimberly, Jared Dziuba, Colin Hamilton, Richard Carlson, 
Joel Jackson, and Peter Sklar, 2018. ``The Lithium Ion Battery and 
the EV Market: The Science Behind What You Can't See.'' BMO Capital 
Markets, February 2018. https://bmo.bluematrix.com/docs/pdf/079c275e-3540-4826-b143-84741aa3ebf9.pdf); MIT Energy Initiative. 
2019. Insights into Future Mobility, at 77. Cambridge, MA: MIT 
Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
    \1198\ Schipper, Florian, Evan M. Erickson, Christoph Erk, Ji-
Yong Shin, Frederick Francois Chesneau, and Doron Aurbach. 2017. 
``Review--Recent Advances and Remaining Challenges for Lithium Ion 
Battery Cathodes I. Nickel-Rich, LiNixCoyMnzO2.'' Journal of the 
Electrochemical Society 164, no. 1 (1): A6220-A6228. https://doi.org/10.1149/2.0351701jes.
---------------------------------------------------------------------------

    For the final rule analysis, based on these considerations, the 
agencies in consult with Argonne determined that it was reasonable to 
model HEV, PHEV, and BEV batteries using NMC622 as the cathode active 
material, as shown in Table VI-91 below.
[GRAPHIC] [TIFF OMITTED] TR30AP20.220

    The agencies recognize that there will be advancements in battery 
chemistries during the rulemaking timeframe. As discussed further in 
Section (3), below, the analysis accounts for the potential that 
battery costs will decrease, but in a technology-agnostic manner. The 
agencies used BatPaC to model battery costs for the analysis by 
modeling battery prices in a specific year--in this case, MY 2020--and 
then used learning curves to reduce the cost of batteries over time. 
The learning curves act as a proxy for potential future improvements in 
battery chemistry and other battery-related advancements that would 
reduce costs. Using the learning curves in this way makes it 
unnecessary to make inherently uncertain projections of potential 
future improvements in battery chemistry over time.
    BatPaC version 4.0, which contains NMC811 as a chemistry option, 
was released after the analysis for this rule was completed. However, 
the cost estimates generated in BatPaC version 3.1 using NMC622, with 
discussed learning curves applied resulted in estimated $/kWh battery 
pack costs, during the rule making time frame within a reasonable range 
of other estimated projections that considered NMC811 as the 
predominant battery chemistry. As discussed further in Section (3), a 
significant shift in battery chemistry alone is only one factor 
required to significantly lower battery costs; other developments like 
increases in battery pack production quantities and cell yield (plant 
efficiencies) would be required to reach the commonly-cited $100/kWh 
target.
    The agencies recognize that the specific chemistries manufacturers 
may choose for future model years may or may not be the same as the 
chemistries selected by the agencies for the analysis. However, this 
approach mirrors the approach taken to modeling technology 
effectiveness and cost used across the analysis; the modeled technology 
effectiveness and cost represents a level of performance representative 
of the typical range of performance across industry. If the agencies 
modeled pre-production battery chemistries unlikely to be widely 
adopted by the industry for several years, the analysis would likely 
under-predict the actual cost and effectiveness of electrification 
technology application. Accordingly, the agencies determined that using 
LFP-G as the modeled chemistry of choice for mild hybrids and NMC622 as 
the modeled chemistry of choice for strong HEVs, PHEVs, and BEVs was 
reasonable.
    The agencies also refined other inputs and assumptions used for 
modeling battery costs in BatPaC, based on a review of public comments 
and subsequent review of market research, technical publications, and 
other information.
    Argonne continuously studies the battery pack designs of existing 
electrified vehicles in the market, using, among other information, 
detailed battery pack teardown analysis reports spanning a range of 
electrified vehicle types and vehicle classes produced over a range of 
MYs. For the final rule, Argonne utilized detailed battery pack 
teardown analysis reports for 10 MY

[[Page 24500]]

2013 to MY 2018 vehicles from A2mac1,\1199\ as shown in the Table VI-92 
below.
---------------------------------------------------------------------------

    \1199\ Argonne Vehicle Modeling for Safer Affordable Fuel 
Efficient (SAFE) Vehicles Final Rulemaking, Section 5.9 Battery 
Performance and Cost Model (BatPaC), referencing A2Mac1 Automotive 
Benchmarking, https://a2mac1.com.
[GRAPHIC] [TIFF OMITTED] TR30AP20.221

    The teardown analysis reports were used to evaluate different 
battery pack design criteria, including battery pack power, battery 
pack energy, battery pack configuration, total number of cells per 
module, number of modules per pack, battery pack mass, energy density 
(cell/pack), cell voltage, battery pack voltage, cathode chemistry, 
cell capacity, and pack capacity. The metrics data collected from 
teardown analysis were used to estimate the battery pack manufacturing 
cost and mass (energy density-Wh/kg) in BatPaC for these exemplar 
vehicles from the A2Mac1 database. The data collected was also used to 
validate the battery pack design assumptions in BatPaC for the final 
rule. The four metrics that BatPaC provides are: Battery pack 
manufacturing cost, battery pack weight (energy density-Wh/kg), battery 
pack capacity (Ah) and nominal battery pack voltage. Since the A2mac1 
teardown reports do not avail the manufacturing costs of these battery 
packs, the analyses and comparisons were limited to the scope of the 
other three criteria.
    For the NPRM, Argonne used the U.S. Department of Energy VTO 
targets for battery energy density (Wh/kg) for high energy and power 
density-(W/kg) for high powered batteries.\1200\ As a result of the 
analysis discussed above Argonne updated the method of estimating 
battery pack weight for each battery pack design in the final rule 
analysis. The analysis revealed greater influences on battery pack 
design by usable energy density characteristics then was initially 
assumed for the NPRM. For the final rule analysis BatPaC was used for 
battery pack weight estimates along with manufacturing cost estimates.
---------------------------------------------------------------------------

    \1200\ Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, ANL/CSE-19/2.
---------------------------------------------------------------------------

    As discussed further in Section VI.C.3.e)(1)(c) Battery Pack Costs, 
the number of cells per pack influenced total battery pack costs for 
the final rule. As result of the analysis discussed above Argonne 
updated the number of cells in each battery. For the final rule 
analysis battery cell counts increased or decreased for some battery 
pack designs, while battery counts for some designs remained the same. 
Argonne's process for evaluating different design criteria for 
electrified vehicles is detailed further in the Argonne model 
documentation.\1201\
---------------------------------------------------------------------------

    \1201\ A Detailed Vehicle Simulation Process To Support CAFE and 
CO2 Standards for the MY 2021-2026 Final Rule Analysis, Section 5.9 
Battery Performance and Cost Model (BatPaC).
---------------------------------------------------------------------------

    The agencies also updated other BatPaC inputs and assumptions based 
on additional market information or research. For the NPRM, the 
agencies modeled battery packs in BatPaC using the default values 
associated with the baseline manufacturing plant, including an annual 
production rate of 100,000 batteries.\1202\
---------------------------------------------------------------------------

    \1202\ See Nelson, Paul A., Gallagher, Kevin G., Bloom, Ira D., 
and Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles--SECOND EDITION (2012), at 62. 
Available at https://publications.anl.gov/anlpubs/2015/05/75574.pdf.
---------------------------------------------------------------------------

    The estimate for battery pack costs incorporates an assumption of 
the battery pack production volume. Both BatPaC version 3.0, used in 
the NPRM, and BatPaC version 3.1, used in the final rule, include a 
default value assumption of 100,000 battery pack units manufactured per 
year per manufacturing plant as well as the plant efficiency (cell 
yield) of 95 percent. For the final rule, the agencies adjusted the 
production volume assumption used in BatPaC version 3.1 to 25,000 
battery pack units, based on the analysis presented below.
    As described in the BatPaC model documentation, the BatPaC models 
the differences in pack designs and how they affect the costs of one or 
more steps in the battery production process and the physical plant 
layout.\1203\ For example, increasing the power of the battery packs 
without increasing the number of cells, or cell capacity, results in 
the model increasing the area of the cells and decreasing the electrode 
coating thickness. This results in an increased cost of the coating 
equipment, the floor area occupied by the equipment, and the direct 
labor for the process.1204 1205 The agencies are aware that 
each manufacturer (not brand) has a unique battery pack design that 
differs from other manufacturers. Accordingly, it is likely that each 
manufacturer's BEV models had distinct characteristics, such as unique 
battery packaging space, energy requirements, thermal control systems, 
and safety systems, which cause battery pack designs to vary between 
each manufacturer.
---------------------------------------------------------------------------

    \1203\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and 
Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, Third Edition (2019), at 100. 
Available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
    \1204\ Kupper et al, The Future of Battery Production for 
Electric Vehicles, Boston Consulting Group, (Sept. 11, 2018), 
https://www.bcg.com/publications/2018/future-battery-production-electric-vehicles.aspx.
    \1205\ Id.
---------------------------------------------------------------------------

    Thus, the agencies determined that even though one battery 
manufacturer

[[Page 24501]]

might manufacture batteries for multiple vehicle manufacturers, the 
default BatPaC assumption of 100,000 battery pack units manufactured 
per plant likely did not account for all of the cost differences in 
pack designs between manufacturers. Therefore, the agencies assumed the 
production volume of each battery pack type was reasonably represented 
by the BEV production volume for each manufacturer. The agencies also 
assumed that battery pack manufacturing plants operated at reasonable 
capacity during that timeframe, which would produce the lowest cost 
assumption.
    The agencies analyzed BEV sales for MYs 2016-2019, referencing data 
collected by the Department of Energy.\1206\ Table VI-93 shows that 
individual manufacturer U.S. BEV sales are substantially below 100,000 
units per year except for Tesla, beginning in MY 2018 Tesla is a 
vertically integrated battery and BEV manufacturer, which is not the 
model the remainder of the industry has implemented, or intends to, 
based on the agencies current understanding. More specifically, Tesla 
sold more BEVs than all manufacturers combined in MYs 2016, 2018, and 
2019. 2017 was the only year in which all other manufacturers combined 
sold more BEVs than Tesla. Ultimately, in selecting a battery pack 
volume estimates for an industry-wide assessment, the agencies sought 
to accurately account for both the representative production volumes 
and representative practices applicable to the industry. As such, the 
agencies evaluated the average per manufacturer volumes, less the 
outlying and vertically integrated volumes of Tesla (shown in Table VI-
94). As depicted in Table VI-93 and Table VI-94, the data show that the 
average annual sales of BEVs for individual manufacturers, excluding 
Tesla, is just 5% of the default battery pack production volume in 
BatPaC.
---------------------------------------------------------------------------

    \1206\ Light Duty Electric Drive Vehicles Monthly Sales Updates, 
Argonne National Laboratory Energy Systems Division, https://www.anl.gov/es/light-duty-electric-drive-vehicles-monthly-sales-updates (last visited March 2, 2020); Maps and Data, Alternative 
Fuels Data Center, https://afdc.energy.gov/data/ (last visited March 
2, 2020).
[GRAPHIC] [TIFF OMITTED] TR30AP20.222


[[Page 24502]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.223

    In consideration of this data, when estimating the production 
volume in the final rule analysis, the agencies selected a value of 
25,000 units per year per manufacturer as a reasonable estimate for the 
average industry for MY 2020, which is the base model year for 
estimated battery pack costs using BatPaC version 3.1. As discussed in 
Section VI.C.3.e)(3) Electrification Learning Curves, other model year 
battery pack costs are estimated using cost learning. Using the default 
production volume of 100,000 units per year per manufacturer, the 
agencies would have underestimated the actual cost of battery pack 
production for MY 2020, as the model assumes that production costs 
decrease as production volumes increase. By selecting the value of 
25,000 units per year per manufacturing plant, the battery cost 
estimate from the BatPaC model better aligned with the cost estimate 
published in industry-recognized reports such as the UBS MY 2016 Chevy 
teardown report.1208 1209 1210
---------------------------------------------------------------------------

    \1207\ Note, for the assessment, Nissan and Mitsubishi are 
considered a single manufacturer.
    \1208\ Proposed Determination TSD at 2-127.
    \1209\ Based on the battery cell to battery pack ratio of 1.3 to 
1.5, the 2015-2019 cell-level figure of $145 per kWh used in the MY 
2016 Chevy Bolt would translate to approximately $190 to $220 per 
kWh on a pack level.
    \1210\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
---------------------------------------------------------------------------

    The agencies performed a sensitivity study for production volume 
using BatPaC version 3.1. The cost of the battery pack dropped by 15 
percent on average when the production volume was changed from 25,000 
to 100,000 units per year. The sensitivity analysis showed that 
manufacturing plant volume has a significant impact on battery pack 
costs and therefore it is important to use realistic production volume 
estimates for the battery pack cost analysis.
    Manufacturing plant efficiency is another parameter important to 
estimate battery pack costs. BatPaC version 3.1 defines manufacturing 
plant efficiency in terms of cell yield, or the number of cells that 
are usable out of the total number of cells that the plant 
produced.\1211\ Since battery pack technology and battery pack 
manufacturing processes are proprietary, the data on plant efficiencies 
are not widely reported. While BatPaC uses a default cell yield (plant 
efficiency) value of 95 percent, Argonne battery experts have used an 
85 percent cell yield value to represent the current production yield 
for internal DOE studies.\1212\ By selecting an 85 percent cell yield 
value for the final rule analysis, the agencies aligned the cell yield 
value assumption with internal DOE studies.
---------------------------------------------------------------------------

    \1211\ Cells might not be usable because of, for example, 
manufacturing defects, among other reasons.
    \1212\ Argonne National Laboratory, BatPaC Model Software, 
https://www.anl.gov/cse/batpac-model-software (last visited March 
19, 2020). Argonne used an 85% cell yield assumption in its 
Estimated Cost of EV Batteries 2018-19 analysis.
---------------------------------------------------------------------------

    In addition, as discussed in detail above, the final rule analysis 
was performed using BatPaC version 3.1, with NMC622 assumed as the 
battery chemistry for HEVs, PHEVs, and BEVs. Separate from the inputs 
and assumptions discussed here, the Argonne battery experts made a 
number of changes to BatPaC version 3.1, and these are extensively 
documented in the BatPaC manual,\1213\ as well as in Argonne model 
documentation for final rule.
---------------------------------------------------------------------------

    \1213\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and 
Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, Third Edition (2019), 
available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
---------------------------------------------------------------------------

(b) Comments on Information Availability
    In addition to comments that the agencies' battery pack costs were 
too high, the agencies received comments that the analysis for battery 
pack costs was unclear and not well documented. ICCT stated that the 
agencies largely obscured the BEV cost sources and calculations, which 
made it ``nearly impossible for even very interested researchers to 
understand how all the BatPaC costs translate into BEV costs that can 
be compared with other full-BEV costs in the literature.'' \1214\ ICCT 
stated that to enable meaningful public comments, the sources and cost 
calculations must be made explicit and the agencies must provide an 
additional public comment opportunity.\1215\
---------------------------------------------------------------------------

    \1214\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1215\ Id.
---------------------------------------------------------------------------

    CARB claimed that it could not comment meaningfully on the battery 
modeling for the NPRM analysis without extensive additional 
information.\1216\ As such, CARB submitted a letter to the agencies' 
NPRM docket posing, under FOIA, a number of questions pertaining to 
battery assumptions used for the modeling. This requested information 
concerned what version of BatPaC was used in the NPRM analysis, inputs 
incorporated into the BatPaC model; and information about how battery 
costs were generated for the analysis.
---------------------------------------------------------------------------

    \1216\ California Air Resources Board, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    Specifically, CARB's initial comments alleged that the agencies had 
not disclosed the exact version of BatPaC used, and had simply claimed 
to use the ``most up-to-date'' version of BatPaC,

[[Page 24503]]

and further that the agencies had not disclosed ``the BatPaC modeling 
files that were used, clear statements about what version of the model 
was used, or thorough descriptions of the inputs to those modeling 
runs.'' CARB claimed that without that information, ``there is no way 
to know what assumptions were made for raw material pricing, battery 
cell yields, pack electrical connection topology, battery production 
volume assumptions, or if any additional parameters were modeled, like 
rapid charging capability.'' CARB argued that these pieces were 
critical to understanding whether the BatPaC model was estimating 
proper battery pack cost values.
    In a subsequent docketed comment submitted as an administrative 
appeal to NHTSA's FOIA response, CARB reasserted that, in fact, the 
``most recent version'' of BatPaC had not been used, because the FOIA 
response stated clearly that version 3.0 had been used and Argonne had 
updated to version 3.1 in October 2017, which was the last version 
released before the NPRM was published. CARB further argued that NHTSA 
was ``choosing to withhold information about battery pack 
configurations,'' and that the agencies had not posted the BatPaC model 
version and files used for the NPRM to the agencies' dockets, 
inhibiting meaningful comment.
    The majority of information sought by CARB's comment was already 
published in supporting documents and materials posted to the agencies' 
dockets and online websites for the NPRM. Nevertheless, in an effort to 
answer CARB's specific questions, NHTSA also processed the initial 
comment as a FOIA request and provided a written response directly to 
CARB within the comment period. This response both pointed CARB to the 
locations where the sought material could be located among the 
published NPRM materials, and expressly answered several of CARB's 
questions for clarification, such as identifying the specific version 
of BatPaC utilized in the NPRM analysis. For example, although the 
Argonne model documentation describing the battery modeling for the 
NPRM was included in the docket, the agencies' response directed CARB 
to the precise location in the docket where it could be found.
    The agencies believe that the NPRM docket contained enough 
information for stakeholders to comment meaningfully. This is apparent 
from the voluminous comments the agencies received regarding the NPRM's 
electrification analysis--including from CARB. For example, as 
discussed above, CARB submitted extensive comments on each element of 
the battery cost modeling that CARB claimed the agencies did not 
adequately explain. As discussed above, CARB stated that the agencies' 
selected battery chemistries represented a step backward from previous 
analysis done for the Draft TAR. CARB noted that regardless of whether 
NMC441 or NMC333 was chosen for PHEVs and BEVs in the NPRM analysis, 
the biggest lithium-ion production companies have indicated that they 
will use NMC811 for BEVs, and therefore neither NMC441 nor NMC333 would 
represent current technology going into BEVs or near-future BEV battery 
technology. CARB stated that NMC811 technology is expected to come to 
market in 2019, which, the agencies note, is far sooner than 
anticipated, even in the agencies' prior analyses. CARB was accordingly 
able to communicate its opinion that NMC881 should have been used to 
model battery chemistries for the NPRM analysis, and that NMC441 or 
NMC333 should not be used.
    As these comments demonstrate, in addition to the extensive 
comments listed above, the expansive information, data, and 
documentation concerning the Argonne BatPaC modeling analysis for the 
NPRM sufficiently enabled commenters to submit voluminous technical 
analysis regarding the electrification analysis. Moreover, while the 
docketed and published NPRM materials themselves afforded sufficient 
notice on these topics, the agencies even undertook the additional step 
of directly responding to CARB in writing in an attempt to address 
specific questions raised by CARB. This written correspondence both 
directed CARB to specific locations on the rulemaking dockets and 
agencies' websites where information CARB was seeking could be 
accessed, and even directly answered several of CARB's questions 
through narrative responses. Both CARB and other commenters submitted 
subsequent comments, which referenced the material described in this 
written response. Accordingly, the agencies consider the information 
provided with the NPRM sufficient to enable meaningful comment, which 
is underscored by the voluminous technical comments received on the 
electrification issues.
    For this final rule, the BatPaC model version 3.1 (June 2018) model 
documentation has been included in the docket for this 
rulemaking.\1217\ Furthermore, Argonne's detailed documentation 
describing the modeling process used to support this final rule 
provides information and specific assumptions that Argonne's experts 
used to simulate batteries and their associated costs for the full 
vehicle simulation modeling.\1218\ These resources, in addition to the 
detailed description of the battery cost modeling process provided here 
and in the FRIA provide interested stakeholders the necessary tools to 
understand the battery cost modeling analysis.
---------------------------------------------------------------------------

    \1217\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and 
Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2), 
available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
    \1218\ A Detailed Vehicle Simulation Process To Support CAFE and 
CO2 Standards for the MY 2021-2026 Final Rule Analysis.
---------------------------------------------------------------------------

c) Final Rule Battery Pack Costs
    As discussed above, based on comments and additional research, the 
agencies updated the battery cost analysis for the final rule by 
relying on BatPaC version 3.1.\1219\ In addition, as outlined above and 
explained in more detail in the Argonne Model Documentation for this 
final rule, several inputs and assumptions were updated based on public 
comments, market research, and additional literature review. The 
agencies computed the average battery pack cost across all road load 
combinations for electrification technologies that could be reasonably 
compared between the NPRM and final rule.\1220\
---------------------------------------------------------------------------

    \1219\ Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2) 
provides a complete list of changes and assumptions incorporated in 
BatPaC version 3.1.
    \1220\ Costs data is from the CAFE Model core file 
Battery_Costs.csv.
---------------------------------------------------------------------------

    Table VI-95 to Table VI-99 show the differences between battery 
pack costs presented in the NPRM and final rule.\1221\ The tables show 
absolute cost differences between battery packs, which can vary for 
battery packs with different energy and power combinations. For 
example, as shown in Table VI-96, the cost difference between the NPRM 
and final rule for a Mild HEV battery pack with a 1kWh energy and 10kW 
power rating is -28 percent. Similarly, the cost difference in an HEV 
battery pack with a 1kWh battery energy and 40kW power rating is 5 
percent. In summary, the percentage increase or decrease in the table 
represents the

[[Page 24504]]

absolute cost differences between the battery packs used in NPRM and in 
final rule.
---------------------------------------------------------------------------

    \1221\ The absolute cost differences shown here is by comparing 
the cost of battery pack with similar number of cells in the NPRM to 
the final rule cost lookup tables for compact and medium car. The 
cost differences between the NPRM and the final rule cost lookup 
tables for small SUV, medium SUV and Pickup trucks will be different 
from the table shown here.
---------------------------------------------------------------------------

    Figure VI-40 to Figure VI-42 shows the average battery pack costs 
across all road load combinations for each applicable vehicle 
technology class for SHEVPS, PHEV50, and BEV200s between the NPRM and 
final rule.\1222\ Since the battery pack size varies for different road 
load combinations, the battery pack cost across different road load 
combinations varies as well. For example, there are 105 combinations of 
different mass reduction, aerodynamic improvements and rolling 
resistance improvements. The battery pack size for an initial road load 
condition that includes MR0, AERO0 and ROLL0 is larger, and therefore, 
the cost of the battery pack is higher as well. The battery pack size 
is smaller for the highest level of road load reduction such as in MR6, 
AERO20 and ROLL20, and the cost of battery pack is less as well.
---------------------------------------------------------------------------

    \1222\ The agencies did not simulate SHEVPS and BEV200 
powertrain architectures on pickup trucks in the NPRM, so those are 
not included in the comparison.
---------------------------------------------------------------------------

    Table VI-95 shows the cost difference in Micro HEV battery packs. 
The cost reduction is from the reduced number of cells in the battery 
pack.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.224

    Table VI-96 shows percentage cost differences for mild hybrid 
(BISG) battery packs. The cost difference is due, in part, to 
accounting for BISG-related hardware costs, such as the battery 
management system, as part of the electric machine costs in this final 
rule.\1223\
---------------------------------------------------------------------------

    \1223\ In the NPRM, additional hardware component costs were 
included as part of the battery pack cost.
[GRAPHIC] [TIFF OMITTED] TR30AP20.225

    Table VI-97 shows the percentage cost differences for HEV battery 
packs. Even as the battery chemistry changed to NMC622, the cost 
increase is from the different battery pack production volume and plant 
efficiency assumptions used in the final rule.

[[Page 24505]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.226

    Figure VI-40 shows the difference in battery pack costs for SHEVPS 
applications between the NPRM and final rule. Power-split hybrids could 
not be used in pickup trucks due to their unique power and towing 
requirements, so those technology classes are not shown. In general, 
the cost of the battery pack in the final rule analysis increased due 
to the updated battery pack production volume and plant efficiency 
assumptions.
[GRAPHIC] [TIFF OMITTED] TR30AP20.227

    Table VI-98 shows the percentage cost differences between the NPRM 
and final rule for PHEV50 battery packs. The cost increase in the 
PHEV50 battery pack shown here is mainly due to the increase in number 
of cells per pack as well as the other updated BatPaC assumptions.

[[Page 24506]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.228

    Table VI-94 shows the difference in average PHEV50 battery pack 
costs between the NPRM and final rule for all technology combinations.
[GRAPHIC] [TIFF OMITTED] TR30AP20.229

    Table VI-99 shows the percentage cost differences for BEV battery 
packs. In the example shown in Table VI-99, the agencies compared the 
cost lookup table from the NPRM with 300 cells to the cost lookup table 
in the final rule analysis with 320 cells. The cost increase in the 
higher energy packs is due to the different battery pack production 
volume and plant efficiency value assumptions, along with the different 
battery chemistry assumption.

[[Page 24507]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.230

BILLING CODE 4910-59-C
    Figure VI-42 shows the average cost of BEV200 battery packs across 
all technology combinations for technology classes that could be 
compared between the NPRM and final rule. As shown, for the final rule 
analysis, the average cost of a BEV200 battery pack is lower than the 
average cost of the NPRM BEV200 battery pack. For the final rule 
analysis, the agencies updated the motor efficiency map for BEVs (as 
explained in Section VI.C.3.d) Electrification Technology 
Effectiveness) and updated the glider share of the vehicles from 50 
percent of the curb weight to 71 percent of the vehicle curb weight (as 
explained in Section VI.C.4 Mass Reduction). In addition, the updated 
motor weight resulted in further reduced vehicle weights. This 
combination of improved vehicle assumptions resulted in reduced energy 
and power requirements in BEVs.
    The agencies also observed that even as the number of cells in the 
battery pack increased from 300 to 320, and changes in production 
volume and plant efficiency values resulted in marginal cost increases 
for higher energy packs, the overall battery capacity requirement went 
down due to overall reduction in power and energy demand from electric 
vehicles.\1224\ A reduction in battery capacity leads to reduced cell 
size in a pack with number of cells and voltage. A reduction in cell 
size leads to cost reductions at the cell level and at the pack level. 
In general, a higher capacity battery pack is more expensive than a 
lower capacity battery pack due to the increase in cell size for a 
given number of cells and voltage.1225 1226
---------------------------------------------------------------------------

    \1224\ As explained above, the energy density values in the NPRM 
were kept constant. For the final rule analysis, the power density 
varied to meet different power and energy requirements, as was 
observed through market research.
    \1225\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and 
Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion 
Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2), 
at 15 (battery design worksheet). Available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
    \1226\ The amount of electrode materials and electrode area of 
the cells are determining cost factors in the battery. Higher 
capacity battery packs require additional manufacturing steps to 
increase the energy density of the pack.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24508]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.231

    The graphs demonstrate the range of cost changes observed, with the 
other electrification technologies falling somewhere in between the 
extremes. In summary, the agencies observed that the BEV200 technology 
showed a cost reduction in battery packs across all vehicle platforms 
with the largest reductions occurring for the largest battery packs. In 
contrast the PHEV50 technology showed a cost increase in battery packs 
across all vehicle platforms with the smallest increase for the largest 
battery packs and the largest increase for the smallest battery packs. 
It is worth noting the cost decreases seen across the technologies are 
generally larger than the cost increases.
    For the final rule, when possible, the calculated battery pack 
weight and manufacturing cost was also compared with the battery pack 
cost and weight data obtained through various benchmarking studies. For 
example, UBS reported a battery pack manufacturing cost of $12,500 from 
its 2017 Chevrolet Bolt teardown analysis.\1227\ Using a production 
volume of 25,000 packs per year per plant and similar battery pack 
design, BatPaC estimated a manufacturing cost of $10,680.\1228\ These 
comparisons were used to verify the different assumptions used in 
BatPaC and helps represent the battery packs for electrified vehicles 
used in representative market volume. Table VI-100 shows a comparison 
of specifications estimates for 60 kWh and 160 kW battery packs from 
the 2016 DOE VTO report 1229 1230 and BatPaC version 3.1 
(June 2018), and the Chevrolet Bolt. The comparison shows modeled and 
actual battery packs are in close agreement.
---------------------------------------------------------------------------

    \1227\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
    \1228\ $178/kWh x 60kWh = $10,680.
    \1229\ Peter Faguy, Overview of the DOE Advanced Battery R&D 
Program (June 2015), https://www.energy.gov/sites/prod/files/2015/06/f23/es000_faguy_2015_o.pdf.
    \1230\ Freyermuth, Vincent. Rousseau, Aymeric. ``Impact of 
Vehicle Technologies Office Targets on Battery Requirements.'' ANL/
ESD-16/22. Energy Systems Division, Argonne National Laboratory 
(2016).

---------------------------------------------------------------------------

[[Page 24509]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.232

    In addition, the agencies compared the battery pack cost estimates 
generated using BatPaC to other current studies or studies cited by 
commenters. Table VI-101 summarizes battery pack estimates from 
selected studies in MYs for which that information was available.
---------------------------------------------------------------------------

    \1231\ Not each study distinguished a DMC source year, so these 
values vary slightly based on inflation.
    \1232\ Sources generally provided estimates for 2018 or 2020.
    \1233\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
    \1234\ Mosquet et al., The Electric Car Tipping Point, BCG (Jan. 
11, 2018), https://www.bcg.com/publications/2018/electric-car-tipping-point.aspx. This study provided cell cost estimates that the 
agencies converted to pack cost estimates using a multiplier of 1.3, 
as outlined in the Draft TAR at 5-124.
    \1235\ Nic Lutsey and Michael Nicholas, Update on electric 
vehicle costs in the United States through 2030, ICCT (April 2, 
2019), available at https://theicct.org/publications/update-US-2030-electric-vehicle-cost. The presented values are $/kWh pack costs for 
mid-range electric cars/crossovers and SUVs.
    \1236\ McKerracher et al., Electric Vehicle Outlook 2019--Free 
Interactive Report, Bloomberg New Energy Finance (May 2019), https://about.bnef.com/electric-vehicle-outlook/.
    \1237\ Logan Goldie-Scot, A Behind the Scenes Take on Lithium-
ion Battery Prices, Bloomberg New Energy Finance (March 5, 2019), 
https://about.bnef.com/blog/behind-scenes-take-lithium-ion-battery-prices/. BNEF projected the pack costs in 2018$ for 2018 as $176, 
and used the same value in the Electric Vehicle Outlook 2019 to 
describe pack cost levels ``today.''
    \1238\ MIT Energy Initiative. 2019. Insights into Future 
Mobility. Cambridge, MA: MIT Energy Initiative. Available at http://energy.mit.edu/insightsintofuturemobility.
    \1239\ Islam, E., Kim, N., Moawad, A., Rousseau, A., ``A Large-
Scale Vehicle Simulation Study To Quantify Benefits & Analysis of 
U.S. Department of Energy VTO & FCTO R&D Goals.'' Report to U.S. 
Department of Energy. Contract ANL/ESD-19/10 (forthcoming).

---------------------------------------------------------------------------

[[Page 24510]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.233

    As shown in the table above, there are a range of cost estimates 
for battery packs. Each individual cost estimate is derived based on 
certain set of assumptions to arrive at a rate of cost reduction. Among 
all the different cost estimates, Bloomberg New Energy Finance (BNEF) 
has the most aggressive year-over-year cost reductions, based on the 
historical learning rate of 18% and their battery demand 
forecast.\1240\ Similar to other sources of cost estimates BNEF assumes 
improved battery chemistry and battery density increasing greater than 
200Wh/kg by 2030. In order for the battery manufacturer to achieve 
economies of scale, BNEF assumes a global battery manufacturing 
facility capable of producing battery packs for both stationary energy 
storage and vehicle applications.
---------------------------------------------------------------------------

    \1240\ Logan Goldie-Scot, A Behind the Scenes Take on Lithium-
ion Battery Prices, Bloomberg New Energy Finance (March 5, 2019), 
https://about.bnef.com/blog/behind-scenes-take-lithium-ion-battery-prices/.
---------------------------------------------------------------------------

    A recent report from the Massachusetts Institute of Technology 
(MIT), the MIT Energy Initiative's Insights into Future Mobility, has 
the most conservative estimate among all the cost sources listed the 
Table VI-101. The authors use a more rigorous two-stage method of 
estimating composite battery learning curves independently for (a) 
battery material synthesis and minerals costs, and (b) battery pack 
production processes. The learning rates are defined as the cost 
reduction that results from cumulative volume doubling, and produce 
separate cost learning rates for the two stages of 3.5 percent and 16.5 
percent, respectively. The study argues that there are greater 
opportunities for cost learning in the production stage than the 
chemical synthesis stage, which is more mature. These cost estimates 
produce global EV fleet penetration rates that may not be as aggressive 
as other estimates, reaching only 33 percent by 2050. This study also 
assumes NMC811 will be available by 2030.
    The cost estimates from other sources referenced above also include 
assumptions about higher levels of battery pack production and higher 
density battery cells. Most cost estimates assume improved battery 
chemistry, such as NMC811. As discussed above, the agencies determined 
that modeling assuming NMC622 was reasonable, based on current 
production vehicles, the relative uncertainty surrounding large-scale 
NMC811 deployment in the rulemaking timeframe, and the ability to 
account for lower battery pack costs over time with cost learning. The 
agencies also believe that, based on the market analysis and from the 
teardown analysis, improvements in battery chemistry may be slow to be 
applied in a widespread manner, and therefore the economies of scale 
required to achieve considerable cost reductions solely from 
improvements in chemistry may remain effusive during the rulemaking 
timeframe.
    For these reasons, the agencies believe that the BatPaC-generated 
battery cost estimates using the updated inputs and assumptions are 
reasonable.
2) Non-Battery Electrification Component Costs
    Battery components are the biggest driver of the cost of 
electrification, however, non-battery electrification components also 
add to the total cost required to electrify a vehicle. In this 
analysis, the agencies accounted for the following non-battery 
component costs: Electric motor(s), inverter, and other power 
electronics including a bi-directional DC/DC converter, a voltage step 
down DC/DC converter, and an on-board charger. Collectively, these 
components (except for the on-board charger) are referred to as the 
electric traction drive systems (ETDS), or the electric machine. Non-
plug-in hybrid electric vehicles include all of the listed components 
except for an on-board charger; PHEVs include all of the listed 
components; and BEVs include all of the listed components except, in 
some cases, a second motor.
    For the NPRM, the agencies accounted for battery pack costs and 
ETDS costs independently.\1241\ The Alliance commented broadly in 
support of separating electrification hardware costs and battery costs, 
and stated that it was a positive change to the modeling.\1242\ The 
Alliance correctly noted that the separation allowed for separate 
learning rates and cost differentiation between the two distinct pieces 
of electrification technologies.
---------------------------------------------------------------------------

    \1241\ PRIA at 362.
    \1242\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
12073, at 140.

---------------------------------------------------------------------------

[[Page 24511]]

    As stated in the PRIA,\1243\ the agencies derived the cost values 
for the EDTS using Argonne National Laboratory's ``Assessment of 
Vehicle Sizing, Energy Consumption, and Cost through Large-Scale 
Simulation of Advanced Vehicle Technologies'' report.\1244\ Generally, 
the agencies referred to this report in the PRIA as the DOE VTO report, 
as it was a report that reviewed results of the DOE VTO. Some 
commenters seemed confused by this alternative reference--even 
questioning why the agencies didn't rely on recent Argonne National 
Laboratory reports.\1245\ To clarify, this report was written by 
Argonne National Laboratory, and to avoid further confusion it is 
referred to using the full title throughout this rule.
---------------------------------------------------------------------------

    \1243\ 83 FR 43047; PRIA at 362.
    \1244\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and 
Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption 
and Cost Through Large Scale Simulation of Advanced Vehicle 
Technologies (ANL/ESD-15/28). United States (2016), available at 
https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
    \1245\ California Air Resources Board, NHTSA-2018-0067-11973, at 
130-31.
---------------------------------------------------------------------------

    CARB expressed concerns with non-battery component effectiveness 
values, arguing that the agencies inappropriately relied on outdated 
data for electric machines and inverter efficiencies across all 
electrification applications, and further claiming that the agencies 
did not project any efficiency gains in those components over 
time.\1246\ Broadly, as these comments on effectiveness related to the 
NPRM non-battery component cost estimates, CARB claimed that the 
agencies failed to consider new data, including the 2015 ORNL Annual 
Progress Report for the Power Electronics and Electric Motors Program, 
and two Argonne studies, which rendered the analysis unrepresentative 
of actual technology costs.
---------------------------------------------------------------------------

    \1246\ California Air Resources Board, NHTSA-2018-0067-11973, at 
130.
---------------------------------------------------------------------------

    CARB also commented that the agencies did not provide any 
substantive discussion or documentation of how non-battery component 
costs were developed for the NPRM analysis. CARB claimed that 
dissonance existed between the PRIA description of voltage systems and 
associated costs needed for different performance classes, the 
Autonomie files, and the technologies input file, and that this served 
as an example of how the agencies failed to include information 
regarding how costs and cost differences were derived, or any component 
changes from previous analyses.
    CARB also commented that the lack of disclosure of non-battery cost 
development information was an issue for other electrification 
technologies. CARB cited the increase in parallel (P2) and power-split 
(PS) hybrid systems costs relative to costs used in past agency 
analyses, noting that there was no discussion on what changed from the 
past analyses. CARB referenced a 2010 FEV teardown (Light Duty 
Technology Cost Analysis, Power-Split and P2 HEV Case Studies, EPA-420-
R-11-015) study that the agencies had previously relied on for 
component costs, noting that not only did the agencies ignore that 
study in the NPRM, but that ICCT had commented 2010 FEV report 
overstated strong hybrid costs at the time of the study, making it 
likely that costs are likely to be lower now and even more so in the 
future. CARB claimed that the agencies provided no justification or 
rationale for the increases in strong hybrid modeled costs for the 
proposal, and that there was no meaningful way to comment on the exact 
components or cost changes that the agencies relied upon. Similarly, 
CARB cited EPA's 2016 Proposed Determination and associated public 
comments from Ford and Tesla on the Draft TAR for the proposition that 
non-battery costs, which were lower in the Draft TAR than the NPRM, 
were conservative and not overly optimistic.
    Finally, in addition to the ORNL and Autonomie group studies that 
CARB referenced as examples of sources that provided updated data on 
non-battery component effectiveness and costs, CARB claimed that newer 
data existed from a UBS Global Research report that examined the 
component costs of a MY 2016 Chevrolet Bolt, and the agencies did not 
discuss why the newer data was not used in the NPRM analysis. CARB 
stated the significant upward adjustment in non-battery costs from 
previous analyses was not supported by industry input, analysis 
conducted by other outside sources, or by the agencies' previous 
analyses.
    As explained above, for the NPRM the agencies relied on Argonne's 
``Assessment of Vehicle Sizing, Energy Consumption, and Cost through 
Large-Scale Simulation of Advanced Vehicle Technologies'' for EDTS 
costs. In turn, the Assessment of Vehicle Sizing, Energy Consumption, 
and Cost through Large-Scale Simulation of Advanced Vehicle 
Technologies report referenced electric machine data provided by OEMs, 
suppliers, and Oak Ridge National Laboratory.\1247\ Regarding CARB's 
assertion that the agencies did not refer to the UBS Global Research 
report on the MY 2016 Chevy Bolt teardown for the NPRM, the agencies 
agree. The UBS Global Research report was not available at the time the 
CAFE model inputs were finalized for the NPRM analysis. That study, 
among others, was considered for the final rule.
---------------------------------------------------------------------------

    \1247\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and 
Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption 
and Cost Through Large Scale Simulation of Advanced Vehicle 
Technologies (ANL/ESD-15/28), at 32.
---------------------------------------------------------------------------

    For the final rule analysis, the agencies carefully considered 
comments and the referenced studies, as well as other studies. The 
agencies determined the cost and component efficiency estimates from 
U.S. DRIVE's October 2017 report, Electrical and Electronics Technical 
Team (EETT) Roadmap,\1248\ provided reasonable estimates to use in the 
final rule. The EETT Roadmap report reflected considerable work by the 
DOE VTO collaboratively with U.S. DRIVE, a government-industry 
partnership. The EETT Roadmap report estimated the 2017 manufacturing 
cost of a commercial on-road 100kW ETDS consisting of a single electric 
traction motor and inverter. The reported costs were approximately 
$1,800, with the cost of the electric motor accounting for $800, and 
approximately $1,000 for the inverter, equaling $18/kW for the ETDS.
---------------------------------------------------------------------------

    \1248\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
---------------------------------------------------------------------------

    The agencies also referenced the UBS MY 2016 Chevy Bolt teardown 
report to compare the cost of the ETDS.\1249\ To compare the costs, the 
agencies applied the $18/kW metric for ETDS as determined by EETT 
Roadmap report to the 150kW ETDS used in the MY 2016 Chevy Bolt ($18kW 
x 150kW = $2700). As shown in Table VI-102, the cost estimate from the 
above computation aligned with UBS MY 2016 Chevy Bolt teardown cost 
estimate. As a result, the agencies determined that it was appropriate 
to use $18/kW to estimate the cost of the ETDS for all hybrid and 
electric vehicle architectures for the final rule.
---------------------------------------------------------------------------

    \1249\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
---------------------------------------------------------------------------

    The EETT Roadmap report did not explicitly estimate the cost of 
other electrical equipment present in PHEVs and BEVs, such as on-board 
chargers, DC to DC converters, and charging cables, but recommended 
cost targets for the years 2020 and 2025. As a consequence, the 
agencies relied on the

[[Page 24512]]

UBS MY 2016 Chevy Bolt teardown report to estimate the cost of on-board 
chargers, DC to DC converters, and charging cables. Table VI-102 shows 
the cost estimate for the ETDS from the EETT Roadmap report and from 
the UBS MY 2016 Chevy Bolt teardown report, and the cost estimate for 
other electrical equipment from the same UBS report.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.234

[GRAPHIC] [TIFF OMITTED] TR30AP20.235

    While the EETT Roadmap report estimated the cost of the ETDS at the 
system level, the report did not itemize the cost of individual 
components in electric motor and inverter in 2017. However, the EETT 
Roadmap report provided target cost estimates for the motor and 
inverter system for the year 2025. As shown in Table VI-104, the EETT 
Roadmap report estimated a cost reduction of 73 percent for the 
inverter and 59 percent for the motor relative to 2017. Using the 
percentage cost reductions from 2025 to the on-road status as defined 
in the EETT Roadmap report, the agencies developed an estimated motor 
and inverter component cost for 2017. The resulting cost estimate for 
2017 using the scaling factor matches the $18/kW for motor and inverter 
($10/kW for Inverter + $8/kW for motor). Since the motor and inverter 
component costs are developed based on a $/kW basis, the agencies 
applied the same $/kW metric for all hybrid and electric vehicle 
applications for the final rule analysis.
---------------------------------------------------------------------------

    \1250\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
    \1251\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
    \1252\ T U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
    \1253\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap, at 18 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.

---------------------------------------------------------------------------

[[Page 24513]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.236

[GRAPHIC] [TIFF OMITTED] TR30AP20.237

[GRAPHIC] [TIFF OMITTED] TR30AP20.238

BILLING CODE 4910-59-C
    In addition, the EETT Roadmap report provided notably newer data 
than the 2010 FEV teardown study referenced by commenters. Based on 
these considerations, the agencies determined that the EETT Roadmap 
report provided reasonable costs to estimate the cost of EDTS 
components in the rulemaking timeframe.
---------------------------------------------------------------------------

    \1254\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap, at 23 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
---------------------------------------------------------------------------

(3) Electrification Learning Curves
    The total incremental costs of electrification powertrain 
technologies are comprised of the DMC as modified by the learning 
curves for each individual powertrain component, which include 
batteries, non-battery components, and IC engines and transmissions 
(for hybrids and PHEVs). The PRIA showed the learning curves for 
battery and non-battery electrification technologies,\1255\ and listed 
the sources used to develop those curves, including the 2015 NAS 
report,

[[Page 24514]]

Wright-based learning curves,\1256\ and Argonne's 2016 Assessment of 
Vehicle Sizing, Energy Consumption, and Cost through Large-Scale 
Simulation of Advanced Vehicle Technologies.\1257\ Learning rates for 
batteries were also derived using Argonne's BatPaC model.
---------------------------------------------------------------------------

    \1255\ PRIA at 380.
    \1256\ Wright, T. P. (1936). Factors Affecting the Cost of 
Airplanes. Journal of Aeronautical Sciences, vol. 3 124-125. http://www.uvm.edu/pdodds/research/papers/others/1936/wright1936a.pdf.
    \1257\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and 
Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption 
and Cost Through Large Scale Simulation of Advanced Vehicle 
Technologies (ANL/ESD-15/28). United States (2016). Available at 
https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
---------------------------------------------------------------------------

    For the NPRM, to develop the learning curves for non-battery 
components, the agencies consulted Argonne's 2016 Assessment of Vehicle 
Sizing, Energy Consumption, and Cost through Large-Scale Simulation of 
Advanced Vehicle Technologies report. The report provided estimated 
cost projections from the 2010 lab year to the 2045 lab year for 
individual vehicle components.1258 1259 The agencies 
considered the component costs used in electrified vehicles, and 
determined the learning curve by evaluating the year over year cost 
change for those components.
---------------------------------------------------------------------------

    \1258\ ANL/ESD-15/28 at 116.
    \1259\ DOE's lab year equates to five years after a model year, 
e.g., DOE's 2010 lab year equates to MY 2015.
---------------------------------------------------------------------------

    The agencies used BatPaC version 3.0 to develop the NPRM learning 
curves for batteries. As discussed above, BatPaC calculations are based 
on generic pack design for a given set of inputs that could reasonably 
represent potential current and future designs. Because BatPaC does not 
simulate battery costs as a function of time, the agencies modified the 
battery volume inputs for MY 2015, MY 2020, MY 2025 to show costs in 
each of those MYs. Like the non-battery component analysis, a learning 
curve was developed from the year over year cost change, and this rate 
was used to develop the learning curves used in the NPRM.
    CARB stated that publicly available data supported lower costs in 
the near term than what the applied learning curve rates would do to 
the battery costs developed by the agencies, and the agencies failed to 
consider new information or data to adjust battery costs.\1260\ CARB 
stated that considering the substantial volume of publicly available 
information and public input to the agencies' previous analysis, 
projected battery costs should have been adjusted even further downward 
for the NPRM. CARB stated that instead, the agencies moved costs upward 
without sufficient justification, and in contrast, the analysis for the 
Proposed Determination and 2016 Draft TAR provided far more 
justification for those modeled battery costs.
---------------------------------------------------------------------------

    \1260\ California Air Resources Board, NHTSA-2018-0067-11873, at 
142-43.
---------------------------------------------------------------------------

    As discussed in Section VI.B.4.d) Cost Learning, above, ICCT 
commented broadly on the change in approach to learning curves since 
the Draft TAR, stating that this change in approach led to lower 
decreases in costs over time in the NPRM than the Draft TAR analysis. 
ICCT compared EPA's Draft TAR learning curves and NPRM learning curves 
for batteries in MYs 2016-2025, concluding that there was a 29% 
reduction in learning for batteries from EPA's Draft TAR analysis to 
the NPRM analysis.
    The agencies considered an array of both present and future cost 
estimates from various public and private sector organizations to 
validate the rate at which battery pack costs declined over time. These 
estimates, in addition to estimates submitted by commenters as 
discussed in BatPaC Inputs and Assumptions and Final Rule Battery Pack 
Costs are shown in Table VI-101. In addition, the agencies had to 
consider how to project learning rates out through 2050, as discussed 
in Section VI.B.4.d) Cost Learning and Section VI.C.3.e)(3) 
Electrification Learning Curves.
    The agencies also assessed and reviewed literature evaluating more 
recent battery technology development.1261 1262 The NPRM 
analysis used a three percent learning rate per year from MY 2033 to MY 
2050. Learning rate forecasts from MY 2033 to MY 2050 for this final 
rule analysis were scaled down in steps from the previous analysis 
based on literature, market research, and Wright's learning curve 
assumptions.
---------------------------------------------------------------------------

    \1261\ MIT Energy Initiative. 2019. Insights into Future 
Mobility. Cambridge, MA: MIT Energy Initiative. Available at http://energy.mit.edu/insightsintofuturemobility.
    \1262\ Islam, E., Kim, N., Moawad, A., Rousseau, A., ``A Large-
Scale Vehicle Simulation Study To Quantify Benefits & Analysis of 
U.S. Department of Energy VTO & FCTO R&D Goals.'' Report to U.S. 
Department of Energy. Contract ANL/ESD-19/10. (forthcoming).
---------------------------------------------------------------------------

    It is difficult to predict which battery chemistry and production 
processes will be prevalent for electrified vehicles in MY 2030, let 
alone for MY 2050. The agencies reviewed potential battery chemistries 
that could come into readiness for adoption at different timeframes, 
such as MY 2030s to MY 2039, and MY 2040 to MY 2050.\1263\ It is 
possible that costs based on other lithium-ion based chemistries will 
learn at the same rate as lithium-ion NMC development. However, the 
same learning effect in battery production may not be additive across 
different chemistries, especially in learning effects related to 
battery production. Accordingly, the learning rates applied between MY 
2030 to MY 2039 considered development and increased volume for the 
same or similar battery chemistries as an NMC battery platform.\1264\ 
Learning curves beyond MY 2040 were flattened further to ensure that 
the cost of batteries did not lower beyond the projected price of the 
raw materials. Further, new chemistries introduced in later years may 
learn at different rates than the curve identified for NMC-based 
chemistries. The battery pack cost learning rate that resulted from 
this exercise produced the schedule that appears in Table VI-96, which 
shows this final rule analysis battery pack cost reduction as function 
of time. By MY 2040, the pack cost has reduced by 54 percent. 
Accordingly, the estimated battery pack cost between MY 2040 and MY 
2050 as shown in Figure VI-43 below shows flatter curve.
---------------------------------------------------------------------------

    \1263\ MIT Energy Initiative. 2019. Insights into Future 
Mobility. Cambridge, MA: MIT Energy Initiative, at p. 79. Available 
at http://energy.mit.edu/insightsintofuturemobility.
    \1264\ For example, an NMC lithium-ion-based platform could move 
from a cathode composition of NMC622 to NMC811.

---------------------------------------------------------------------------

[[Page 24515]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.239

    The reference cost is defined for MY 2020 vehicles, and vehicles 
produced in subsequent years (as well as earlier years) use a per kWh 
cost that is a percentage of the 2020 cost. As the figure shows, the 
cost reduction is rapid through MY 2030, after which cost reductions 
slow considerably. As discussed above, the cost projections assumed 
different battery chemistries and different rates of cost learning.
    The agencies expect there will be incremental improvements in 
battery chemistry, energy density, plant efficiency, and production 
volume over the timeframe modeled in the analysis. While each of these 
factors may have an impact on the rate at which battery costs decline 
over time, the agencies determined that using the same cost learning 
projection method from the NPRM to project learning rates out through 
2050 provided a reasonable method for accounting for something that is 
inherently uncertain. Accordingly, the learning curve used in the NPRM 
and in the final rule represent a composite learning curve irrespective 
of the type of battery chemistry, the production volume necessary to 
achieve economies of scale, or energy density of the battery pack. For 
the final rule, the agencies have performed sensitivity analyses 
varying the battery pack learning rate, and these analyses are 
presented in FRIA Chapter VII.E Sensitivity cases.
(4) Electrified Powertrain Costs
    For the NPRM analysis and carried forward for the final rule 
analysis, the total electrified powertrain costs were developed by 
summing individual component costs. The costs associated with the IC 
engine, transmissions, electric machines, and battery packs were 
combined to create a full-system cost, per Section VI.C.3.e)(2) Non-
battery Electrification Component Costs, Section VI.C.3.e)(1) Battery 
Pack Modeling, Section VI.C.1.g) Engine Costs, and VI.C.2.e) 
Transmissions Costs. This approach assured all technologies 
appropriately contributed to the total system cost.
    The Alliance commented in support of the agencies' accounting 
separately for the subsystems' costs and benefits for CISG, BISG, P2 
hybrid, power split hybrid (PS), and PHEV technologies.\1265\ The 
Alliance noted that these distinctions are important to capture the 
differences between various technologies, which can have separate 
packaging requirements, efficiency potentials, and vehicle 
applications. Ford echoed the Alliance comments on the modeling of 
electric vehicles in the NPRM, stating they supported the use of 
separate cost and benefits modeling for P2 and power split strong 
hybrid technologies.\1266\ Additionally, Ford commented that the 
modeling ``better reflects market realities by recognizing that 
manufacturers cannot simply pass on the entire incremental costs of 
hybrid, plug-in hybrid, and battery electric vehicles to the 
customers.''
---------------------------------------------------------------------------

    \1265\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
12073, at 140.
    \1266\ Ford Motor Company, NHTSA-2018-0067-11928, at 10.
---------------------------------------------------------------------------

    Comments from other stakeholders generally stated that the NPRM 
powertrain sizing approach resulted in costs for complete powertrains 
that were too high compared to other studies or market observations. In 
addition, as discussed in Section VI.C.1.g) Engine Costs, CARB also 
commented that the costs associated with IC engines were not excluded 
from the final costs of BEV vehicles.\1267\ CARB continued, stating 
that ``the final costs of BEV vehicles are higher due to the inclusion 
of the base absolute costs, to which the assigned BEV incremental cost 
would be added.'' The agencies agreed with CARB that inclusion of IC 
engine costs in the BEV cost was an error in the analysis.
---------------------------------------------------------------------------

    \1267\ NHTSA-2018-0067-11873 at p.122.
---------------------------------------------------------------------------

    In response to this comment, the agencies developed absolute costs 
for baseline engines for the CAFE Model so the absolute costs for IC 
engines could be removed from BEVs. In the final rule analysis, when a 
vehicle adopted BEV technology, the costs associated with IC powertrain 
systems were removed. As the vehicle walks through the technology tree, 
becoming a battery electric vehicle, the motor and inverter (ETDS) 
costs replaced the internal IC engine costs. Since the cost of the ETDS 
accounted for significant portion of the

[[Page 24516]]

total cost of electrification, it was important to accurately 
characterize the motor size (motor rating). To do this, the agencies 
used the MY 2017 market data file to compute the average engine power 
for each technology class.
    For SHEVPS and SHEVP2 vehicles, as explained further in Section 
VI.C.3.e)(4)(c) Strong Hybrid Costs, the agencies computed the average 
rating for traction and generator motors across all road load 
combinations using Autonomie simulation runs. Since motor sizing varies 
based on road load levels, the average motor sizes acted as a mid-range 
representation for motor ratings across all road load combinations. The 
full range of motor sizes are driven by road load limits; the motor 
size for initial road load levels (MR0, AERO0 and ROLL0) would be 
larger compared to the motor size for highest level of road load 
reduction (MR6, AERO20 and ROLL20). After calculating the average motor 
size, the agencies applied the $18/kW metric (derived from the EETT 
Roadmap report) for both traction motors and generator motors. As 
discussed earlier, the agencies also used the cost of the CVTL2 as 
proxy to represent the cost of the eCVT used in power-split hybrid 
vehicle systems, and used the cost of the AT8L2 as proxy for the cost 
of the planetary gear set used in the P2 parallel hybrid system. The 
total cost of electrification for power-split hybrid vehicles includes 
the cost of the eCVT transmission, and the total cost of 
electrification for the P2 parallel hybrid vehicles includes the cost 
of the planetary gear set transmission.
    CARB also submitted supplemental comments attempting a cost walk 
for electrified powertrain technologies, stating that inconsistencies 
in the model files and PRIA and lack of documentation about how the 
costs were derived ``[left] the public without the ability to 
understand why the costs are what they are and what should be 
applied.'' \1268\ Accordingly, a cost walk for a vehicle adopting an 
electrified powertrain is shown below. Additional comments on 
electrified powertrain costs are discussed in each individual 
technology section below, along with a discussion of changes made for 
the final rule in response to these comments.
---------------------------------------------------------------------------

    \1268\ California Air Resources Board, NHTSA-2018-0067-12428, at 
25.
---------------------------------------------------------------------------

    For the final rule analysis, the agencies have updated several 
electrification inputs and assumptions in response to these comments, 
as discussed in the previous sections. An example of how the costs are 
applied to a simulated vehicle platform's technology cost is discussed 
here, to assist CARB and other stakeholders in assessing 
electrification technology costs for the final rule analysis. The 
example shows the costs for a vehicle with conventional engine and 
transmission technology as it adds electrification technology.
    The application of the electrification costs to an existing 
platform follows the same basic process for each technology on the 
electrification path. All technology costs used are for the model year 
of the electrification technology application. The first step is the 
process is the removal of the costs associated with the conventional 
drivetrain technologies. The next step is the application of the costs 
associated with the electrification technology. The costs include the 
cost of the engine, if applicable, transmission, non-battery 
components, and the battery pack. After the electrification costs are 
applied, other technology costs, such as aerodynamic or rolling 
resistance technologies are applied.
    The specific example is the Toyota Rav4 LE AWD/XLE AWD simulated 
platform. The platform data were used from the reference run CAFE model 
standard setting vehicle_report.csv result file, augural standards 
results. The change in technology for the simulated platform was 
between MY 2023 and MY 2024. Table VI-107 shows the costing change 
between the MYs.
[GRAPHIC] [TIFF OMITTED] TR30AP20.240

    Table VI-108 shows the costs, and where to find them, for the 
drivetrain components subtracted from the MY 2023 version of the 
platform. The costs for current engine and transmission were 
subtracted. To properly cost the engine it is important to note the 
engine was designated as a 4C1B engine, or, 4 cylinder 1 bank engine 
type. For more information about engine geometry designation in the 
technology input file please see Section VI.A.7 Structure of Model 
Inputs and Outputs.

[[Page 24517]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.241

    The costs for the new electrification technology were then applied. 
For the specific example the simulated vehicle platform is being 
converted to a PHEV20 powertrain. For all the technologies in the 
electrification path two major component groups were always added, the 
battery pack and the non-battery components. Hybrid electric 
technologies will also include the cost for an engine. Table VI-109 
shows the costing data for the non-battery pack electrification 
technology components, and where the cost data can be found.
[GRAPHIC] [TIFF OMITTED] TR30AP20.242

    The battery pack is cost is determined by multiplying the baseline 
battery pack cost by the learn curve factor. Table VI-110 shows the 
calculation of the battery pack costs. The baseline battery costs are 
determined per discussions in Section VI.C.3.e)(1) Battery Pack 
Modeling.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.243

    Table VI-111 shows a summary of the total cost application for the 
technology transition of the Rav4 example platform. The added costs of 
the addition of the LDB technology, improvement from AERO15 to AERO20, 
improvement from MR0 to MR1 are summarized. However, the costing data 
for these technologies can be found in the Technology Input file on the 
`SmallSUV' tab under each technology's respective rows.

[[Page 24518]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.244

    The following sections discuss specific electrification component 
cost comments on the NPRM, responses, and any relevant assumptions for 
the final rule analysis.
a) Micro Hybrid Cost
    As stated in PRIA, the cost of SS12V in NPRM included the cost of 
the battery, learning rate and retail price equivalent.\1269\ The 
assumed direct manufacturing cost (DMC) was the same as was used for 
the Draft TAR and the Proposed Determination,\1270\ but adjusted for 
learning and updated from 2013 to 2016 dollars. Cost learning made the 
cost of SS12V presented in the NPRM slightly lower than the Proposed 
Determination.
---------------------------------------------------------------------------

    \1269\ Footnote n. 364 in PRIA; Table 6-32 and Table 6-33.
    \1270\ Draft TAR Table 5.210.
---------------------------------------------------------------------------

    ICCT compared the agencies' NPRM cost effectiveness estimate for 
SS12V with EPA's Proposed and Final Determination analyses, and 
concluded that the latter analyses found SS12V cost nearly $100 less 
than the agencies found in the NPRM, with a higher effectiveness 
benefit.\1271\ ICCT noted its difficulty in evaluating whether SS12V 
technology was actually cost-effective, since the NPRM CAFE model added 
the incremental cost of BISG over SS12V. ICCT stated that because SS12V 
is not as cost effective as other technologies in the electrification 
technology pathway, such as BISG, the analysis' estimate of SS12V costs 
was exaggerated and resulted in an unrealistic increase in compliance 
costs.
---------------------------------------------------------------------------

    \1271\ International Council on Clean Transportation, 
``Attachment 3_ICCT 15page summary and full comments appendix,'' 
NHTSA-2018-0067-11741, at I-63.
---------------------------------------------------------------------------

    While BISG is more expensive than the SS12V, BISG provides 
additional benefits such as smoother start-stop (reduced vibration 
during each start-stop event), launch assist and/or torque assist 
(during certain sudden acceleration while passing or load at low speed 
for short burst of time). Therefore, the effectiveness of SS12V should 
not be compared to BISG. The agencies have always considered BISG as a 
separate technology. Also, the effectiveness of SS12V in the Proposed 
Determination was determined using ALPHA modeling. A peer reviewer 
noted that ``[a]ccording to the documentation review, ALPHA's stop/
start modeling appears to be very simplistic.'' \1272\ As discussed in 
Section VI.B.3 Autonomie model, the Autonomie tool simulates the 
technology as part of the full vehicle system, accounting for 
interactions with other technologies, and therefore the agencies 
believe the full-vehicle simulations provide more realistic 
effectiveness estimates than the value from the Proposed Determination. 
For these reasons, the agencies disagree with ICCT's assertions. For 
SS12V, the agencies continued to use the costs from the NPRM, which are 
consistent with the Draft TAR and Proposed Determination. The ETDS 
costs presented in the final rule do not include the cost of the 
battery.
---------------------------------------------------------------------------

    \1272\ Peer Review of ALPHA Full Vehicle Simulation Model, at C-
4, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
---------------------------------------------------------------------------

b) Mild Hybrid Cost
    The belt integrated starter generator (BISG) and crank integrated 
starter generator (CISG), sometimes referred to as mild hybrid systems, 
provide idle-stop capability and use a higher voltage battery with 
increased energy capacity over typical automotive batteries. The higher 
voltage allows the use of a smaller, more powerful and efficient 
electric motor/generator which replaces the standard alternator. For 
the NPRM the agencies developed the costs for the mild hybrid systems 
assuming the use of a 115V system. The battery, motor, and supporting 
components were sized and costed based on this voltage level.
    Many commenters asserted that the costs presented in the NPRM 
analysis for BISG and CISG systems were inflated or incorrect.\1273\ 
ICCT noted that because mild hybrid systems were

[[Page 24519]]

widely adopted by the fleet under the augural standards, the high cost 
of those systems had a significant impact on the costs of the 
standards.\1274\
---------------------------------------------------------------------------

    \1273\ International Council on Clean Transportation, NHTSA-
2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
12039; Fiat Chrysler Automobiles, NHTSA-2018-0067-11943; Alliance of 
Automobile Manufacturers, NHTSA-2018-0067-12073; California Air 
Resources Board, NHTSA-2018-0067-11873.
    \1274\ International Council on Clean Transportation, NHTSA-
2018-0067-11741, at I-24.
---------------------------------------------------------------------------

    Meszler Engineering Services noted that the NPRM documentation 
presented BISG/CISG battery costs that were ``not unreasonable,'' and 
that the CAFE model database of battery costs used for NPRM analysis 
included estimates for those electrification technologies that were 
$259 higher than those presented in the NPRM documentation.\1275\ 
Meszler surmised that it initially appeared as if the model may have 
been applying a redundant RPE factor to BISG/CISG costs, but noted that 
the determination that the costs differed from those documented by a 
constant absolute offset made that assumption an unlikely possibility.
---------------------------------------------------------------------------

    \1275\ Meszler Engineering Services, NHTSA-2018-0067-11723 
Attachment 2.
---------------------------------------------------------------------------

    ICCT and UCS both noted the discrepancy between the reported 
battery costs in the PRIA and costs reported in the NPRM Autonomie 
simulation databases.\1276\ ICCT disagreed with the agencies' approach 
to modeling batteries in the NPRM analysis, stating that ``[n]ot only 
is [the Argonne] database exceedingly difficult to access to modify 
battery costs (as battery costs should be a user input), but it makes 
it much harder to see how battery costs affect mild hybrid costs over 
time.'' \1277\ Claimed difficulties aside, ICCT concluded that the 
battery costs were outdated and grossly overstated, based on the tables 
in section 6.3.9.12 of the PRIA and the outputs of the low battery cost 
sensitivity case, which ICCT stated were more closely aligned with EPA 
and other research on battery costs. ICCT presented its own best 
estimate of NPRM BISG costs, stating that they were not able to make 
the PRIA and datafile costs match up.
---------------------------------------------------------------------------

    \1276\ International Council on Clean Transportation, NHTSA-
2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
12039.
    \1277\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
---------------------------------------------------------------------------

    Several commenters noted that the costs of BISG/CISG systems were 
higher for Small Cars/SUVs and Medium Cars than for Medium SUVs and 
Pickup trucks, which the Alliance and FCA described as ``implausible'' 
and ``misaligned with industry understanding,'' and which ICCT 
described as ``contrary to basic engineering logic, which holds that a 
system which would be smaller and have lower energy and power 
requirements would be less expensive, not more.'' \1278\ Both ICCT and 
UCS stated that regardless of alleged errors in costs between 
technology classes, even the lower of the values presented in the PRIA 
overestimated the cost of mild hybrid batteries.\1279\
---------------------------------------------------------------------------

    \1278\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1279\ International Council on Clean Transportation, NHTSA-
2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
12039.
---------------------------------------------------------------------------

    The Alliance and FCA urged the agencies to update the CAFE model to 
address this issue so that the cost of compliance was properly 
reflected in the results. To estimate the impact of the error, the 
Alliance and FCA modified the technology input file so that the Medium 
SUV and Pickup truck electrification costs were changed to be identical 
to the Small Car/SUV and Medium Car costs for SS12V, BISG, and CISG, 
and re-ran the CAFE model to show an estimated $13 billion increase in 
compliance costs under the augural standards with the error 
corrected.\1280\
---------------------------------------------------------------------------

    \1280\ Fiat Chrysler Automobiles, NHTSA-2018-0067-11943; 
Alliance of Automobile Manufacturers, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Conversely, CARB modified the fuel consumption improvement 
estimates for BISG systems to match those predicted by Argonne in a 
recent report after calculating the smallest modified improvement from 
MYs 2015-2025 for five vehicle classes, resulting in efficiency 
improvements of 8.5-11 percent.\1281\ CARB also reduced the non-battery 
costs for Small Car/SUVs to match the non-battery costs for Medium SUV 
and Pickup trucks, which CARB stated still reflected higher costs than 
those previously used by EPA in the Proposed Determination. CARB did 
not modify the battery costs, but did comment that they were overstated 
by approximately 50 percent ``due to the erroneous oversizing of the 
battery.'' CARB's modified run decreased average vehicle technology 
costs by a range of $300-$500 per year, ``reflecting an approximate 25 
percent drop in 2029 model year incremental technology costs to meet 
the existing standards relative to the rollback standards.''
---------------------------------------------------------------------------

    \1281\ California Air Resources Board, NHTSA-2018-0067-11873 
(``Specifically, the fuel consumption improvements modeled by ANL in 
the most recent report for DOE were utilized in place of the 
assumptions used for the Agencies' analysis. As noted above, ANL, 
via Autonomie modeling, identified efficiencies between 8.5 percent 
to 12.7 percent for mild hybrids, relative to both gasoline spark 
ignited and relative to turbocharged gasoline spark ignited across 
five different vehicle classes. Using approximately the smallest 
modeled improvement across the 2015 to 2025 model years for each of 
the five classes, improvements of 8.5 percent-11 percent were 
utilized for a modified CAFE Model run.'').
---------------------------------------------------------------------------

    Commenters also pointed to prior agency analyses, studies, and 
applications of BISG systems to provide examples of what they believed 
BISG system costs should be, with ICCT arguing that the agencies' cost 
values for BISG/CISG systems were contrary to the research and 
evidence.\1282\ HDS noted that the 2018 PRIA estimate was approximately 
double the estimate from the 2016 Draft TAR, that the difference in 
battery costs between those two analyses did not explain the 
difference, and that there was no discussion in the PRIA that did 
so.\1283\
---------------------------------------------------------------------------

    \1282\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1283\ H-D Systems, NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    UCS stated that BISG system costs have already reached that which 
was predicted in EPA's first Final Determination, published in 2017, 
for 2025, and would decline further because of continued volume-based 
learning.\1284\ UCS also cited a 2018 Argonne report that estimated the 
battery component cost for a mild hybrid system to be $159.35, and a 
Chevrolet Malibu eAssist teardown study that estimated total battery 
subsystem direct costs at $166, and battery modules, power 
distribution, and covers at $120 in direct manufacturing costs.\1285\ 
UCS summarized that the aforementioned costs are less than half the 
costs listed in the PRIA and approximately one quarter of the 
``BatPaCCost'' value given in the Argonne input files. UCS also cited 
cost estimates from the 2015 NAS report and two EPA reports, and 
concluded that the agencies did not sufficiently explain why the NPRM 
cost data differed so substantially from this other available 
information.
---------------------------------------------------------------------------

    \1284\ Union of Concerned Scientists, NHTSA-2018-0067-12039.
    \1285\ Id. (citing [Component Cost, ANL 2017k]).
---------------------------------------------------------------------------

    ICCT cited its own 2016 study of supplier costs with estimates for 
48V mild hybrid systems, estimating the system cost at $600-$1,000 
(with costs on the lower side for cars and the higher side for light 
trucks) in the 2025 timeframe.\1286\ ICCT pointed to the RAM 1500 
pickup truck as an example of a vehicle with a BISG system that ``has 
already validated the ICCT figures in 2019.'' ICCT noted that the BISG 
system, branded as eTorque, was first offered as a ``free standing'' 
option on the RAM 1500 truck for $800, and that price was recently 
raised to $1,450. ICCT stated that even with the higher price, applying 
the agencies' RPE of 1.5 means

[[Page 24520]]

that the direct manufacturing cost is less than $1,000, which is less 
than the $1,616 direct manufacturing cost estimate in the NPRM for 2016 
pickup trucks.\1287\ Similarly, UCS cited the $500 premium that General 
Motors charged for the technology on its Chevrolet Silverado pickup 
trucks with eAssist.\1288\
---------------------------------------------------------------------------

    \1286\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1287\ ICCT also stated that the eTorque system offered improved 
performance and driveability and contributes to higher payload and 
towing ratings for 2019 compared with 2018, and noted that the 
agencies ``have completely failed to account for the consumer value 
of the utility benefits'' from the system. The agencies' approach to 
simulating performance neutrality and the consumer benefit of 
increased performance are discussed in Section VI.B.3.a)(6) 
Performance Neutrality.
    \1288\ Union of Concerned Scientists, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    The agencies reviewed all of the comments and information provided. 
It appears there may have been confusion about what costs were used for 
the Draft TAR and NPRM. For the Draft TAR, non-battery BISG costs, 
including learning and RPE, were $1,701 compared to $1,186 for the NPRM 
(both costs in 2018 dollars). Therefore, the costs for the NPRM were 
lower than for the Draft TAR when cost accounting is on an equivalent 
basis.
---------------------------------------------------------------------------

    \1289\ Table 5.131 in Draft TAR ($1,045 x 1.5 = $1567.5 in 
2013$. (Absolute cost, without batteries. This includes learning and 
Retail Price Equivalent).
    \1290\ Table 6-32 in PRIA (Absolute Electrification Cost without 
batteries. This includes learning and Retail Price Equivalent).
    \1291\ See Table I 19--Cost and Mass Estimate of BISG 
components.
[GRAPHIC] [TIFF OMITTED] TR30AP20.245

    The agencies also determined the cost presented by EPA in Draft TAR 
(see Table 5.131 in Draft TAR) was the direct manufacturing cost of the 
BISG system, and not the retail price equivalent. The Draft TAR cost 
estimate in Table VI-112 includes the RPE and costs updated from 2013 
to 2018 dollars. The agencies agree with the commenters about the 
discrepancy in the cost of the battery pack for the BISG system 
presented in PRIA and in CAFE model. To avoid any confusion, Table VI-
112 shows the non-battery costs of the BISG system.
    After considering the comments and reviewing the approach used in 
the NPRM, the agencies agreed updating the cost of the BISG system was 
appropriate for the final rule analysis. Adjustments were based on 
using a 48V BISG system instead of the 115V system used for the NPRM. 
For the final rule, the agencies considered several cost sources, 
including the EPA-sponsored FEV report titled: Light-Duty Vehicle 
Technology Cost Analysis on 2013 Chevrolet Malibu ECO with eAssist BAS 
Technology Study.\1292\ Based on the teardown study, EPA estimated the 
direct manufacturing cost of the BISG system (without batteries) to be 
$1,045 in 2013 dollars. This included a cost adjustment for reduced 
voltage insulation. The agencies also considered the 2019 Dodge Ram 
eTorque system retail price. A cost of $1,195 for water-cooled system 
and $1,450 for air-cooled system in 2018 dollars was deduced from the 
retail price of eTorque assist (BISG) system. The 2015 NAS report 
estimated the cost range of BISG technology at $888 to $1,164 in 2010 
dollars in 2025.\1293\ This is equivalent to a range of $1,020 to 
$1,337.27 in 2018 dollars in 2025. The agencies also reviewed 
confidential business information on BISG cost and mass estimates 
provided by manufacturers.
---------------------------------------------------------------------------

    \1292\ Light Duty Vehicle Technology Cost Analysis 2013 
Chevrolet Malibu ECO with eAssist BAS Technology Study, FEV P311264 
(Contract no. EP-C-12-014, WA 1-9).
    \1293\ Cost, Effectiveness and Deployment of Fuel Economy 
Technologies for Light-Duty Vehicles, National Academy of Sciences, 
2015.
---------------------------------------------------------------------------

    For the final rule analysis, the agencies used the A2Mac1 database 
to develop a bill of materials for BISG systems. The agencies sourced 
cost estimates for the motor, inverter and DC-DC converter from the 
2017 EETT roadmap report.\1294\ The agencies used BatPaC model version 
3.1 to perform a standalone analysis determining the cost of a battery 
pack for the 48V system.1295 1296 Table VI-113 shows the 
cost and mass estimates for BISG components used in the final rule.
---------------------------------------------------------------------------

    \1294\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
    \1295\ A Detailed Vehicle Simulation Process To Support CAFE and 
CO2 Standards for the MY 2021--2026 Final Rule Analysis, at Table 
50.
    \1296\ BatPac 10032018 BISG Version 3.1--28June2018_FINAL.

---------------------------------------------------------------------------

[[Page 24521]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.246

    The agencies compared the cost estimates in the 2017 EETT roadmap 
report and found they aligned well with cost estimates from sources 
cited by commenters. For reference, Table VI-113 above showed the cost 
estimate for BISG system (without the battery) used in Draft TAR, NPRM 
and in Final Rule. Furthermore, the agencies considered the Alliance 
and FCA analysis, provided in their respective comments, recommending 
the use of the same BISG system cost for both cars and 
trucks.1297 1298 This analysis, supplemented with CBI data, 
demonstrated that the costs for implementing BISG systems on different 
vehicle classes was not appreciably different. The agencies agree with 
this assessment. For the final rule analysis, the cost of the BISG 
system is the same for cars, SUVs, and pickups.
---------------------------------------------------------------------------

    \1297\ Fiat Chrysler Automobiles, NHTSA-2018-0067-11943, at 85.
    \1298\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
12073, at 140-42.
---------------------------------------------------------------------------

(c) Strong Hybrid Cost
    In the NPRM and this final rule analysis, the total cost for strong 
hybrids (SHEVP2 and SHEVPS) included the electric machine, battery 
pack, IC engine, and transmission. Discussed earlier in Section 
VI.C.3.d) Electrification Effectiveness Modeling, each strong hybrid 
powertrain is optimized for the given vehicle class by appropriate 
sizing of the electric machine, IC engine and battery pack. 
Accordingly, the costs represent the optimized system. For the NPRM, 
the agencies referred to the ``Assessment of vehicle sizing, energy 
consumption, and cost through large-scale simulation of advanced engine 
technologies'' report to estimate the cost and effectiveness for 
different hybrid systems for the NPRM.\1299\ For the final rule, as 
discussed in Section 2) and further below, the agencies sourced cost 
estimates from the October 2017 U.S. DRIVE report, ``Electrical and 
Electronics Technical Team Roadmap.'' \1300\
---------------------------------------------------------------------------

    \1299\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and 
Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption 
and Cost Through Large Scale Simulation of Advanced Vehicle 
Technologies (ANL/ESD-15/28). United States (2016), available at 
https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
    \1300\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
---------------------------------------------------------------------------

    SHEVP2 and SHEVPS have different characteristics and in turn have 
different costs, as reflected in both the NPRM and this final rule 
analysis. The cost for engines and transmissions for SHEVP2s are based 
on estimates discussed further in Sections VI.C.1 Engine Path and 
VI.C.2 Transmission Path, respectively. The cost for SHEVP2 electric 
machines and battery packs were dependent on their sizes, which were 
optimized by the Autonomie sizing algorithm. SHEVPS total powertrain 
costs includes the optimized battery pack, electric machine, an 
Atkinson engine, and the CVT.
    Many commenters generally stated that the costs of hybrid 
technology were overestimated in comparison to prior agency estimates 
and other publicly available sources, and that the agencies' 
documentation of hybrid system costs was unclear.
    Meszler Engineering Services commented that the net costs of 
vehicles that apply SHEVP2 technology were in error, resulting from the 
way that the CAFE model applied HCR, CEGR and TURBO technology in 
combination with the SHEVP2 strong hybrid system.\1301\
---------------------------------------------------------------------------

    \1301\ Meszler Engineering Services, NHTSA-2018-0067-11723.
---------------------------------------------------------------------------

    HDS claimed that cost estimates for both SHEVP2 and SHEVPS were 
significantly higher than the Draft TAR estimates, differing by a 
factor of about 2 for SHEVP2 and by a factor of 2.5 for SHEVPS, with no 
justification given for the increase in costs.\1302\ HDS noted that the 
SHEVPS cost estimates were particularly surprising since the costs have 
been investigated extensively since that technology was introduced to 
the market over a decade ago. HDS stated that the 2016 TAR estimates 
were in line with other analyses like the NAS

[[Page 24522]]

estimate, and consistent with actual retail price increments observed 
in the market.
---------------------------------------------------------------------------

    \1302\ H-D Systems, NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    HDS also pointed to cost estimates based on teardown studies 
sponsored by EPA and the European Union,\1303\ public cost data 
disclosed by suppliers of hybrid systems, and the retail prices of 
available hybrid vehicles as estimates that contradict the agencies' 
NPRM cost estimates. HDS compared the European Vehicle Market Phase 1 
FEV cost analysis to the costs published by EPA in the TAR, concluding 
that the EU costs ``even at [levels adjusted for the strength of the 
Euro] are quite similar to EPA estimates of $2,650 to $3,300 (depending 
on vehicle size) published in the TAR for the P2 hybrid, and also shows 
that the PS hybrid is just 7 percent more expensive than the P2 
hybrid.'' HDS stated that battery costs have also certainly decreased 
since 2012 when the report was written, so current costs are estimated 
to be approximately $400 less than the values cited above.
---------------------------------------------------------------------------

    \1303\ Id., citing FEV, Light-Duty Vehicle Technology Cost 
Analysis-European Vehicle Market (Phase 1), (2012, updated 2013), 
available at https://www.theicct.org/.
---------------------------------------------------------------------------

    HDS also cited a methodology to estimate costs from retail price 
increments in the market,\1304\ stating that a typical cost-to-retail 
price ratio is 1.5. Applying this methodology, the cost of the SHEVPS 
hybrid as used by Ford and Toyota would be in the $2,500 to $3,000 
range, the cost of a SHEVP2 as used by Hyundai Kia would be $2,250, and 
the cost of a low volume and/or luxury model system would be estimated 
at $3,300 for a SHEVP2.
---------------------------------------------------------------------------

    \1304\ Id. (citing Vincentric Hybrid Analysis, executive 
summary, www.vincentric.com/Home/IndustryReports/HybridAnalysis 
October2014.aspx.).
---------------------------------------------------------------------------

    Similarly, ICCT stated that the agencies failed to analyze properly 
the dozens of hybrid vehicles in the marketplace, their costs which 
were lower than the agencies assumed, and their rapid improvements from 
automakers and suppliers competitively developing lower cost components 
for those vehicles.\1305\ ICCT observed an incremental price increase 
in the analysis for hybrid vehicles under the augural standards of 
approximately $6,600 per hybrid vehicle in 2017 and $4,800 in 2025, and 
concluded that this was not a plausible result considering hybrid 
component costs and full-vehicle prices in the marketplace in 2016 as 
well as the technology improvement that continues to enter the fleet. 
ICCT stated that the agencies must set a maximum cost premium for full 
hybrids of $2,500 in 2017, declining linearly to $1,400 by 2025 for 
mid-size cars and crossovers, with cost components likely scaling by 
vehicle power requirements (up for pickups, down for smaller cars), 
which it stated the agencies must also account for in the modeling.
---------------------------------------------------------------------------

    \1305\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
---------------------------------------------------------------------------

    ICCT stated that the agencies must disclose the basis for the 
``unrealistically high'' hybrid system cost estimates, such that the 
public can clearly connect the bottom-up cost components to full 
vehicle costs for all vehicle models that have hybrid cost 
applied.\1306\ ICCT stated that hybrid system cost estimates are ``one 
of the most important technology cost estimations to assess the Augural 
standards' compliance cost, as the NPRM projects that 22 percent of 
vehicles will need full hybrid systems to meet the augural standards,'' 
and accordingly after disclosing those costs, the agencies must provide 
another opportunity for public comment. Similarly, CARB stated that it 
was unable to decipher the hybrid cost components, and without that 
information could only guess as to why the costs increased relative to 
costs in the Draft TAR and EPA's Proposed Determination.\1307\ As such, 
CARB stated they could not make a conclusion as to whether improper 
battery resizing, incorrectly modeled batteries, or oversized electric 
motors contributed to the overestimation of costs for strong hybrid 
systems.
---------------------------------------------------------------------------

    \1306\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1307\ California Air Resources Board, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    The agencies believe comparing the retail price of P2 or PS hybrid 
to conventional vehicles could be misleading. Even though hybrid 
vehicles may have higher direct manufacturing costs, manufacturers may 
choose not to price it higher than the conventional version of the 
vehicle. In other words, manufacturers may choose to subsidize the cost 
of hybrid technologies to gain overall credit for fleetwide compliance. 
Therefore, the agencies believe that comparing retail price between 
hybrid and conventional vehicles should be done only when other sources 
of information are available to corroborate the differences in retail 
price.
    The agencies also referred to an EPA-sponsored teardown and cost 
estimate report as suggested by HDS. Table VI-114 shows the absolute 
cost of P2 and PS hybrid systems as estimated in the EPA sponsored 
teardown report and the absolute cost estimated in the final rule in 
2018$. As indicated above, the absolute cost in the final rule includes 
the cost of transmissions for the PS and P2 hybrid systems. The EPA 
teardown cost estimate includes the cost of the eCVT for the PS hybrid 
systems only. The P2 hybrid system costs do not include the cost of 
engine and transmission in the table below.
    Although ICCT suggested that the agencies cap the maximum cost 
premium for full hybrids of $2,500 in 2017 and linearly decrease the 
cost to $1,400 by 2025, ICCT did not provide any supporting material to 
suggest that maximum upper limit of $2,500 for full hybrid is 
economically feasible, nor did they provide an example of an existing 
full hybrid vehicle in the marketplace with a technology increase of 
$2,500 in 2017. ICCT also did not make it clear if the costs suggested 
would be applicable to P2 or PS hybrid architecture.
    Based on the comments, the agencies reassessed SHEVP2 and SHEVPS 
cost estimates for the final rule. As discussed above, the agencies 
referred to U.S. DRIVE's October 2017 report, ``Electrical and 
Electronics Technical Team Roadmap'' \1308\ to estimate the cost of 
motors and inverters. The agencies also agreed with commenters and 
referenced the MY 2016 Chevrolet teardown report by UBS to estimate the 
cost of other hybrid components such as wiring harness, cables, 
voltage-step-down DC to DC converters, and on-board chargers. Per 
Section VI.C.3.e)(2) Non-battery Electrification Component Costs, for 
the final rule, the cost of non-battery hybrid system components 
includes the cost of traction motor, motor/generators, high voltage 
cables and connectors, charging cord, and on-board chargers. The cost 
of the planetary gear set is also included in the cost of non-battery 
components. Per Section VI.B.4 Technology Costs, for the final rule, 
the cost of hybrid systems is presented as absolute cost, and not as an 
incremental to some previous technology (absolute cost includes the 
retail price equivalent). The agencies used the cost of the AT8L2 
transmission as a cost proxy for the planetary gear set in P2 hybrid 
systems, and used the cost of CVTL2 transmission as a cost proxy for 
planetary gear set for PS hybrid systems. It should also be noted the 
costs shown here do not include the cost of engine coupled to the 
hybrid system.
---------------------------------------------------------------------------

    \1308\ U.S. DRIVE, Electrical and Electronics Technical Team 
Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
---------------------------------------------------------------------------

    The agencies reviewed the FEV 2010 Ford Fusion HEV teardown report, 
Light Duty Technology Cost Analysis, Power-

[[Page 24523]]

Split and P2 HEV Case Studies.\1309\ In a Split-HEV architecture, there 
are two motors; one motor provides torque while the other motor act as 
a generator to recapture the energy during regenerative braking. The 
report does not capture the cost of motor-generator and the cost of the 
DC to DC converter. The report did not include an extensive teardown of 
a P2 hybrid vehicle, but rather made a cost adjustment for the PS motor 
and inverter to reflect additional cost. Table VI-114 shows the 
breakdown of cost estimates for the electric machine in the 2010 Ford 
Fusion HEV.\1310\ Since the costs were developed in 2009$, the cost 
estimates for the same components are presented in 2018$. Table VI-115 
shows the cost estimate for electric machines for a midsize passenger 
car for MY 2017 in 2018$.\1311\ The cost is estimated using the EETT 
Roadmap report as explained earlier. Since EPA uses indirect cost 
multiplier (ICM) to determine the final retail price, and ICMs vary for 
different technologies, the agencies compared the direct manufacturing 
cost from report to the direct cost estimate in the final rule.
---------------------------------------------------------------------------

    \1309\ Light Duty Technology Cost Analysis, Power-Split and P2 
HEV Case Studies, EPA-420-R-11-015 (November 2011), available at 
https://nepis.epa.gov/Exe/ZyPDF.cgi/P100EG1R.PDF?Dockey=P100EG1R.PDF.
    \1310\ Table D-4 (components considered are transmission, power 
distribution cables and Inverter). The cost of inverter is from 
Table D-11.
    \1311\ Average peak power for the traction motor used in this 
final rule is 72kW, and 37kW continuous power for the generation 
motor.
---------------------------------------------------------------------------

    The direct manufacturing cost estimated in the Light Duty 
Technology Cost Analysis, Power-Split and P2 HEV Case Studies published 
for EPA is $3,689.28 in 2018$, and direct manufacturing cost estimated 
for electric machines in this final rule is $4,355.82. As mentioned 
before, the cost of the motor-generator and the cost of the DC to DC 
converter is not captured in that report.
[GRAPHIC] [TIFF OMITTED] TR30AP20.247

[GRAPHIC] [TIFF OMITTED] TR30AP20.248

(d) PHEV Cost
    Plug-in hybrid vehicles' costs were developed similar to strong 
hybrids for the NPRM analysis and the final rule analysis. The plug-in-
hybrid system components were optimized, per Section VI.C.3.d)(2) 
Modeling and Simulating Vehicles with Electrified Powertrains in 
Autonomie and the resultant systems were used to determine costs, per 
Battery Pack Modeling and Non-battery Electrification Component Costs. 
Per Section VI.C.3.c) Electrification Adoption Features, the agencies 
used one engine technology and one transmission technology per plug-in 
hybrid architecture type.
    For PHEVs following SHEVP2 on the hybrid/electric architecture 
path, per Section VI.C.3.a)(1) Electrification technologies, the total 
cost of the technology package was determined from summing the costs of 
the TURBO1 engine, the AT8L2 transmission, and the battery and non-
battery electrification technology components. For PHEVs following 
SHEVPS on the hybrid/electric architecture path, per Section 
VI.C.3.a)(1) Electrification technologies, the total cost of the 
technology package was determined from summing the costs of the 
Atkinson engine, the CVT transmission, and the battery and non-battery 
electrification technology components.
    CARB provided observations about non-battery component costs for 
PHEVs,

[[Page 24524]]

arguing that what the agencies asserted for the incremental costs of a 
PHEV over a strong hybrid vehicle are not supported in the 
market.\1312\ CARB cited the Toyota Prius Prime and Hyundai Sonata as 
examples of vehicles that share most of their components with their 
non-plug-in hybrid counterparts, with components like the on-board 
charger and higher voltage, larger energy capacity battery pack 
excepted. CARB stated the agencies' lack of discussion about how non-
battery component costs were developed made it ``virtually impossible 
to understand what the drivers are for the increases in costs relative 
to the Agencies' previous analysis for the 2016 Draft TAR and EPA's 
Proposed Determination.'' CARB concluded that the available PHEV market 
offerings do not support the higher costs relative to the Draft TAR and 
EPA's Proposed Determination analyses, and no justification was 
provided for the change.
---------------------------------------------------------------------------

    \1312\ California Air Resources Board, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    The agencies agree with CARB that the incremental costs of PHEV 
over strong hybrid costs were too high, and that values were not 
supported by the market. In response to this comment, the agencies 
updated the non-battery component costs as well as the battery costs to 
better reflect the market values. In addition, the agencies have 
optimized the Autonomie modeling in a way to maintain the same engine, 
transmission and other components from a SHEVP2 or SHEVPS moving to a 
PHEV20/50 or PHEV20T/50T.\1313\ For further discussions on PHEV 
modeling and updates, see Section VI.C.3.a)(1) Electrification 
technologies and Section VI.C.3.d) Modeling and Simulating Vehicles 
with Electrified Powertrains in Autonomie. The updates discussed here 
and applied to the final analysis resulted in values that more 
accurately represented PHEV technology costs.
---------------------------------------------------------------------------

    \1313\ I.e., a SHEVP2 with a turbocharged engine may adopt 
PHEV20T or PHEV50T technology, but a SHEVPS will only ever adopt 
PHEV20 or PHEV50 technology, as the SHEVPS do not use turbocharged 
engines.
---------------------------------------------------------------------------

(e) BEV Cost
    For the NPRM and this final rule analysis, the total costs of BEVs 
included optimized battery pack and electric machine costs. Like the 
other electrified powertrains, Autonomie optimized both the size of the 
battery pack and electric machine to fulfill the performance neutrality 
requirements for each vehicle. Further discussion on electrification 
technology component sizing and optimization is provided in Section 
VI.C.3.d) Modeling and Simulating Vehicles with Electrified Powertrains 
in Autonomie. Discussion on electrification component costing is 
provided in Battery Pack Modeling and Non-battery Electrification 
Component Costs. When computing the total cost of a vehicle, the 
agencies remove the costs of the IC engines and transmission when a 
conventional or hybridized powertrain adopts BEV technologies. In 
Section VI.C.1 Engines Path and Section VI.C.22 Transmission, the 
agencies discussed the absolute costs used for engine and transmission 
technologies in the final rule analysis.
    ICCT stated that if the agencies had considered BEV battery and 
other component costs correctly, cost parity would be reached with 
conventional combustion vehicles in the 2025-2027 timeframe.\1314\ ICCT 
went on to allege that if the agencies removed all constraints on 
electric vehicles,\1315\ they would appropriately realize that the 2025 
standards are more cost-effective if electric vehicles are included.
---------------------------------------------------------------------------

    \1314\ International Council on Clean Transportation, NHTSA-
2018-0067-11741.
    \1315\ As discussed above, the agencies believe that ICCT 
misunderstood the agencies' statutory obligations and the 
differences between the standard setting modeling scenario and the 
``real-world'' modeling scenario. The agencies did not apply 
additional constraints on BEVs in the NPRM analysis.
---------------------------------------------------------------------------

    The agencies disagree with ICCT's statement that BEVs would reach 
parity to IC engines by the 2025-2027 timeframe. For this final rule 
analysis, the agencies have updated the battery pack costs, electric 
machine costs, and excluded costs of IC engines and transmission when a 
vehicle was converted to a BEV. However, the costs still did not reach 
parity within the rulemaking time frame. Furthermore, NHTSA notes that 
the decision to exclude BEV technology from the CAFE program standard-
setting analysis is not a choice made by the agency, but a statutory 
requirement.\1316\
---------------------------------------------------------------------------

    \1316\ See 49 U.S.C. 32902(h).
---------------------------------------------------------------------------

(f) FCV Cost
    For the NPRM and the final rule analysis the agencies considered 
fuel cell vehicle technology advancements in hydrogen storage tanks, 
sensors and control systems, and market penetration.\1317\ The agencies 
are also considered the availability of hydrogen refueling stations 
across the country and cost of compressed hydrogen.1318 1319 
Although the agencies did not receive any comments on the cost of fuel 
cell vehicles, the agencies updated the cost of hydrogen storage tanks 
and fuel cells based on a cost analysis from Department of Energy 
(DOE), Office of Energy Efficiency and Renewable Energy (EERE), Fuel 
Cell Technologies Office.\1320\
---------------------------------------------------------------------------

    \1317\ The agencies referenced EPA's 2018 Automotive Trends 
Report, available at https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF, for information about FCV market 
penetration.
    \1318\ MIT Energy Initiative. Insights into Future Mobility 
(2019). Cambridge, MA: MIT Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
    \1319\ U.S. Department of Energy, Alternative Fuels Data Center: 
Alternative Fueling Station Counts by State: https://afdc.energy.gov/stations/states (last visited January 3, 2020).
    \1320\ James et al., Final Report: Hydrogen Storage System Cost 
Analysis (September 2016), available at https://www.osti.gov/servlets/purl/1343975.
---------------------------------------------------------------------------

    The DOE estimates that the cost of a compressed gas storage system 
is around $28/kWh (assumed rate of production of 10,000 units per 
year). The hydrogen fuel price ranges from $12.85 to $16 per kilogram, 
which translates to approximately $5.60 per gallon on an equivalent 
energy basis.\1321\
---------------------------------------------------------------------------

    \1321\ California Fuel Cell Partnership: https://cafcp.org/content/cost-refill (last visited January 3, 2020).
---------------------------------------------------------------------------

    Table VI-116 shows the evolution of the fuel cell vehicle costs 
from the Draft TAR to final rule (costs include the fuel cell, control 
systems, motors, inverters, hydrogen storage tanks, wiring harness, 
hydrogen fuel sending lines, safety systems, sensors and hardware for 
mounting and installation). The cost of the battery pack and battery 
management system is not included in the cost of the fuel cell vehicle.
[GRAPHIC] [TIFF OMITTED] TR30AP20.249


[[Page 24525]]


4. Mass Reduction
    Mass reduction is a relatively cost-effective means of improving 
fuel economy and reducing CO2 emissions, and vehicle 
manufacturers are expected to apply various mass reduction technologies 
to meet fuel economy and CO2 standards. Reducing vehicle 
mass can be accomplished through several different techniques, such as 
modifying and optimizing vehicle component and system designs, part 
consolidation, and adopting lighter weight materials (advanced high 
strength steel, aluminum, magnesium, and plastics including carbon 
fiber reinforced plastics). The cost for mass reduction depends on the 
type and amount of materials used, the manufacturing and assembly 
processes required, and the degree to which changes to plants and new 
manufacturing and assembly equipment is needed. In addition, 
manufacturers may develop expertise and invest in certain mass 
reduction strategies that may affect the approaches for mass reduction 
they consider and the associated costs. Manufacturers may also consider 
vehicle attributes like noise-vibration-harshness (NVH), ride quality, 
handling, and various acceleration metrics when considering how to 
implement any mass reduction strategy. See Section VI.B.3.a)(5) 
Maintaining Vehicle Attributes for more details.
    The automotive industry uses different metrics to measure vehicle 
weight. Some commonly used measurements are vehicle curb weight,\1322\ 
gross vehicle weight (GVW),\1323\ gross vehicle weight rating 
(GVWR),\1324\ gross combined weight (GCVW),\1325\ and equivalent test 
weight (ETW),\1326\ among others.
---------------------------------------------------------------------------

    \1322\ This is the weight of the vehicle with all fluids and 
components but without the drivers, passengers, and cargo.
    \1323\ This weight includes all cargo, extra added equipment, 
and passengers aboard.
    \1324\ This is the maximum total weight of the vehicle, 
passengers, and cargo to avoid damaging the vehicle or compromising 
safety.
    \1325\ This weight includes the vehicle and a trailer attached 
to the vehicle, if used.
    \1326\ For the EPA two-cycle regulatory test on a dynamometer, 
an additional weight of 300 lbs. is added to the vehicle curb 
weight. This additional 300 lbs. represents the weight of the 
driver, passenger, and luggage. Depending on the final test weight 
of the vehicle (vehicle curb weight plus 300 lbs.), a test weight 
category is identified using the table published by EPA according to 
40 CFR 1066.805. This test weight category is called ``Equivalent 
Test Weight'' (ETW).
---------------------------------------------------------------------------

    The vehicle curb weight is the most commonly used measurement when 
comparing vehicles. A vehicle's curb weight is the weight of the 
vehicle including fluids, but without a driver, passengers, and cargo.
    A vehicle's glider weight, which is vehicle curb weight minus the 
powertrain weight, is used to track the potential opportunities for 
weight reduction not including the powertrain. A glider's subsystems 
may consist of the vehicle body, chassis, interior, steering, 
electrical accessory, brake, and wheels systems. However, as noted in 
the PRIA, the definition of a glider may vary from study to study (or 
even simulation to simulation).
    Each of the subsystems presents an opportunity for weight 
reduction; however, some weight reduction is dependent on the weight 
reduction of other subsystems. The agencies characterize mass reduction 
as either primary mass reduction or secondary mass reduction. Primary 
mass reduction involves reducing mass of components that can occur 
independent from the mass of other components. For example, reducing 
the mass of a hood (e.g., replacing a steel hood with an aluminum hood) 
or reducing the mass of a seat are examples of primary mass reduction 
because each can be implemented independently. Other components and 
systems that may contribute to primary mass reduction include the 
vehicle body, chassis, and interior components.
    When significant primary mass reduction occurs, other components 
designed based on the mass of primary components may be redesigned as 
well. An example of a subsystem where secondary mass reduction can be 
applied is the brake system. If the mass of primary components is 
reduced sufficiently, the resulting lighter weight vehicle could safely 
maintain braking performance and attributes with a lighter weight brake 
system. Other examples of components where secondary mass reduction can 
be applied are wheels and tires.
    For this analysis, the agencies consider mass reduction 
opportunities from the glider subsystems of a vehicle first, and then 
consider associated opportunities to downsize the powertrain, which are 
accounted for separately.\1327\ As explained later, in the Autonomie 
simulations, the glider system includes both primary and secondary 
systems from which a percentage of mass is reduced for different glider 
weight reduction levels; specifically, the glider includes the body, 
chassis, interior, electrical accessories, steering, brakes and wheels. 
The model sizes the powertrain based on the glider weight and the mass 
of some of the powertrain components in an iterative process. The mass 
of the powertrain depends on the powertrain size. Therefore, the weight 
of the glider impacts the weight of the powertrain.\1328\ See Section 
VI.B.3.a)(3) Vehicle models for Autonomie and Section VI.B.3.a)(4) How 
Autonomie Sizes Powertrains for Full Vehicle Simulation for more 
details.
---------------------------------------------------------------------------

    \1327\ When the mass of the vehicle is reduced by an appropriate 
amount, the engine may be downsized to maintain performance. See 
Section VI.B.3.a)(5) Maintaining Vehicle Attributes] and Section 
VI.B.3.a)(6) Performance Neutrality for more details.
    \1328\ Since powertrains are sized based on the glider weight 
for the analysis, glider weight reduction beyond a threshold amount 
during a redesign will lead to re-sizing of the powertrain. For the 
analysis, the glider was used as a base for the application of any 
type of powertrain. A conventional powertrain consists of an engine, 
transmission, exhaust system, fuel tank, radiator and associated 
components. A hybrid powertrain also includes a battery pack, 
electric motor(s), generator, high voltage wiring harness, high 
voltage connectors, inverter, battery management system(s), battery 
pack thermal system, and electric motor thermal system.
---------------------------------------------------------------------------

    The agencies use glider weight to apply non-powertrain mass 
reduction technology, and use Autonomie simulations to determine the 
size of the powertrain and corresponding powertrain weight for the 
respective glider weight. The combination of glider weight (after mass 
reduction) and re-sized powertrain weight equal the vehicle curb 
weight. See Section VI.C.4.d)(1) glider mass and mass reduction 
subsection below for more detail on glider mass and glider mass 
reduction.
(a) Mass Reduction in the CAFE Model
    Several studies have explored the amount of vehicle mass reduction 
that is feasible in the rulemaking timeframe and the cost for that mass 
reduction.1329 1330 1331 1332 Those studies were sponsored 
by the agencies, CARB, ICCT, the automotive industry, and material 
manufacturers, and are discussed in Section VI.C.4.e)(1), below. All of 
the studies showed that the maximum feasible amount of mass reduction 
that can be applied in the rulemaking timeframe is around 20 percent of 
a baseline vehicle's curb weight. The National Academies of Sciences 
similarly concluded, based on some of these same studies along with 
other information, that it is feasible to

[[Page 24526]]

reduce up to 20 percent of the mass of the vehicle.\1333\
---------------------------------------------------------------------------

    \1329\ DOT HS 811 692: Investigation of Opportunities for 
Lightweight Vehicles Using Advanced Plastics and Composites.
    \1330\ A Review of the Safety of Reduced Weight Passenger Cars 
and Light Duty Trucks by Michigan Manufacturing Technology Center, 
October 2018.
    \1331\ ATG Silverado Body Light weighting Study, Aluminum 
Technology Group, January 2017.
    \1332\ 2013 NanoSteel Intensive Body-In-White, EDAG and 
NanoSteel Company Inc.
    \1333\ Cost, Effectiveness and Deployment of Fuel Economy 
Technologies for Light-Duty Vehicles, National Academy of Sciences, 
2015, at 212 .
---------------------------------------------------------------------------

    As discussed in Section VI.C.4.e), the mass reduction studies show 
that the cost for mass reduction increases progressively as the amount 
of mass reduction increases. In other words, lower levels of mass 
reduction are more cost effective than higher levels of mass reduction. 
As in past rulemakings, the agencies have considered multiple levels of 
mass reduction to provide options similar to what manufacturers could 
consider at vehicle redesigns.
    For the NPRM, the agencies included five levels of mass reduction 
with a maximum of 20 percent glider mass reduction, corresponding to 10 
percent curb mass reduction, using the assumption that the glider was 
50 percent of curb weight. Table VI-117 shows the glider and curb 
weight mass reduction levels for each level of mass reduction 
considered in the NPRM analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.250

    The agencies received a number of comments suggesting that the 
amount of mass reduction allowed should be 20 percent of curb weight, 
as well as suggestions that the agencies should assume the glider 
represents 75 percent of the vehicle's curb weight. These comments are 
addressed in more detail in Section VI.C.4.d) below, but some 
understanding of how the glider share assumption affects the maximum 
amount of mass reduction allowed in the CAFE model is required here.
    Several commenters stated that the agencies should allow further 
levels of mass reduction technology improvements in the CAFE model. For 
example, ICCT commented that the agencies must revise their treatment 
of mass reduction because studies have demonstrated that at least 20% 
mass reduction of curb weight is available for adoption across vehicle 
classes by 2025. \1334\ ICCT stated that based on these studies, the 
agencies must increase the maximum available mass reduction potential 
levels to include up to 20% and 25% mass reduction of curb weight, as 
the industry ``will cost-effectively deploy at least 15% vehicle curb 
mass reduction in the 2025 timeframe at net zero cost.'' ICCT caveated 
that amount of mass reduction seems less likely in smaller cars, which 
typically employ lower levels of mass reduction, so a constraint of 7.5 
percent mass reduction as was applied in the Draft TAR would be 
appropriate for those vehicles.
---------------------------------------------------------------------------

    \1334\ NHTSA-2018-0067-11741. ICCT also alleged that the 
agencies intentionally disregarded the studies that presented this 
result; those comments are discussed in Section VI.C.4.e) Mass 
Reduction Costs, below.
---------------------------------------------------------------------------

    ICCT also commented that there were numerous material improvements 
in development that were not considered in the rule, including but not 
limited to higher strength aluminum, improved joining techniques for 
mixed materials, third-generation steels with higher strength and 
enhanced ductility, a new generation of ultra-high strength steel cast 
components, and metal/plastic hybrid components, among other 
technologies mentioned in ICCT's working paper on light-weighting.
    In assessing these comments, the agencies reconsidered the mass 
reduction studies and available reports and agreed that additional 
levels of mass reduction should be available for the final rule 
analysis. In response to comments, the agencies made two adjustments to 
allow higher levels of mass reduction in the analysis. First, as 
explained in Section VI.C.4.d)(1), below, the agencies increased the 
glider percentage of vehicle curb weight used for the analysis from 50 
percent to 71 percent. As explained in that section, increasing the 
glider percentage also increases the amount of curb weight reduction 
for all levels of mass reduction. Second, the agencies created another 
level of mass reduction (MR6) in the CAFE model, which represents a 
significant application of carbon fiber in the vehicle to achieve 
nearly 30 percent reduction in glider weight (which approximately 
translates to 20 percent reduction in vehicle curb weight). For 
example, incorporating a carbon fiber tub,\1335\ or a carbon fiber 
monocoque with aluminum sub frame in the front and back,\1336\ or a 
carbon fiber splitter and carbon fiber wheels,\1337\ allows for greater 
levels of mass reduction, albeit at a very high cost. These 
technologies are not ready for high volume production vehicles.
---------------------------------------------------------------------------

    \1335\ The BMW i3 and BMW i8, which are about 20 percent lighter 
than an average MY 2017 vehicle, use a carbon fiber tub.
    \1336\ The Alfa Romeo 4c/4c Spider, which is about 20 percent 
lighter than an average MY 2017 vehicle, uses this design.
    \1337\ The Ford Shelby GT350R which is about 20 percent lighter 
than an average MY 2017 vehicle, uses this design.
---------------------------------------------------------------------------

    Table VI-118 shows the levels of mass reduction technology 
available for application in the final rule analysis, with the 
associated glider weight percentage reduction and the percentage curb 
weight reductions for passenger cars and light trucks. As discussed in 
Section VI.C.4.c) below, the agencies declined to place a constraint on 
the amount of mass reduction technology that smaller cars could adopt.

[[Page 24527]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.251

    The agencies continue to believe the maximum feasible mass 
reduction levels identified in comprehensive design studies, such as 
those discussed in Section VI.C.4 Mass Reduction Costs are the most 
reliable for projecting the maximum amount of mass reduction in the 
rulemaking timeframe, and therefore have determined MR6 is the highest 
level that should be used for the final rule analysis. While the 
information provided by ICCT on newer materials and manufacturing and 
assembly methodology is interesting and relevant, this information, by 
itself, is insufficient to assess the amount of mass reduction that is 
feasible and the cost for the mass reduction. ICCT did not provide a 
comprehensive analysis showing a design concept that maintains vehicle 
attributes and performance, such as noise, vibration and harshness, 
stiffness, handling, compliance with NHTSA safety standards, good 
performance under NHTSA NCAP and IIHS rating systems, and other 
criteria. The various studies in Section VI.C.4.e) Mass Reduction 
considered those factors to varying degrees. Without that rigorous 
analysis, the actual amount of mass reduction that could be enabled 
through the use of those materials and methods described by ICCT, and 
the cost of achieving that mass reduction, would be highly speculative. 
As explained in Section VI.C.4.e) Mass Reduction below, the agencies 
determined the NHTSA-sponsored design studies remain a reasonable basis 
for estimating a feasible amount of mass reduction and the cost for 
mass reduction in the rulemaking timeframe, because those studies 
considered a wide range of materials (including advanced materials) and 
design solutions.
(b) Analysis Fleet Mass Reduction Assignments
    The agencies included an estimated level of mass reduction 
technology for each vehicle model in the MY 2016 analysis fleet for the 
NPRM, and have updated the estimates for the MY 2017 analysis fleet for 
the final rule analysis. The methodology used to provide each vehicle 
model an appropriate initial mass reduction technology level for 
further improvements was described in detail in the Draft TAR (when 
NHTSA first employed this methodology), in the PRIA accompanying the 
NPRM, and is reproduced here, in part, to provide additional context to 
the agencies' responses to comments on analysis fleet mass reduction 
assignments. The methodology used in this final rule was unchanged from 
the NPRM.
    For the Draft TAR, NHTSA/Volpe Center staff developed regression 
models to estimate curb weights based on other observable attributes. 
With regression outputs in hand, Volpe evaluated the distribution of 
vehicles in the analysis fleet. In addition, vehicle platforms were 
evaluated based on the sales-weighted residual of actual vehicle curb 
weights versus predicted vehicle curb weights. Based on the actual curb 
weights relative to predicted curb weights, platforms (and the 
subsequent vehicles) were assigned a baseline mass reduction level (MR0 
through MR6). For the NPRM and final rule analysis, the agencies 
followed a similar procedure for the MY 2016 and MY 2017 analysis 
fleets.
    To develop the curb weight regressions, the agencies grouped 
vehicles into three separate body design categories for analysis: 3-
Box, 2-Box, and Pick-up.
[GRAPHIC] [TIFF OMITTED] TR30AP20.252

    For the NPRM and final rule analysis, the agencies retained the MY 
2015 regressions for 3-Box and 2-Box vehicles, however the pickup 
category regression was updated in response to comments on the Draft 
TAR. The

[[Page 24528]]

agencies trained a new regression with EPA MY 2014 data and added pick-
up bed length as an independent variable. As a result of stepping back 
to MY 2014 data for the pick-up regression, the training data did not 
include the all-aluminum body Ford F-150 in the calculation of the 
baseline. The advanced F-150 in the MY 2015 pick-up regression 
meaningfully affected Draft TAR regression statistics because the F-150 
accounted for a large portion of observations in the analysis fleet, 
and the F-150 included advanced weight savings technology.
    The agencies leveraged many documented variables in the analysis 
fleet as independent variables in the regressions. Continuous 
independent variables included footprint (wheelbase x track width) and 
powertrain peak power. Binary independent variables included strong HEV 
(yes or no), PHEV (yes or no), BEV or FCV (yes or no), all-wheel drive 
(yes or no), rear-wheel drive (yes or no), and convertible (yes or no). 
In addition, for PHEV and BEV/FCV vehicles, the capacity of the battery 
pack was included in the regression as a continuous independent 
variable. In some body design categories, the analysis fleet did not 
cover the full spectrum of independent variables. For instance, in the 
pickup body style regression, there were no front-wheel drive vehicles 
in the analysis fleet, so the regression defaulted to all-wheel drive 
and left an independent variable for rear-wheel drive.
    Furthermore, the agencies evaluated alternative regression 
variables in response to comments from vehicle manufacturers on the 
NHTSA/Volpe analysis in the Draft TAR.\1338\ The agencies evaluated 
regressions including overall dimensions of vehicles, such as height, 
width, and length, instead of and in addition to just wheelbase and 
track width. The experimental regression variables only marginally 
changed predicted curb weight residuals as a percentage of predicted 
curb weight, at an industry level and for most manufacturers. The 
results were not significantly different, and therefore the agencies 
opted not to add these variables to regressions or replace independent 
variables presented in Draft TAR with new variables.
---------------------------------------------------------------------------

    \1338\ PRIA at 407.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.253


[[Page 24529]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.254


[[Page 24530]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.255

BILLING CODE 4910-59-C
    Each of the three regressions produced outputs effective for 
identifying vehicles with a significant amount of mass reduction 
technology in the analysis fleet. Many coefficients for independent 
variables provided clear insight into the average weight penalty for 
the utility feature. In some cases, like battery size, the relatively 
small sub-sample size and high collinearity with other variables 
confounded coefficients.
    By design, no independent variable directly accounted for the 
degree of weight savings technology applied to the vehicle. Residuals 
of the regression captured weight reduction efforts and noise from 
other sources.
    The agencies received many comments on the Draft TAR encouraging 
the use of observed technologies in each vehicle, and in each vehicle 
subsystem to assign levels of mass reduction technology. As a practical 
matter, the agencies cannot conduct a tear down study and detailed cost 
assessment for every vehicle in every model year. However, upon review 
of many vehicles and their subsystems, the agencies recognized a few 
vehicles with MR0 or MR1 assignments in NHTSA's analysis of the Draft 
TAR that contained some advanced weight savings technologies, yet these 
vehicles and their platforms still produced ordinary residuals. 
Engineers from industry confirmed important factors other than glider 
weight savings and the independent variables considered in the 
regressions may factor into the use of lightweight technologies. Such 
factors included the desire to lower the center of gravity of a 
vehicle, improve the vehicle weight distribution for handling, optimize 
noise-vibration-and-harshness, increase torsional rigidity of the 
platform, offset increased vehicle content, and many other factors. In 
addition, engineers highlighted the importance of sizing shared 
components for the most demanding applications on the vehicle platform; 
optimum weight savings for one platform application may not be suitable 
for all platform applications. For future analysis, the agencies will 
look for practical ways to improve the assessment of mass reduction 
content and the forecast of incremental mass reduction costs for each 
vehicle.
    Figure VI-44 below shows results from the pickup truck regression 
on predicted curb weight versus actual curb weight. Points above the 
solid regression line represent vehicles heavier than predicted (with 
lower mass reduction technology levels); points below the solid 
regression line represent vehicles lighter than predicted (with higher 
mass reduction technology levels). The dashed lines in the Figure VI-44 
show the thresholds (5, 7.5, 10, 15, 20 and 28 percent of glider 
weight). Final rule glider weight assumption is 71 percent of vehicle 
curb weight.
BILLING CODE 4910-59-P

[[Page 24531]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.256

BILLING CODE 4910-59-C
    For points with actual curb weight below the predicted curb weight, 
the agencies used the residual as a percent of predicted weight to get 
a sense for the level of current mass reduction technology used in the 
vehicle. Notably, vehicles approaching -20% curb weight widely use 
advanced composites throughout major vehicle systems, and few examples 
exist in the MY 2016 fleet.\1339\
---------------------------------------------------------------------------

    \1339\ This evidence suggests that achieving a 20% curb weight 
reduction for a production vehicle with a baseline defined with this 
methodology is extremely challenging, and requires very advanced 
materials and disciplined design.
---------------------------------------------------------------------------

    Generally, residuals of regressions as a percent of predicted 
weight appropriately stratified vehicles by mass reduction level. Most 
vehicles showed near zero residuals or had actual curb weights close to 
the predicted curb weight. Few vehicles in the analysis fleet were 
identified with the highest levels of mass reduction. Most vehicles 
with the largest negative residuals have demonstrably adopted advanced 
weight savings technologies at the most expensive end of the cost 
curve.
    To validate the residuals, the agencies estimated the mass 
reduction technology level for several vehicle models in the analysis 
fleet and compared those estimates to the numerical results from the 
regression analysis. To estimate the mass reduction technology level 
for the selected vehicles, the agencies conducted an in-depth review of 
available information on the materials, design, and last redesign year 
for those vehicle models, and compared that information with the 
designs and materials used in the mass reduction feasibility and cost 
studies summarized in Section VI.C.4.e), below.

[[Page 24532]]

That comparison showed good agreement with the technology levels from 
the regression analysis.
    The agencies believe the regression methodology is a technically 
sound methodology for estimating mass reduction levels in the analysis 
fleet.
    As part of their comments stating the NPRM modeling reflected 
reality better than the Draft TAR and Proposed Determination analyses, 
Toyota commented broadly that the MY 2016 baseline fleet used in the 
NPRM encompassed powertrain and tractive energy (including mass 
reduction) improvements more representative of vehicles on the road 
today.\1340\ Toyota noted that the 2016 baseline fleet generally 
contained higher levels of technology compared to the MY 2014 and MY 
2015 baseline fleets, and included a comparison of its initial fleet 
mass reduction assignments in the Draft TAR and the NPRM. Toyota showed 
how moving further up the technology tree (e.g., starting with a 
baseline that includes higher levels of technology) for certain 
pathways such as mass reduction increased costs exponentially. Toyota 
stated that the NPRM underestimated mass reduction cost values.
---------------------------------------------------------------------------

    \1340\ NHTSA-2018-0067-12098.
---------------------------------------------------------------------------

    While a more specific discussion of costs is located in Section 
VI.C.4.e), the agencies agree with Toyota's assessment that the costs 
for mass reduction technology increase exponentially as progressively 
higher levels of mass reduction are incorporated. Having an accurate 
assessment of baseline technology levels ensures that the subsequent 
application of technology and its associated costs is correctly 
accounted for.
    C.A.R produced a report in response to the Draft TAR that generally 
agreed with the regression methodology of using observed vehicle 
attributes for estimating mass reduction levels, as opposed to 
comparing vehicle curb weight from a newer model year to a previous 
generation of the same vehicle, pointing to several of the limitations 
discussed above.\1341\
---------------------------------------------------------------------------

    \1341\ EPA Mass Reduction Analysis--Observations and 
Recommendations, Center for Automotive Research, October 2017 (page 
15), available at https://www.cargroup.org/wp-content/uploads/2017/10/EPA-MR-Analysis-Critique_Oct-5_final.pdf.
---------------------------------------------------------------------------

    Both ICCT and H-D Systems commented on the methodology for 
identifying mass reduction technology levels in the analysis fleet, 
with ICCT broadly stating that by placing additional mass reduction 
technology in the baseline, the agencies artificially removed ``the 
most cost-effective lightweighting from future use, which incorrectly 
increases the costs of all subsequent mass-reduction in the compliance 
modeling.'' \1342\
---------------------------------------------------------------------------

    \1342\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    ICCT claimed that the agencies unjustifiably increased the amount 
of vehicle mass reduction technology present in the 2016 baseline fleet 
from the 2015 baseline used in the Draft TAR, stating that the 2015 
Draft TAR fleet had 26 percent of vehicles sold with some level of mass 
reduction applied (MR1 or a higher level), whereas the 2016 NPRM fleet 
had 47 percent of vehicles sold with some level of mass reduction 
applied. In addition to faulting the agencies for not acknowledging the 
change and not attempting to justify it, ICCT stated that the 2016 
analysis fleet mass reduction assignments were overstated, as ``it 
appears that the agencies have applied mass reduction technology to 
vehicles in the model that did not have mass reduction applied in the 
real world.'' ICCT stated that the effect of this change was to 
``render[] unavailable mass reduction technologies for these vehicles 
in the model,'' causing the model to select less cost-effective 
technologies instead and driving the modeled compliance costs higher.
    ICCT argued that to substantiate the changes made to the baseline 
fleet mass reduction assignments, the agencies must show data on how 
these improvements are evident in the fleet and to quantify and include 
their realized benefits in the analysis, including a detailed and 
justified explanation of all mass reduction technologies deemed already 
to have been applied to the MY 2016 analysis fleet. More specifically, 
ICCT stated that the agencies ``must clearly and precisely share their 
estimated percent (and absolute pounds) mass reduction amount for each 
vehicle make and model in the baseline fleet (rather than simply 
showing binned categories), and their technical justification for each 
value,'' and ``[t]o not do so obscures the agencies' new methods and 
data sources from public view, rendering their lightweighting 
calculations a black box.''
    In addition, ICCT recommended that the agencies conduct two 
sensitivity analyses, one assuming that every baseline make and model 
has not yet applied any lightweighting (setting the baseline to 0% mass 
reduction), and one assuming that each vehicle model has applied Draft 
TAR baseline mass reduction assignments, to demonstrate how much the 
agencies' decision to load up more baseline technology affects the 
compliance scenarios.
    ICCT concluded that because the changes in baseline mass reduction 
assignments from prior analyses to the NPRM ``are opaquely buried in 
the agencies' datafiles and unexplained, we believe the agencies have 
to reissue a new regulatory analysis and allow an additional comment 
period for review of their methods and analysis.''
    To address ICCT's comment, it is important to understand the mass 
reduction baseline technology assignment methodology previously used by 
EPA in the Draft TAR and Proposed Determination.\1343\ As stated in the 
Draft TAR, the curb weight of each vehicle model in the MY 2008 
analysis fleet (used for the 2012 rulemaking to establish MYs 2017-2025 
standards) was assumed to be at a baseline MR0 level. The mass 
reduction technology level in the MY 2014 analysis fleet was determined 
by comparing the curb weight of the MY 2014 vehicle to the most similar 
vehicle in the MY 2008 analysis fleet.\1344\ The curb weight of the 
newer model year vehicle was adjusted to account for changes in the 
vehicle footprint and changes in mass due to added safety technology. 
If a vehicle did not have a previous generation vehicle, then the sales 
weighted average percent mass reduction over the manufacturer's name 
plate product line was used to represent the expectation of mass 
reduction technology available within the vehicle.
---------------------------------------------------------------------------

    \1343\ Draft TAR at 5-395.
    \1344\ Draft TAR at 5-395.
---------------------------------------------------------------------------

    EPA listed some limitations to this methodology in the Draft 
TAR,\1345\ and others are also addressed here. First, assuming that 
every vehicle started with MR0 technology did not account for the 
actual varying levels of mass reduction technology that existed in the 
MY 2008 fleet. Second, for each vehicle model, there was no accounting 
for the mass associated with different powertrain configurations. This 
was particularly problematic because the method did not account for 
light weight technology already available in the vehicle structure to 
counter the increased mass associated with more advanced powertrains, 
such as HEV, PHEV, and EV technologies.\1346\ Third, there was no 
sales-weight accounting for the various configurations in estimating 
the vehicle model mass reduction technology level, meaning that if a 
high-sales-volume vehicle employed significant mass reduction 
technology, that vehicle was not credited as such in the analysis

[[Page 24533]]

fleet. Fourth, there was no accounting for mass increases due to the 
addition of future regulatory requirements like potential safety 
regulations. Fifth, there was no accounting for mass associated with 
changes in vehicle attributes and utility, such as the addition of 
infotainment systems and crash avoidance technologies. These 
limitations all individually had the effect of overestimating mass 
reduction technology effectiveness and undercounting mass reduction 
technology costs across the fleet, and accordingly their combined 
effect was significant. The lack of controls for these items introduced 
errors into the mass reduction technology level effectiveness 
estimates.
---------------------------------------------------------------------------

    \1345\ Draft TAR at 5-395.
    \1346\ PRIA at 413.
---------------------------------------------------------------------------

    After considering the comments, the agencies determined the use of 
the regression method, based on observable attributes, is the best 
available methodology to provide a reasonable estimate of mass 
reduction technology for the analysis fleet. The agencies believe that, 
contrary to ICCT's assertion, the regression methodology used in the 
NHTSA Draft TAR, NPRM, and final rule analyses provides a more 
transparent method for calculating baseline mass reduction technology 
assignments. The methodology was fully explained in the Draft TAR and 
PRIA, and avoided the limitations identified by EPA by using data from 
the analysis fleet, and not requiring the use of or assumptions about 
the exact mass reduction levels of vehicles in a prior model year 
fleet. In addition, the regression accounted for differences in 
powertrains between trim levels, including non-ICE powertrains by 
accounting for these factors in the regression analysis.
    Also, because manufacturers generally apply mass reduction 
technology at a vehicle platform level (i.e., using the same components 
across multiple vehicle models that share a common platform) to 
leverage economies of scale and to manage component and manufacturing 
complexity, conducting the regression analysis at the platform level 
leads to more accurate estimates for the real-world vehicle platform 
mass reduction levels. The platform approach also addresses the impact 
of potential weight variations that might exist for specific vehicle 
models, as all of the individual vehicle models are aggregated into the 
platform group, and are effectively averaged using sales weighting, 
which minimizes the impact of any outlier vehicle configurations.
    The agencies also disagree that the changes in baseline mass 
reduction assignments were unexplained. The PRIA discussed reasons that 
baseline mass reduction assignments differed from prior analyses, 
including that, ``[s]ince the Draft TAR, many platforms have not been 
redesigned, but in some cases the sales-weighted residuals for 
carryover platforms have moved. In the case of 2-Box and 3-Box 
vehicles, the analysis attributes such changes to differences in sales 
mix year-over-year and other updates to reported curb weights and 
platform designations. In the case of platforms with pick-up trucks, 
the analysis updated the pick-up regression since the Draft TAR, so 
that may be a contributing factor.'' \1347\
---------------------------------------------------------------------------

    \1347\ PRIA at 424.
---------------------------------------------------------------------------

    To the extent that the NPRM glider weight assumption impacted the 
NPRM MY 2016 analysis fleet baseline mass reduction assignment values, 
the agencies presented a table in the PRIA showing how different glider 
weight assumptions impacted mass reduction technology levels for the 
analysis fleet.\1348\ The following Table VI-123 recreates that table 
in part, with updates based on the glider weight values used for the 
final rule.
---------------------------------------------------------------------------

    \1348\ PRIA at 422.
---------------------------------------------------------------------------

    For example, from the regression analysis, the Ford F-150 has a 
predicted curb weight (residual) of 12.4 percent of the actual curb 
weight. If the glider weight assumption is 50 percent of the vehicle 
curb weight (like in NPRM), then the agencies would assign MR5 as an 
initial mass reduction assignment in the analysis fleet. With this high 
level of mass reduction technology already applied, the opportunity for 
further mass reduction would be limited. However, if the glider weight 
is assumed to be 71 percent of the vehicle curb weight, then Ford F-150 
would be assigned MR4, and would have an opportunity to apply another 
level of mass reduction albeit at higher cost.

[[Page 24534]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.257

    The agencies also disagree that the amount of vehicle mass 
reduction technology present in the 2016 baseline fleet was 
``unjustifiably increased'' from the 2015 baseline used in the Draft 
TAR. Table VI-124 shows the percent mass reduction technology used in 
Draft TAR, NPRM, and in final rule. It is clear from the table below 
that total percentage of MY 2016 vehicle fleet used in the NPRM had 
nearly the same level of some mass reduction technology applied 
compared to the Draft TAR. Similar to ICCT's observations, 28 percent 
of the MY 2015 vehicle fleet used in the Draft TAR had some level of 
mass reduction technology (MR1 to MR5) and 26 percent of MY 2016 
vehicle fleet had some mass reduction technology applied. Since the 
agencies assumed a reduced glider share in the NPRM, the percentage of 
vehicles assigned a MR4 or MR5 technology level increased compared to 
Draft TAR. In addition, for this final rule, the agencies observed that 
many of the vehicles in the MY 2017 fleet had been redesigned, which 
provided the opportunity to incorporate additional mass reduction 
technologies.
[GRAPHIC] [TIFF OMITTED] TR30AP20.258

    The agencies considered a sensitivity case that assumed no mass 
reduction, rolling resistance, or aerodynamic improvements had been 
made to the MY 2017 fleet (i.e., setting all vehicle road levels to 
zero--MRO, AERO and

[[Page 24535]]

ROLL0), in response to ICCT's comment. While this is an unrealistic 
characterization of the initial fleet, the agencies conducted a 
sensitivity analysis to understand any affect it may have on technology 
penetration along other paths (e.g., engine and hybrid technology). 
Under the CAFE program, the sensitivity analysis shows a slight 
decrease in reliance on engine technologies (HCR engines, turbocharge 
engines, and engines utilizing cylinder deactivation) and hybridization 
(strong hybrids and plug-in hybrids) in the baseline (relative to the 
central analysis). The consequence of this shift to reliance on lower-
level road load technologies is a reduction in compliance cost in the 
baseline of about $300 per vehicle (in MY 2026). As a result, cost 
savings in the preferred alternative are reduced by about $200 per 
vehicle. Under the CO2 program, the general trend in 
technology shift is less dramatic (though the change in BEVs is larger) 
than the CAFE results. The cost change is also comparable, but slightly 
smaller ($200 per vehicle in the baseline) than the CAFE program 
results. Cost savings under the preferred alternative are further 
reduced by about $100. With the lower technology costs in all cases, 
the consumer payback periods decreased as well. These results are 
consistent with the approach taken by manufacturers who have already 
deployed many of the low-level road load reduction opportunities to 
improve fuel economy.
    Second, as discussed above, EPA's Draft TAR baseline mass reduction 
assignments had identified limitations that the regression methodology 
has addressed. Moreover, as discussed above, the regression methodology 
was updated from the Draft TAR to characterize data better on pickup 
trucks. The agencies do not believe that conducting sensitivity 
analyses using these outdated or limited assumptions would be useful 
for this final rule.
    More narrowly, HDS commented that while the regression coefficients 
between 2-box and 3-box vehicles for footprint seemed consistent, the 
regression coefficients for horsepower between the 2-box and 3-box 
vehicles seemed incorrect because both types of vehicles use similar 
engines.\1349\ HDS stated that ``[c]ollinearity between footprint and 
HP or other effects caused by having electric vehicles (with electric 
motor HP ratings) in the regression data is the probable cause of these 
inconsistent coefficients for HP, but this cannot be confirmed without 
access to the same database used by NHTSA.'' HDS concluded that 
``[r]evisions to the regression could have a significant effect on the 
baseline assignment of vehicles, as the current assignment for vehicles 
like the 2016 Mazda MX5 as having the highest level of weight reduction 
technology (MR5) and the 2016 Chevy Malibu as having MR4 technology 
appear incorrect as their curb weights are comparable to other similar 
MY 2016 vehicles in their respective class.''
---------------------------------------------------------------------------

    \1349\ H-D Systems, NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    While many of the vehicles share same the same powertrain for 
passenger cars and SUVs or for cars and pickup trucks, the utility and 
functionality of the vehicle in SUVs and pickup trucks (2-box) is 
different than passenger cars (3-box). The presence of additional 
structure for towing or higher capacity towing, rear cross member, 
higher capacity suspension, and other differences, enable SUVs and 
pickup trucks to have towing and heavier payload capability. For 
example, Ford uses the nearly similar displacement and horsepower 
engines in Mustang Ecoboost Coupe and in F150 2WD XL, Regular Cab, Long 
Box. However, the curb weight for the pickup truck is higher than the 
Mustang. Directionally, this supports that the 2-box weight per 
horsepower coefficient should be greater than the 3-box coefficient, 
just as it is in the for the regression. The coefficient for passenger 
cars and SUVs has not changed since the Draft TAR (based on MY2015 
vehicle fleet). Based on the comments to Draft TAR, for the NPRM, a new 
set of coefficients were generated for pickups using the MY 2014 
vehicle fleet. This was done so that coefficients were not skewed due 
to presence of the aluminum intensive Ford F150 pickup truck. Hence, 
the agencies believe the coefficients used in the regression analysis 
are directionally correct and disagree with HDS's assertion. The 
agencies further note that HDS did not suggest any alternate 
methodology or specific coefficients to use in the regression analysis.
(c) Mass Reduction Technology Adoption Features
    The agencies described in the NPRM that given the degree of 
commonality among the vehicle models built on a single platform, 
manufacturers do not have complete freedom to apply unique technologies 
to each vehicle that shares the platform: while some technologies 
(e.g., low rolling resistance tires) are very nearly ``bolt-on'' 
technologies, others involve substantial changes to the structure and 
design of the vehicle, and therefore often necessarily affect all of 
the vehicle models that share that platform. In most cases, mass 
reduction technologies are applied to platform level components and 
therefore the same design and components are used on all of the vehicle 
models that share the platform.
    As discussed in Section Analysis Fleet, above, each vehicle in the 
analysis fleet is associated with a specific platform. Similar to the 
application of engine and transmission technologies, the CAFE model 
defines a platform ``leader'' as the vehicle variant of a given 
platform that has the highest level of observed mass reduction present 
in the analysis fleet. If there is a tie, the CAFE model begins mass 
reduction technology on the vehicle with the highest sales in model 
year 2017. If there remains a tie, the model begins by choosing the 
vehicle with the highest Manufacturer Suggested Retail Price (MSRP) in 
MY 2017. As the model applies technologies, it effectively levels up 
all variants on a platform to the highest level of mass reduction 
technology on the platform. So, if the platform leader is already at 
MR3 in MY 2017, and a ``follower'' starts at MR0 in MY 2017, the 
follower will get MR3 at its next redesign (unless the leader is 
redesigned again before that time, and further increases the mass 
reduction level associated with that platform, then the follower would 
receive the new mass reduction level).
    Important for analysis fleet mass reduction assignments (discussed 
above), and for understanding adoption features as well, is the 
agencies' handling of vehicles that traditionally operated on the same 
platform but had a mix of old and new platforms in production when the 
analysis fleet was created. As described in the PRIA, the Honda Civic 
and Honda CR-V traditionally share the same platform. In MY 2016, Honda 
redesigned the Civic and updated the platform to include many mass 
reduction technologies. Also in MY 2016, Honda continued to build the 
CR-V on the previous generation platform--a platform that did not 
include many of the mass reduction technologies on the all new MY 2016 
Civic. In MY 2017, Honda launched the new CR-V that incorporated 
changes to the Civic platform, and the Civic and CR-V again shared the 
same platform with common mass reduction technologies. The NPRM and 
final rule analyses treat the old and new platforms separately to 
assign technology levels in the baseline, and the CAFE model brings 
vehicles on the old platform up to the level of mass reduction 
technology on the new shared platform at the first available redesign 
year.
    Furthermore, as stated in the NPRM and PRIA, unlike the analysis 
presented in the Draft TAR that restricted high

[[Page 24536]]

levels of mass reduction for cars to show a safety neutral pathway to 
compliance, the NPRM analysis did not artificially restrict mass 
reduction to achieve a safety neutral outcome.\1350\ The NPRM CAFE 
model considered MR0 through MR5 for all vehicles at redesign, and 
similarly for the final rule, the CAFE model considers MR0 through MR6 
for all vehicles at redesign.
---------------------------------------------------------------------------

    \1350\ PRIA at 494.
---------------------------------------------------------------------------

    Ford commented in support of the removal of ``previously applied 
modeling rules that disallowed the mass reduction technology pathway 
for certain vehicle classes since this restriction was not supported by 
an adequate technical justification.'' \1351\ ICCT commented that a 
constraint of 7.5 percent mass reduction to smaller cars, as was 
applied in the Draft TAR, would be appropriate for those vehicles.
---------------------------------------------------------------------------

    \1351\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    The agencies considered ICCT's comment that mass reduction on small 
passenger cars should be limited to 7.5 percent, and Ford's comment 
supporting the removal of ``previously applied modeling rules that 
disallowed the mass reduction technology pathway for certain vehicle 
classes.'' Neither CAFE standards nor this analysis mandate mass 
reduction, or mandate that mass reduction occur in any specific manner. 
The mass reduction cost subsection below shows mass reduction is a 
cost-effective technology for improving fuel economy and CO2 
emissions. The steel, aluminum, plastics, composite, and other material 
industries are developing new materials and manufacturing equipment and 
facilities to produce those materials. In addition, suppliers and 
manufacturers are optimizing designs to maintain or improve functional 
performance with lower mass. Manufacturers have stated that they will 
continue to reduce vehicle mass to meet more stringent standards, and 
therefore, this expectation is incorporated into the modeling analysis 
supporting the standards to: (1) Determine capabilities of 
manufacturers; and (2) predict costs and fuel consumption effects of 
CAFE standards. The CAFE and CO2 rulemakings in 2012, and 
the Draft TAR and EPA Proposed Determination, imposed an artificial 
constraint that limited vehicle mass reduction in some small vehicles 
to achieve a desired safety-neutral outcome. For the current 
rulemaking, this artificial constraint is eliminated so the analysis 
reflects manufacturers' applying the most cost effective technologies 
to achieve compliance with the regulatory alternatives and the final 
standards; this approach allows mass reduction to be applied across the 
fleet. This approach is consistent with industry trends. To the extent 
that mass reduction is only cost-effective for the heaviest vehicles, 
the CAFE model would create the outcome predicted by commenters. In 
reality, however, mass reduction is a cost-effective means of improving 
fuel economy and does take place across vehicles of all sizes and 
weights. Accordingly, the model reflects that manufacturers may reduce 
vehicle mass--regardless of vehicle class--when doing so is cost 
effective.
    The agencies have included one additional mass reduction level for 
the final rule in response to comments by ICCT and others, and to 
account for carbon fiber use in vehicles. For the NPRM, the maximum 
level of mass reduction was limited to 10 percent of a vehicle's curb 
weight, and that amount of mass reduction could be applied during the 
rulemaking timeframe. For the final rule, based on the current state of 
mass reduction technology and the application rate of different levels 
of mass reduction technologies, the agencies applied phase-in caps for 
MR5 and MR6 (15 percent and 20 percent reduction of a vehicle's curb 
weight, respectively). The agencies applied a phase-in cap for MR5 
level technology so that 15 percent of the vehicle fleet starting in 
2016 employed the technology, and the technology could be applied to 
100 percent of the fleet by MY 2022. This cap is consistent with the 
NHTSA lightweighting study that found that a 15 percent curb weight 
reduction for the fleet is possible within the rulemaking 
timeframe.\1352\ The agencies also applied a phase in cap for MR6 
technology so that one percent of the vehicle fleet starting in MY2016 
employed the technology, and the technology could be applied to 13 
percent of the fleet by MY2025. The agencies believe that this phase-in 
cap appropriately functions as a proxy for the cost and complexity 
currently required (and that likely will continue to be required until 
manufacturing process evolve) to produce carbon fiber components. 
Again, MR6 technology in this analysis reflects the use of a 
significant share of carbon fiber content, as seen through the BMW i3 
and Alfa Romeo 4c as discussed above.
---------------------------------------------------------------------------

    \1352\ DOT HS 811 666: Mass Reduction for Light Duty Vehicles 
for Model Years 2017-2025: Figure 397 at page 356.
---------------------------------------------------------------------------

(d) Mass Reduction Technology Effectiveness
    As discussed in Section VI.B.3, Argonne developed a database of 
vehicle attributes and characteristics for each vehicle technology 
class that included over 100 different attributes like frontal area, 
drag coefficient, fuel tank weight, transmission housing weight, 
transmission clutch weight, hybrid vehicle component weights, and 
weights for components that comprise engines and electric machines, 
tire rolling resistance, transmission gear ratios, and final drive 
ratio. Argonne used these attributes to ``build'' each vehicle that it 
used for the effectiveness modeling and simulation. Important for 
precisely estimating the effectiveness of different levels of mass 
reduction is an accurate list of initial component weights that make up 
each vehicle subsystem, from which Autonomie considered potential mass 
reduction opportunities.
    As stated above, glider weight, or the vehicle curb weight minus 
the powertrain weight, is used to determine the potential opportunities 
for weight reduction irrespective of the type of powertrain.\1353\ This 
is because weight reduction can vary depending on the type of 
powertrain. For example, an 8-speed transmission may weigh more than a 
6-speed transmission, and a basic engine without variable valve timing 
may weigh more than an advanced engine with variable valve timing. 
Autonomie simulations account for the weight of the powertrain system 
inherently as part of the analysis, and the powertrain mass accounting 
is separate from the application and accounting for mass reduction 
technology levels (MR0-MR6) that are applied to the glider in the 
simulations. Similarly, Autonomie also accounts for battery and motor 
mass used in hybrid and electric vehicles separately. This secondary 
mass reduction is discussed further, below.
---------------------------------------------------------------------------

    \1353\ Depending on the powertrain combination, the total curb 
weight of the vehicle includes glider, engine, transmission and/or 
battery pack and motor(s).
---------------------------------------------------------------------------

    Accordingly, in the Autonomie simulation, mass reduction technology 
is simulated as a percentage of mass removed from the specific 
subsystems that make up the glider, as defined for that set of 
simulations (including the non-powertrain secondary mass systems such 
as the brake system).
(1) Glider Mass and Mass Reduction
    Autonomie accounts for the mass of each subsystem that comprises 
the glider. For the NPRM, the glider subsystems included the vehicle 
body and the chassis, but did not include mass from subsystems such as 
the interior system, brake system, electrical accessory system, and 
steering and

[[Page 24537]]

wheel systems. The agencies described in the PRIA that based on 
advances in active and passive safety technologies that add some mass 
to the interior system, certain subsystems were not considered for 
potential light-weighting to maintain safety performance.\1354\ For the 
NPRM, the A2Mac1 database was used to estimate the average mass of each 
subsystem considered as part of the glider based on the subsystem 
assumptions, and to compute the average glider share of vehicle curb 
weight.\1355\ That analysis showed the glider accounted for 50 percent 
of the vehicle curb weight. The agencies solicited comment on whether 
systems or components beyond the vehicle body and chassis should be 
included as part of the glider, and also indicated that the glider 
weight assumption might increase for the final rule based on further 
research.
---------------------------------------------------------------------------

    \1354\ PRIA at 411-12.
    \1355\ The A2Mac1 database was used and this analysis was 
presented in ANL report docketed here: NHTSA-2018-0067-1490. The 
mass data in the database were obtained from vehicle teardown 
studies.
---------------------------------------------------------------------------

    The agencies received several comments on the NPRM glider weight 
assumptions, with the overarching theme of the comments being that the 
NPRM did not include all systems and components that should be 
included, and if those systems and components were included, the glider 
share would be higher. Commenters also stated that the 50 percent 
glider share value used for the NPRM reduced the amount of mass 
reduction that could be applied to vehicles in the analysis.
    UCS stated that representing the glider as a reduced fraction of 
the curb weight caused the agencies significantly to underestimate the 
potential for mass reduction. UCS noted that because mass reduction is 
applied at the glider level, reducing the share of the glider 
inherently caps the potential reduction in the curb weight, and this 
single change cut the potential improvement from mass reduction by one-
third. Similarly, CARB stated that the updated glider weight assumption 
severely limited the effectiveness of mass reduction, as the most 
aggressive mass reduction category of 15 to 20 percent mass reduction 
can only reduce the vehicle curb weight by 10 percent.
    UCS cited previous agency analyses and analyses from other 
organizations that stated the total potential for mass reduction by 
2025 is between 15.8 and 32 percent of curb weight, contrasted to the 
NPRM assumption of a maximum 10 percent reduction.\1356\ UCS also cited 
industry data which showed that the glider represented a higher share 
of vehicle curb weight than was assumed in the Draft TAR analysis, and 
both UCS and CARB cited to industry data from vehicles like the Ford F-
150, which UCS stated was able to achieve the NPRM maximum achievable 
mass reduction through the deployment of aluminum alone.\1357\ UCS 
concluded that by capping the total potential for mass reduction at 
such a low level, the agencies artificially reduced the potential for 
the cost-effective technology, which increased the use of more 
expensive and more advanced technologies. CARB concluded that the 
agencies' 10 percent restriction means that real-world improvements 
that have already happened on production vehicles were not considered 
feasible in the NPRM analysis.
---------------------------------------------------------------------------

    \1356\ NHTSA-2018-0067-12039 (citing Caffrey et al. 2013, 
Caffrey et al. 2015, Lotus 2012, NAS 2015, Singh et al. 2012, Singh 
et al. 2016, Singh et al. 2018).
    \1357\ NHTSA-2018-0067-12039. See also NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    Several commenters also stated that the 50 percent glider weight 
assumption was unexplained and unjustified, and argued that the 
agencies' own studies showed that the glider weight percentage should 
range from 75-80 percent.\1358\ UCS stated that both the NHTSA-
sponsored 2011 Honda Accord study, which showed the glider making up 79 
percent of the vehicle, and the NHTSA-sponsored 2014 Chevrolet 
Silverado study, which showed the glider making up 73.6 percent, showed 
values substantially higher than the 50 percent value, and were in line 
with the agencies' prior analyses.\1359\ As part of its comments that 
key assumptions about mass reduction changed from the Draft TAR without 
any supporting rationale, CARB stated that EPA had previously relied on 
four studies (two contracted for by EPA and two contracted for by 
NHTSA), and for the NPRM analysis the agencies only cited two of those 
studies.\1360\ Moreover, ICCT commented that the agencies' previous 
studies showed a glider fraction greater than 75 percent even with 
numerous safety features considered. Accordingly, ICCT stated that the 
agencies must specifically identify the ``safety components'' referred 
to in the NPRM and justify the limitations placed on light weighting in 
response. ICCT affirmatively concluded that the agencies must re-adopt 
the Draft TAR methodology in which glider mass is assumed to be 75 
percent of vehicle mass, or provide detailed justification and evidence 
supporting the new value of 50 percent.\1361\
---------------------------------------------------------------------------

    \1358\ NHTSA-2018-0067-11985; NHTSA-2018-0067-12039; NHTSA-2018-
0067-11873.
    \1359\ NHTSA-2018-0067-12039.
    \1360\ NHTSA-2018-0067-11873.
    \1361\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    The agencies carefully considered these comments and reexamined 
available data and information. The NHTSA-sponsored passenger car light 
weighting study showed a glider mass of 79 percent, and the NHTSA-
sponsored light duty truck light weighting study showed a glider mass 
of 73.6 percent, and the 75 percent value used for the Draft TAR was a 
value between the values from these two studies. The agencies 
determined it would be more rigorous to consider data from a broader 
array of vehicles with various powertrain combinations and trim levels 
to assess the glider share for the final rule, considering that the 
vehicle fleet analyzed in this rule consists of over 2900 vehicle 
models.
    The agencies examined glider weight data available in the A2Mac1 
database.\1362\ The A2Mac1 database tool is widely used by industry and 
academia to determine the bill of materials and mass of each component 
in the vehicle system.\1363\ The A2Mac1 database has been used by the 
agencies to inform past CAFE and CO2 rulemakings. The 
agencies analyzed a total of 147 MY 2014 to 2016 vehicles, covering 35 
vehicle brands with different powertrain options representing a wide 
array of vehicle classes to determine the glider weight for the final 
rule analysis.\1364\
---------------------------------------------------------------------------

    \1362\ A2Mac1: Automotive Benchmarking. (n.d.). Retrieved from 
https://a2mac1.com.
    \1363\ Bill of material (BOM) is a list of the raw materials, 
sub-assemblies, parts and quantities needed to manufacture an end 
product.
    \1364\ The agencies presented this material for comments in the 
ANL report posted in the docket NHTSA-2018-0067-1490.
---------------------------------------------------------------------------

    The agencies also considered that the NHTSA passenger car and light 
truck light-weighting studies examined mass reduction in the body, 
chassis, interior, brakes, steering, electrical accessory, and wheels 
subsystems and had developed costs for light weighted components in 
those subsystems. As a result, the agencies determined it is 
appropriate to include all of those subsystems as available for mass 
reduction as part of the glider. Therefore, all of these systems were 
included for the analysis of glider weight using the A2Mac1 database. 
Table VI-125 shows the average mass for each subsystem and the glider 
share for each of the vehicle classes for all powertrain combinations.
BILLING CODE 4910-59-P

[[Page 24538]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.259

    This data was also compared with the glider weight measured in the 
NHTSA MY 2014 Chevrolet Silverado light weighting study,\1365\ and the 
glider weight data range was similar to the analysis results. Based on 
the comments and the agencies' updated assessment, the agencies have 
increased the glider weight assumption to 71 percent of the vehicle 
curb weight for the final rule.
---------------------------------------------------------------------------

    \1365\ DOT HS 812 487: Mass Reduction for Light-Duty Vehicles 
for Model Years 2017-2025.
---------------------------------------------------------------------------

    As stated above, for the NPRM, the interior, brake system, 
electrical accessory system, and steering and wheel systems were not 
included as part of the glider. The decision not to include the 
interior system was based on an assumption at that time that interior 
system mass reduction might adversely impact safety. In addition, the 
decision not to include the brake system was based on an assumption at 
that time that there would be little or no opportunity for downsizing 
and reducing mass based on the reduced weight from body and chassis 
only. As a result, brake systems were not considered as part of the 
glider in the NPRM. For the final rule, the agencies included the 
interior system based on market observations that light-weighted seats, 
side door trim, frontal dash, and others interior components have been 
incorporated on production vehicles that meet FMVSSs and perform well 
on voluntary NCAP and IIHS safety tests. The agencies also considered 
that interior, brakes, steering, wheel and electrical subsystems were 
included in the NHTSA light weighting studies. By adding the interior, 
steering, wheel subsystems and electrical subsystems as part of glider, 
the agencies believe light weighting the glider increases the 
opportunity for brake system optimization and mass reduction. 
Similarly, there is increased opportunity for mass reduction for wheels 
using gauge optimization, resulting from including more subsystems in 
the glider.
    By including the interior, brake, steering, electrical accessory, 
and wheel subsystems in addition to the body and chassis subsystems in 
the definition of what subsystems comprise the glider, the agencies 
increased the glider weight from 50 percent of the vehicle curb weight 
to 71 percent of the vehicle curb weight. This increase in turn means 
that the potential for vehicle mass reduction was increased from 10 
percent of the vehicle curb weight to 20 percent of the vehicle curb 
weight. Table VI-126 shows the percent of light truck glider weight 
reduction and the corresponding vehicle curb weight reduction for each 
level of mass reduction for the glider shares used in the Draft TAR (75 
percent), NPRM (50 percent), and final rule (71 percent) 
analyses.\1366\
---------------------------------------------------------------------------

    \1366\ Table 6-57 in PRIA showed the vehicle curb weight changes 
for different glider weight assumptions.
---------------------------------------------------------------------------

BILLING CODE 4910-59-C

[[Page 24539]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.260

2) Powertrain Mass Reduction
    As explained above, any mass reduction due to powertrain 
improvements is accounted for separately from glider mass reduction. 
Autonomie considers several components for powertrain mass reduction, 
including engine downsizing, and transmission, fuel tank, exhaust 
systems, and cooling system lightweighting.
    The 2015 NAS report suggested an engine downsizing opportunity 
exists when the glider mass is lightweighted by at least 10%. The 2015 
NAS report also suggested that 10% lightweighting of the glider mass 
alone would boost fuel economy by 3% and any engine downsizing 
following the 10% glider mass reduction would provide an additional 3% 
increase in fuel economy.\1367\ The agencies' lightweighting studies 
applied engine downsizing (for some vehicle types but not all) when the 
glider weight was reduced by 10 percent. Accordingly, the NPRM analysis 
limited engine resizing to several specific incremental technology 
steps; \1368\ important for this discussion, engines in the analysis 
were only resized when mass reduction of 10% or greater was applied to 
the glider mass, or when one powertrain architecture was replaced with 
another architecture.
---------------------------------------------------------------------------

    \1367\ National Research Council. 2015. Cost, Effectiveness, and 
Deployment of Fuel Economy Technologies for Light-Duty Vehicles. 
Washington, DC--The National Academies Press. https://doi.org/10.17226/21744.
    \1368\ 83 FR 43027.
---------------------------------------------------------------------------

    Argonne performed a regression analysis of engine peak power versus 
weight for the NPRM based on attribute data taken from the A2Mac1 
benchmarking database, to account for the difference in weight for 
different engine types. For example, to account for weight of different 
engine sizes like 4-cylinder versus 8-cylinder, Argonne developed a 
relationship curve between peak power and engine weight based on the 
A2Mac1 benchmarking data. For the NPRM analysis, this relationship was 
used to estimate mass for all engine types regardless of technology 
type (e.g., variable valve lift and direct injection). Weight 
associated with changes in engine technology was applied by using this 
linear relationship between engine power and engine weight from the 
A2Mac1 benchmarking database. When a vehicle in the analysis fleet with 
an 8-cylinder engine adopted a more fuel efficient 6-cylinder engine, 
the total vehicle weight would reflect the updated engine weight with 
two less cylinders based on the peak power versus engine weight 
relationship.
    When Autonomie selects a powertrain combination for a lightweighted 
glider, the engine and transmission are selected such that there is no 
degradation in the performance of the vehicle relative to the baseline 
vehicle. The resulting curb weight is a combination of the 
lightweighted glider with the resized and potentially new engine and 
transmission. This methodology also helps in accurately accounting for 
the cost of the glider and cost of the engine and transmission in the 
CAFE model. This is one of the fundamental differences between the 
analysis for this rulemaking the analysis for the Proposed 
Determination. For the Proposed Determination, the cost for mass 
reduction included mass reduction and cost reduction for one specific 
engine downsizing, and applied it to all vehicle classes without regard 
for performance and utility. There also was no accounting for the mass 
of other applied powertrains and the associated effectiveness impacts.
    As explained in the introduction, secondary mass reduction is 
possible from some of the components in the glider after mass reduction 
has been incorporated in primary subsystems (body, chassis, and 
interior). Similarly, engine downsizing and powertrain secondary mass 
reduction is possible after certain level of mass reduction is 
incorporated in the glider. For the analysis, the agencies include both 
primary mass reduction, and when there is sufficient primary mass 
reduction, additional secondary mass reduction. The Autonomie 
simulations account for the aggregate of both primary and secondary 
glider mass reduction, and separately for powertrain mass.
    The agencies received several comments about secondary mass 
reduction and powertrain mass reduction. Broadly, CARB commented that 
the agencies did not include powertrain downsizing and associated 
secondary mass reduction, which was a departure from the analysis done 
by

[[Page 24540]]

EPA for the Draft TAR.\1369\ CARB stated that the agencies 
``inexplicably'' did not consider secondary mass reduction 
opportunities ``including but not limited to drive axles, suspension, 
and braking components (as a result of the overall vehicle being 
lighter); fuel tank (and corresponding weight of fuel during 
certification testing); powertrain (lighter engine and transmission 
needed to power the lighter vehicle); and thermal systems.'' CARB cited 
both EPA and NHTSA light weighting studies for the proposition that 
there are significant opportunities for secondary mass reduction that 
lead to additional cost savings. As a result, CARB stated that the 
agencies inflated the cost of mass reduction as well as the amount of 
mass reduction that is feasible and cost-effective, leading to an over 
estimate in the technology costs to meet the existing standards.
---------------------------------------------------------------------------

    \1369\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    As CARB correctly noted, the NHTSA-sponsored studies have taken 
into consideration secondary mass reduction benefits such as radiator 
engine support, and optimized engine cradles, wheels, and suspension 
systems. As discussed above, in response to comments, the agencies have 
included additional subsystems such as brakes, wheels, steering, 
electrical, and interior systems to the glider for the final rule 
analysis, thereby accounting for mass reduction opportunities for these 
systems.
    Also, as discussed further in Section VI.C.4.e), below, secondary 
mass reduction is integrated into the mass reduction cost curves. 
Specifically, the NHTSA studies, upon which the cost curves were built, 
first generated costs for lightweighting the vehicle body, chassis, 
interior, and other primary components, and then calculated costs for 
lightweighting secondary components. Accordingly, the cost curves 
reflect that, for example, secondary mass reduction for the brake 
system is only applied after there has been sufficient primary mass 
reduction to allow the smaller brake system to provide safe braking 
performance and to maintain mechanical functionality.
    CARB appears to have misunderstood how the analysis accounts for 
powertrain mass reduction. The agencies described in the PRIA that the 
Autonomie simulations recognize that many powertrain packages have 
different weights for each vehicle class; for example, an eight-speed 
transmission may weigh more than a six-speed transmission, and a basic 
engine with variable valve timing may weigh more than a basic engine 
without variable valve timing.\1370\ Autonomie varies the weight of 
these powertrain systems as part of the analysis, and these changes are 
done separately from the glider mass reduction technology levels (MR0 
to MR6) in the simulations. Accordingly, accounting for powertrain mass 
reduction as part of the mass reduction technology analysis would 
double count impacts. The use of separate accounting assures that the 
analysis accounts for mass associated with secondary mass reduction 
from glider, and engine downsizing, as well as mass associated with 
each individual engine, transmission, and electrification technology. 
These mass changes were not accounted for in the Draft TAR and Proposed 
Determination analyses. Moreover, these are accounted for separately in 
the cost accounting, which is discussed further in the Section 
VI.C.4.e), below.
---------------------------------------------------------------------------

    \1370\ PRIA at 418.
---------------------------------------------------------------------------

    HDS commented that some assumptions in the Autonomie modeling 
related to engine weight appeared incorrect, such as the assumption 
that a turbocharged 4-cylinder engine weighed the same as a DOHV V6 
engine with 1.5 times the 4-cylinder's displacement, when in fact that 
engine is often 75 to 100 lbs. lighter.\1371\
---------------------------------------------------------------------------

    \1371\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    HDS also noted that ``mass reduction assumes no reduction of 
powertrain weight for mass reduction levels of 2.5% and 5%. Mass 
reduction effectiveness therefore are somewhat more appropriate for 
reductions over 5% which apparently include some powertrain weight 
reduction. More transparency in the PRIA regarding powertrain weight 
changes will allow more detailed comment on engine weight assumptions 
used.''
    We agree with the comment that certain advanced engines could be 
lighter than a basic engine. For the final rule, the estimated mass 
levels for engines were updated, as discussed in Section VI.B.3 Tech 
Effectiveness, based on the A2Mac1 database and other sources that 
provided more precise mass data for powertrain technologies. Also, the 
agencies improved upon the precision of estimated engine weights by 
creating two curves to represent separately naturally aspirated engine 
designs and turbocharged engine designs.\1372\ This update resulted in 
two benefits. First, small naturally aspirated 4-cylinder engines that 
adopted turbocharging technology reflected the increased weight of 
associated components like ducting, clamps, the turbocharger itself, a 
charged air cooler, wiring, fasteners, and a modified exhaust manifold. 
Second, larger cylinder count engines like naturally aspirated 8-
cylinder and 6-cylinder engines that adopted turbocharging and 
downsized technologies would have lower weight due to having fewer 
engine cylinders. For the final rule analysis, a naturally aspirated 8-
cylinder engine that adopts turbocharging technology and is downsized 
to a 6-cylinder turbocharged engine appropriately reflects the added 
weight of the turbocharging components, and the lower weight of fewer 
cylinders. These refinements address the issues identified in HDS's 
comments.
---------------------------------------------------------------------------

    \1372\ ANL Final Model Documentation for final rule analysis 
Chapter 5.2.9 Engine Weight Determination.
---------------------------------------------------------------------------

    Regarding HDS's second comment, as discussed in the NPRM, to 
address product complexity and economies of scale, engine resizing is 
limited to specific incremental technology changes that would typically 
be associated with a major vehicle or engine redesign.\1373\ As 
discussed further in Section VI.B.3.a)(6) Performance Neutrality, the 
NPRM also referred to the 2015 NAS report conclusion that ``[f]or small 
(under 5 percent [of curb weight]) changes in mass, resizing the engine 
may not be justified, but as the reduction in mass increases (greater 
than 10 percent [of curb weight]), it becomes more important for 
certain vehicles to resize the engine and seek secondary mass reduction 
opportunities.'' \1374\ In consideration of both the NAS report and 
comments received from manufacturers, the agencies determined it would 
be reasonable to allow allows engine resizing upon adoption of 7.1%, 
10.7%, 14.2%, and 20% curb weight reduction, but not at 3.6% and 
5.3%.\1375\ Resizing is also allowed upon changes in powertrain type or 
the inheritance of a powertrain from another vehicle in the same 
platform. The increments of these higher levels of mass reduction, or 
complete powertrain changes, more appropriately match the typical 
engine displacement increments that are available in a manufacturer's 
engine portfolio.
---------------------------------------------------------------------------

    \1373\ See 83 FR 43027 (Aug. 24, 2018).
    \1374\ National Research Council. 2011. Assessment of Fuel 
Economy Technologies for Light-Duty Vehicles. Washington, DC--The 
National Academies Press. http://nap.edu/12924.
    \1375\ These curb weight reductions equate to the following 
levels of mass reduction as defined in the analysis: MR3, MR4, MR5 
and MR6, but not MR1 and MR2; additional discussion of engine 
resizing for mass reduction can be found in Section VI.B.3 
Technology Effectiveness.

---------------------------------------------------------------------------

[[Page 24541]]

3) Summary of Final Rule Mass Reduction Technology Effectiveness
    Figure VI-45 below shows the range of incremental effectiveness 
used for the NPRM analysis. The chart lumps all of the vehicle classes 
for each of the technology types.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.261

BILLING CODE 4910-59-C
    Figure VI-46 below shows the range of incremental effectiveness 
improvement from full vehicle modeling when mass reduction technologies 
were applied to vehicles for the final rule analysis.
BILLING CODE 4910-59-P

[[Page 24542]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.262

BILLING CODE 4910-59-C
e) Mass Reduction Costs
    The PRIA described the decision to use NHTSA's passenger car light 
weighting study based on a MY 2011 Honda Accord and NHTSA's full-size 
pickup truck light weighting study based on a MY 2014 Chevrolet 
Silverado to derive the estimated cost for each of the mass reduction 
technology levels.\1376\ The agencies relied on the results of those 
studies because they considered an extensive range of material types, 
material gauge, and component redesign while taking into account real 
world constraints such as manufacturing and assembly methods and 
complexity, platform-sharing, and maintaining vehicle utility, 
functionality and attributes, including safety, performance, payload 
capacity, towing capacity, handling, NVH, and other characteristics. In 
addition, the agencies described that the baseline vehicles assessed in 
the NHTSA-sponsored studies were reasonably representative of baseline 
vehicles in the MY 2016 analysis fleet.\1377\ The agencies also noted 
they made the decision to rely on these studies after reviewing other 
agency, CARB, ICCT and industry studies.\1378\ The other studies often 
did not consider important factors, made unrealistic assumptions about 
key vehicle systems, and/or applied secondary mass reduction 
inappropriately, resulting in unrealistically low costs. The PRIA also 
described how the cost estimates derived from the NHTSA lightweighting 
studies were adjusted to reflect the NPRM glider share 
assumption.\1379\
---------------------------------------------------------------------------

    \1376\ PRIA at 391; Table 6-38 and Table 6-41 in PRIA.
    \1377\ PRIA at 403.
    \1378\ As described in the PRIA at 390-91, studies by EPA, CARB, 
Transport Canada, the American Iron and Steel Institute (AISI), the 
Aluminum Association, and the American Chemistry Council were all 
reviewed for potential incorporation into the analysis.
    \1379\ See PRIA at 396, Tables 6-38 and 6-39; PRIA at 401, 
Tables 6-41 and 6-42. See also PRIA at 391 (``While the definitions 
of glider may vary from study to study (or even simulation to 
simulation), the agencies referenced the same dollar per pound of 
curb weight to develop costs for different glider definitions. In 
translating these values, the agencies took care to track units ($/
kg vs. $/lb.) and the reference for percentage improvements (glider 
vs. curb weight).'').
---------------------------------------------------------------------------

    Furthermore, the agencies changed the cost of mass reduction 
accounting from a curb weight basis in the Draft TAR to glider weight 
basis in the NPRM.\1380\ Because the mass reduction studies provide 
mass reduction costs for the glider, this change enabled more direct 
use of cost curve data from the studies in the CAFE model. This change 
also allowed independent accounting for powertrain mass, which enabled 
the CAFE model to account more accurately for the unique mass of each 
of the powertrains that are available in each vehicle model. The cost 
of the engine, transmission, and electrification are accounted for 
separately from the glider in the CAFE model.
---------------------------------------------------------------------------

    \1380\ In the Draft TAR, the agencies presented the cost 
estimates from mass reduction studies sponsored by both NHTSA and 
EPA. EPA presented the cost of mass reduction as function of vehicle 
curb weight. To harmonize the cost estimates with EPA, NHTSA also 
presented the cost of mass reduction as a function of vehicle curb 
weight.
---------------------------------------------------------------------------

    The agencies received several comments on the mass reduction costs 
used in the NPRM. FCA commented that the costs and benefits used the 
CAFE model were overly optimistic,

[[Page 24543]]

stating that although its Ram 1500 pickup truck achieved several 
hundred pounds of weight reduction, the cost of achieving that weight 
reduction was greater than that used in the CAFE model.\1381\ 
Similarly, as mentioned above, Toyota commented that mass reduction 
cost values were underestimated.\1382\ Conversely, CARB, UCS, and the 
City of Oakland in California commented that the costs used for mass 
reduction in the NPRM overstated the cost of mass reduction. The 
agencies also received several comments relating to the studies used to 
develop the mass reduction cost curves, how the values from those 
curves were applied in the CAFE model, and costs for secondary mass 
reduction; those comments are discussed in turn.
---------------------------------------------------------------------------

    \1381\ NHTSA-2018-0067-11943.
    \1382\ NHTSA-2018-0067-12098.
---------------------------------------------------------------------------

(1) Studies Used To Develop Mass Reduction Cost Curves
    The agencies described in the PRIA that since the 2012 final rule, 
both agencies conducted lightweighting studies to assess the technical 
feasibility and cost of mass reduction.\1383\ The agencies also stayed 
apprised of studies performed by other agencies, manufacturers, and 
industry trade associations, and reviewed them in development of 
lightweighting assumptions used in the NPRM and final rule 
analysis.\1384\ Among the several lightweighting studies, the agencies 
used NHTSA's passenger car lightweighting study, based on a MY 2011 
Honda Accord, and NHTSA's full-size pickup truck lightweighting study, 
based on a MY 2014 Chevrolet Silverado, to derive the cost estimates to 
achieve different levels of mass reduction for the NPRM and final rule.
---------------------------------------------------------------------------

    \1383\ PRIA at 390.
    \1384\ PRIA at 403.
---------------------------------------------------------------------------

    The agencies described that the decision to rely on those studies 
included that those studies considered materials, manufacturing, 
platform-sharing, functional attribute, performance, and noise-
vibration- and harshness (NVH), among other constraints pertaining to 
cost, effectiveness, and safety considerations, in addition to that 
these vehicles were a reasonable representation of the baseline 
vehicles in the MY 2016 compliance simulation.\1385\ Specifically in 
regards to safety, the agencies described a preference to use studies 
that considered small overlap impact tests conducted by the Insurance 
Institute for Highway Safety (IIHS) and not all studies took that test 
into account. In regards to maintaining vehicle functionality, the 
agencies described that the NHTSA pickup truck study accounted for 
vehicle functional performance for attributes including towing, noise 
and vibration, and gradeability, in addition to considering platform 
sharing constraints.
---------------------------------------------------------------------------

    \1385\ PRIA at 403.
---------------------------------------------------------------------------

    In contrast, the agencies explained that the other studies often 
did not consider many important factors, or those studies made 
unrealistic assumptions about key vehicle systems through secondary 
downsizing, resulting in unrealistically low costs. Specifically, the 
agencies referenced EPA's past analysis of a MY 2010 Toyota Venza as an 
example of a study that employed overly aggressive secondary mass 
reduction, which translated into cost savings for the initial 10% mass 
reduction.\1386\
---------------------------------------------------------------------------

    \1386\ PRIA at 391.
---------------------------------------------------------------------------

    The agencies received several comments on the studies used to 
generate the mass reduction cost curves. Ford commented in support of 
the agencies' decision to exclude mass reduction studies that were 
misaligned with tear-down studies.\1387\ Ford cited the MY 2010 Toyota 
Venza Phase II study used to establish the mass reduction cost values 
used for the Draft TAR and Proposed Determination that suggested the 
first 7-10% of mass reduction could be accomplished with zero or 
reduced cost,\1388\ which Ford characterized as ``a gross 
underestimation of industry investment and material costs associated 
with any weight reduction.''
---------------------------------------------------------------------------

    \1387\ NHTSA-2018-0067-11928.
    \1388\ EPA-420-R-16-021: Proposed Determination Technical 
Support Document at 2-158, November 2016.
---------------------------------------------------------------------------

    ICCT commented that The National Academies of Science 
``specifically endorsed tear-down studies as the most appropriate way 
to get at vehicle technology costs, [as those] studies are typically 
more accurate and far more transparent than the older method of 
surveying manufacturers, and such whole-vehicle studies are key to 
capturing holistic vehicle level mass-reduction technology costs.'' 
ICCT noted that there are many peer-reviewed tear-down studies that 
demonstrate that at least 20 percent mass reduction is available for 
adoption across vehicle classes by 2025, including studies by EDAG, 
FEV, Ford, and Lotus Engineering; however, ICCT alleged that the 
agencies ``have either incorrectly interpreted or invalidly nullified 
the most relevant detailed engineering teardown studies on mass-
reduction technology.'' ICCT noted that the agencies were ``well 
aware'' of these studies, as they were performed by CARB in conjunction 
with the agencies, however, ICCT alleged that the agencies 
``reinterpreted the results of the main study relied upon in the TAR in 
order to inflate costs,'' and that the ``technical assessment by the 
agencies has a clear technical bias towards reducing CAFE and GHG 
standards.'' ICCT concluded that ``[e]xcluding these studies amounted 
to intentionally disregarding the most pertinent and rigorous 
engineering studies that are applicable to the rulemaking timeframe.''
    ICCT recommended the agencies adjust their technology cost inputs 
to reflect the ``best-available technology studies.'' ICCT stated that 
the correct cost assumption from these studies is that ``a 5-10% mass 
reduction by 2025 reduces vehicle cost, and the auto industry will 
cost-effectively deploy at least 15% vehicle curb mass reduction in the 
2025 timeframe at near zero net cost (and consistently less than 
$500).''
    CARB asserted that the agencies inflated the costs of mass 
reduction in the NPRM analysis by only considering NHTSA-sponsored 
studies and improperly excluding the effects of secondary mass 
reduction as documented in those studies.\1389\ CARB provided a table 
of studies that largely mirrored the tables of studies the agencies 
considered in the Draft TAR and PRIA,\1390\ and also included the 
associated mass reduction costs in $/kg included in each study, noting 
that for all excluded studies cited in the table, all mass reduction 
costs were substantially lower than the values used in the agencies' 
analysis.\1391\ Similarly, UCS commented that while the PRIA did state 
that additional studies ``often did not consider many important factors 
or . . . made unrealistic assumptions about key vehicle systems,'' the 
agencies did not specifically identify the factors and assumptions that 
merited disregarding those studies, which were included previously in 
agency analysis as part of the record when deriving previous estimates 
for the costs of mass reduction.\1392\
---------------------------------------------------------------------------

    \1389\ NHTSA-2018-0067-11873.
    \1390\ Draft TAR at 5-168; PRIA at 404-05.
    \1391\ NHTSA-2018-0067-11873.
    \1392\ NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    The agencies agree with ICCT that peer-reviewed tear-down studies 
present an appropriate method to capture holistic vehicle-level mass 
reduction technology costs. The agencies also agree with ICCT that the 
agencies were well aware of studies conducted by EDAG, FEV, Ford, and 
Lotus Engineering; in fact, the agencies

[[Page 24544]]

presented a table listing several of those studies in the PRIA with the 
qualification that those studies were reviewed in developing 
lightweight assumptions for the analysis, but those studies did not 
consider important factors, or those studies made unrealistic 
assumptions about key vehicle systems through secondary downsizing, 
resulting in unrealistically low costs.
    The agencies also agree with UCS' comment that the language could 
have been more specific about identifying the factors and assumptions 
that merited disregarding studies that were previously included as part 
of the record when deriving previous estimates for the costs of mass 
reduction. The following discussion briefly summarizes the record since 
the Draft TAR and differences between NHTSA's and other lightweighting 
studies' approach to factors listed in the PRIA. Important for this 
discussion is an understanding of primary versus secondary mass 
reduction; as described above, when there is sufficient primary mass 
reduction, other components that are designed based on the mass of 
primary components may be redesigned and have lower mass. Recall the 
braking system example used throughout this section; mass reduction in 
the braking system is secondary mass reduction because it requires 
primary mass reduction before it can be incorporated. If the mass of 
primary components is reduced sufficiently, the resulting lighter 
weight vehicle could maintain braking performance, attributes, and 
safety with a lighter weight brake system.
    Several studies were referenced in the Draft TAR that either used 
tear-down analyses and computer-aided engineering (CAE) to produce a 
future engineered lightweight vehicle, or considered future 
technologies and processes for lightweighting vehicle components.\1393\
---------------------------------------------------------------------------

    \1393\ Draft TAR at 5-158 through 5-197.
---------------------------------------------------------------------------

    EPA developed cost curves for cars and CUVs based on the MY 2010 
Toyota Venza study, and pickup truck cost curves based on the MY 2011 
Chevrolet Silverado study.\1394\ The other studies were considered by 
EPA, but not used to develop the Draft TAR, Proposed Determination and 
Final Determination cost curves. In brief, EPA described that the 
Toyota Venza Phase I was a mass reduction opportunity study only, and 
the Phase II study was a holistic vehicle study that examined nearly 
every component in the vehicle for mass reduction potential and 
calculated a related cost and mass saved for each. For the cost curve, 
EPA applied the individual components in sequence from largest cost per 
kilogram savings to largest cost per kilogram increase. For example, 
the cost curve for the Draft TAR and Proposed Determination applied 
engine downsizing and transmission system mass reduction first, and 
before lightweighting the body, chassis, doors and other 
components.\1395\ EPA stated this methodology was chosen based on the 
understanding that OEMs will choose the cost saving technologies first 
and that some cost mass reduction technologies will be paid for by the 
cost save mass reduction technologies, citing a 2016 publication by CAR 
and a GM presentation that stated over $2,000,000,000 was saved in 
material costs through various lightweighting approaches.\1396\
---------------------------------------------------------------------------

    \1394\ Draft TAR at 5-367.
    \1395\ EPA-420-R-16-021: Proposed Determination Technical 
Support Document at 2-161 and 2-162
    \1396\ Draft TAR at 5-172 (citing ``Identifying Real world 
Barriers to Implementing Lightweighting Technologies and Challenges 
in Estimating the Increase in Costs,'' Center for Automotive 
Research, Jay Baron, Ph.D., January 2016 http://www.cargroup.org/?module=Publications&event=View&pubID=128; General Motors, ``General 
Motors 2015 Global Business Conference,'' Presentation, October 1, 
2015, Slides 43-45 in document, https://www.gm.com/content/dam/gm/events/docs/5194074-596155-ChartSet-10-1-2015.).
---------------------------------------------------------------------------

    The NHTSA cost curves were developed by rearranging the 
lightweighted components from the MY 2011 Honda Accord and MY 2014 
Chevrolet Silverado studies based on cost effectiveness, assuming the 
vehicle body, chassis, interior, and other primary components were 
lightweighted first, followed then by lightweighting powertrain 
components and other secondary systems.\1397\ The cost curves based on 
the NHTSA studies reflect that, returning to this example, secondary 
mass reduction for the brake system is only applied after there has 
been sufficient primary mass reduction to allow the smaller brake 
system to provide safe braking performance and to maintain mechanical 
functionality.\1398\
---------------------------------------------------------------------------

    \1397\ Draft TAR at 5-421 (``The powertrain components which 
include engine, transmission, and fuel systems such as fuel filler 
pipe, fuel tank, fuel pump, etc., exhaust systems and cooling 
systems were not considered for application of primary mass 
reduction but benefits of secondary mass reduction were accounted 
for. These powertrain components are assumed to be downsized only 
after the primary vehicle structural components (Body-In-White) 
achieve certain level of mass reduction.'').
    \1398\ Draft TAR at 5-422.
---------------------------------------------------------------------------

    The EPA and NHTSA studies took fundamentally different approaches 
to accounting for the costs of mass reduction technology, and 
accordingly, EPA needed to translate the cost curves from the NHTSA 
studies to use a similar methodology as the cost curves from the EPA 
studies.\1399\ To ``normalize'' the NHTSA studies with the EPA's 
studies, EPA listed components identified for lightweighting in the 
NHTSA studies and reorganized those components from the lowest cost to 
highest cost along with associated mass reduction per the ``whole 
vehicle'' approach mentioned above, distributed mass savings from 
secondary mass reduction to all points along the cost curve, and 
included the mass saved from engine downsizing without taking into 
consideration the cost of added engine technology. This resulted in 
lower-cost secondary mass reduction opportunities being considered 
before primary mass reduction opportunities, which in turn resulted in 
artificially low $/kg costs for mass reduction.
---------------------------------------------------------------------------

    \1399\ Draft TAR at 5-369.
---------------------------------------------------------------------------

    For the NPRM and final rule, the agencies simply used the original 
ordered list of components from the MY 2011 Honda Accord study and MY 
2014 Chevrolet Silverado study, arranged sequentially for cost 
effectiveness based on primary then secondary mass reduction 
opportunities, to generate the cost curves for passenger cars and light 
trucks. Accordingly, the agencies did not ``reinterpret'' the results 
of studies used in the Draft TAR in the NPRM, as ICCT alleged, but 
rather appropriately represented how primary and secondary mass 
reduction opportunities are implemented in the real world (to the 
extent that ICCT is referring to the translation of the study costs to 
the NPRM glider weight assumptions, that is discussed in Section 
VI.C.4.e)(1), below). To maintain utility and performance in the real 
world, primary components must be lightweighted first before the engine 
and transmission can be resized. Moreover, as described in the Draft 
TAR, NHTSA's mass reduction studies did not ``improperly exclude'' the 
effects of secondary mass reduction, rather those effects were 
appropriately accounted for after primary components achieved certain 
levels of mass reduction. As discussed in Section VI.B.3.a)(6) 
Performance Neutrality, this approach aligned with the NAS approach to 
consider powertrain downsizing only after the vehicle structural 
components achieved 10 percent mass reduction.
    OEMs have also disagreed with the conclusion that mass reduction 
could come at a cost savings. For instance, Ford characterized the 
Toyota Venza studies, which concluded the first 7-10% of mass reduction 
could come at a negative cost as ``a gross

[[Page 24545]]

underestimation of industry investment and material costs associated 
with any weight reduction.'' The agencies believe that the approach to 
secondary mass reduction followed in the NHTSA passenger car and pickup 
truck lightweighting studies appropriately incorporated both the costs 
and real-world constraints associated with employing primary and 
secondary mass reduction technologies.
    Aside from the differences in how studies treated secondary mass 
reduction, the agencies opted not to use, or could not use, other 
studies either previously considered in the rulemaking record or 
mentioned by commenters for several reasons:
    Studies were not comprehensive, and therefore could not be used to 
develop a comprehensive cost curve: Some studies narrowly assessed 
lightweighting of a portion of vehicle, such as the body in white 
subsystem, or as stated in the PRIA,\1400\ were limited to material 
substitution of the vehicle components, such as replacing steel with 
aluminum or replacing mild steel with AHSS or replacing mild steel with 
CFRP in selective components. Factors important to vehicle 
functionality, like material joining techniques and the feasibility of 
manufacturing processes or necessary retooling requirements were not 
considered, and therefore could not be used to develop a comprehensive 
cost curve representative of the costs required to reduce mass in a 
vehicle.\1401\
---------------------------------------------------------------------------

    \1400\ PRIA at 391.
    \1401\ An Assessment of Mass Reduction Opportunities for a 2017-
2020 Model Year Vehicle Program, March 2010, Lotus Engineering, at 
p. 6.
---------------------------------------------------------------------------

    Cost curves were not developed or no cost analysis was performed: 
For the CARB Holistic Vehicle Mass Reduction/Cost Study, a cost curve 
was not developed, and the resulting cost per kilogram data points were 
point estimates. The calculated cost per kilogram was used as one data 
point of several to indicate the direction for mass reduction beyond 
EPA's original passenger car/CUV curve.\1402\ Or, as in the case of the 
DOE/Ford/Magna Multi Material Lightweight Vehicle (MMLV) project, no 
cost analysis was performed for the initial project, and later 
project(s) concluded that ``a 37% to 45% mass reduction in a standard 
mid-sized vehicle is within reach if carbon fiber composite materials 
and manufacturing processes are available and if customers are willing 
to accept a reduction in vehicle features and content, as demonstrated 
with the Multi-Materials and Carbon Fiber Composite-Intensive vehicle 
scenarios.'' \1403\
---------------------------------------------------------------------------

    \1402\ Draft TAR at 5-185.
    \1403\ Draft TAR at 5-194.
---------------------------------------------------------------------------

    Engineered vehicles did not meet functional design or manufacturing 
requirements: As noted in the update to EPA's Light-Duty Vehicle Mass 
Reduction and Cost Analysis for the Toyota Venza, the Phase I 
engineered Venza did not meet the design target of no expected NVH 
degradation.\1404\ The Phase II (High Development) study assumed 
significant cost savings from reduced parts manufacturing, but did not 
appropriately explain the methodology used to conclude that the part 
count reduction was feasible.\1405\
---------------------------------------------------------------------------

    \1404\ Light-Duty Vehicle Mass Reduction and Cost Analysis--
Midsize Crossover Utility Vehicle, EPA-420-R-12-026 (August 2012).
    \1405\ Peer Review of Demonstrating the Safety and 
Crashworthiness of a 2020 Model-Year, Mass-Reduced Crossover Vehicle 
(Lotus Phase 2 Report), EPA-420-R-12-028 (September 2012).
---------------------------------------------------------------------------

    In addition, the agencies qualified in the PRIA a preference to use 
studies that considered the small overlap impact test conducted by 
IIHS, and not all studies took that test into account.\1406\ NHTSA's 
``Update to Future Midsize Lightweight Vehicle Findings in Response to 
Manufacturer Review and IIHS Small-Overlap Testing'' based on the MY 
2011 Honda Accord presented results incorporating suggestions from 
Honda regarding NVH and durability, and updating the engineered vehicle 
to achieve a ``good'' rating in seven crash safety tests.\1407\ EPA 
studies also accounted for the IIHS small overlap test through an ad 
hoc estimate of mass and cost, unlike the NHTSA update, which 
explicitly modeled to account for NVH performance and to comply with 
the IIHS small overlap test.
---------------------------------------------------------------------------

    \1406\ PRIA at 391.
    \1407\ Singh, H., Kan, C-D., Marzougui, D., & Quong, S. (2016, 
February). Update to future midsize lightweight vehicle findings in 
response to manufacturer review and IIHS small-overlap testing 
(Report No. DOT HS 812 237). Washington, DC: National Highway 
Traffic Safety Administration.
---------------------------------------------------------------------------

    The agencies continue to believe that the MY 2011 Honda Accord and 
MY 2014 Chevrolet Silverado lightweighting studies are the best studies 
upon which to estimate the costs of mass reduction in the rulemaking 
timeframe.
(2) How the Cost Curves Are Applied in the Model
    Commenters also submitted comments on how the cost curves were 
applied in the model, including that the studies the agencies relied 
upon to generate cost curves, discussed above, did not support the 50 
percent glider share assumption used in the NPRM, and the agencies did 
not correctly scale the costs to match the glider share assumption.
    UCS commented that the agencies based the costs for mass reduction 
on glider weight reduction, however, the need for more expensive 
materials and more advanced engineering and design strategies only 
results from the need for greater levels of absolute mass reduction on 
the vehicle.\1408\ UCS stated that the cost curves had effectively been 
derived from the assumption of reductions as great as 16.8 percent 
reduction in curb weight in the case of the Silverado (Singh et al. 
2018) and as great as 18 percent reduction in curb weight in the case 
of the Honda Accord (Singh et al. 2016), but applied to curb weight 
reductions approximately two-thirds that magnitude. UCS stated that 
approach was ``completely invalid and significantly overstates the 
costs for mass reduction.'' UCS also commented that the agencies 
incorrectly scaled the cost curves based on the agencies' mass 
reduction studies, which refer to direct manufacturing costs as a 
function of vehicle curb weight, not just glider weight. UCS stated 
this incorrectly yielded the same costs for two-thirds the amount of 
mass reduction.
---------------------------------------------------------------------------

    \1408\ NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    CARB similarly commented that the mass reduction costs assigned to 
both passenger cars and light trucks in the CAFE model were 
inappropriately inflated based on incorrect scaling from the glider 
share assumptions used in the Honda Accord and Chevy Silverado studies 
to the NPRM glider share value.\1409\ CARB analyzed two tables in the 
PRIA that showed the agencies' translation of cost numbers derived from 
the two studies to the cost numbers used in the CAFE model, and 
asserted that the agencies improperly used costs from the upper end of 
the mass reduction range rather than the midpoint of the range, leading 
to cost overestimation.
---------------------------------------------------------------------------

    \1409\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    Similarly, HDS commented that the PRIA passenger car cost curves 
used data that were not in agreement with the study that they were 
based upon, noting that the Honda Accord study showed the glider 
accounting for 78% of curb weight, and this limited absolute weight 
reduction.\1410\ HDS noted that the truck weight reduction cost data 
were closer to those cited in the Chevy Silverado teardown study, 
although the glider share for that study was also 73.6% of vehicle curb 
weight.
---------------------------------------------------------------------------

    \1410\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    HDS also commented that although the agencies relied on the same 
Honda Accord study that was used in the Draft

[[Page 24546]]

TAR, ``the costs have been changed significantly [from the Draft TAR] 
for unexplained reasons.'' \1411\ HDS stated that the PRIA showed 
average costs for mass reduction, whereas earlier studies showed the 
cost increment for each 5% mass reduction, noting that with increasing 
incremental cost with increased mass reduction, average cost will 
always be lower than incremental cost. HDS claimed that it was 
``unusual'' for the Draft TAR incremental costs to decrease between 11% 
and 19% mass reduction but increase elsewhere, but also noted the 
unexplained increase in cost, specifically a $536 cost for 175kg weight 
reduction, shown in the PRIA.
---------------------------------------------------------------------------

    \1411\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    HDS also compared manufacturing costs from the Draft TAR to the 
PRIA analysis, noting that the direct manufacturing cost was found to 
be negative (i.e., a cost saving) in the Draft TAR analysis for mass 
reduction up to 15 percent, but EPA assumed the indirect costs were 
positive so that the total cost was a sum of positive and negative 
costs--meaning the total cost could be positive or negative. In 
contrast, HDS noted there were no negative costs in the cost curves 
used for the PRIA analysis, resulting in a very large differential 
between the costs of mass reduction, with the 2018 average cost being 
higher than even the 2016 marginal costs.
    Three notable changes from the NHTSA Draft TAR to NPRM and final 
rule analysis impacted how the cost curves for mass reduction are 
applied in the CAFE Model.
    First, the Draft TAR considered mass reduction in the glider and 
powertrain together, and calculated the percentage mass reduction on a 
vehicle curb weight basis. In the Draft TAR, only one engine and 
transmission combination were considered to account for the mass change 
associated with downsizing the engine, and the cost estimates for mass 
reduction for this one powertrain combination was applied to all 
powertrain combinations. This approach did not account for the mass 
changes associated with the application of powertrain technologies 
(engine, transmission and electrification) technologies, and did not 
account for the corresponding change in glider mass needed to offset 
the powertrain mass change and to achieve the specified curb weight 
mass reduction level. This approach did not reflect the real world, 
where there are many vehicles with different body styles and powertrain 
combinations, and therefore did not account for differences in mass for 
different engines, transmissions, or electrification.
    Accordingly, for the NPRM and final rule, the cost of mass 
reduction was calculated on a glider weight basis so that the weight of 
each powertrain configuration could be directly and separately 
accounted for. This approach provides the true cost of mass reduction 
without conflating the mass change and costs associated with downsizing 
a powertrain or adding additional advanced powertrain technologies. 
Hence, the mass reduction costs in the NPRM reflect the cost of mass 
reduction in the glider and do not include the mass reduction 
associated with engine downsizing, and therefore appear to be higher 
than the cost estimates in the Draft TAR.
    Second, the glider share of curb weight changes from the Draft TAR 
to NPRM and from the NPRM to the final rule analysis also affected the 
absolute amount of curb weight reduction that was applied, and 
therefore for cost per pound for the mass reduction changes with the 
change in the glider share. The cost for removing 20 percent of the 
glider weight when the glider represents 75% of a vehicle's curb weight 
is not the same as the cost for removing 20 percent of the glider 
weight when the glider represents 50% of the vehicle's curb weight. For 
example, the glider share of 79 percent of a 3,000-pound curb weight 
vehicle is 2,370 pounds, while the glider share of 50 percent of a 
3,000-pound curb weight vehicle is 1,500 pounds, and the glider share 
of 71 percent of a 3,000-pound curb weight vehicle is 2,130 pounds. The 
mass change associated with 20 percent mass reduction is 474 pounds for 
79 percent glider share (=[3,000 pounds x 79% x 20%]), 300 pounds for 
50 percent glider share (=[3,000 pounds x 50% x 20%]), and 426 pounds 
for 71 percent glider share (=[3,000 pounds x 71% x 20%]). The mass 
reduction cost studies show that the cost for mass reduction varies 
with the amount of mass reduction. Therefore, for a fixed glider mass 
reduction percentage, different glider share assumptions will have 
different costs.
    To further illustrate, Table VI-127 and Table VI-128 below shows 
the associated curb weight percentage mass reduction and the associated 
average cost per pound for different glider weight assumptions for each 
glider mass reduction technology level used in the final rule analysis. 
For reference, the costs from the passenger car light weighting study 
are presented.\1412\ These costs were the basis for deriving the costs 
for each mass reduction technology level in the Draft TAR, NPRM, and 
final rule analyses, using the unique glider share values for each of 
those analyses. In the light weighting study, NHTSA applied the mass 
reduction technologies identified for the exemplar vehicle on other 
vehicle(s) and vehicle types to understand the level of mass reduction 
that could be achieved. In the case of passenger cars, the maximum 
level of mass reduction was around 15% of the vehicle curb weight if 
all the mass reduction technologies are applied. In other words, 
achieving mass reduction greater than 10% of the curb weight for 
passenger cars will require extensive use of advanced materials such as 
high strength aluminum and carbon fiber composite material.
---------------------------------------------------------------------------

    \1412\ Table 6-39 in PRIA.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24547]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.263

[GRAPHIC] [TIFF OMITTED] TR30AP20.264

BILLING CODE 4910-59-C
    Finally, as explained earlier, to determine the mass reduction 
technology levels for the NPRM 2016 analysis fleet, a distribution of 
the residuals from the regression using 50 percent glider weight 
generally showed a greater percentage of vehicles achieving higher 
levels of mass reduction. With this high level of mass reduction 
already achieved, the opportunities for further mass reduction would be 
limited and have higher costs. For the final rule, since the agencies 
updated the glider share to 71 percent of the vehicle curb weight, the 
distribution of residuals from the regression shifted some vehicles to 
lower baseline mass reduction

[[Page 24548]]

technology levels, providing more opportunity for further mass 
reduction, on average. Even as some of the vehicles start further up on 
the mass reduction cost curve due to higher levels of mass reduction 
technology (MR3, MR4) already present in the vehicles, there are 
additional opportunities for further mass reduction to achieve MR5 and 
above.
    Table VI-127 and Table VI-128 show that for the final rule, cost 
estimates with the 71 percent glider share come closer to the cost 
estimates used in Draft TAR, which assumed a 79 percent glider share.
(3) Secondary Mass Reduction Costs
    As discussed above, the agencies changed the cost of mass reduction 
calculation from a curb weight basis in the Draft TAR to a glider 
weight basis in the NPRM.\1413\ This change allowed us to estimate the 
cost of mass reduction independently of the cost associated with 
downsized advanced engines and advanced transmissions, as the cost of 
downsized advanced engines and transmissions are accounted for 
separately in the CAFE model.
---------------------------------------------------------------------------

    \1413\ In the Draft TAR, the agencies presented the cost 
estimates from mass reduction studies sponsored by both NHTSA and 
EPA. EPA presented the cost of mass reduction as function of vehicle 
curb weight. To harmonize the cost estimates with EPA, NHTSA also 
presented the cost of mass reduction as a function of vehicle curb 
weight.
---------------------------------------------------------------------------

    The MY 2011 Honda Accord and MY 2014 Chevy Silverado studies used 
to develop the NPRM and final rule cost curves for mass reduction 
technologies include some non-powertrain secondary mass reduction 
technologies such as brakes and wheels. The agencies presented the list 
of mass reduction technologies in NPRM.\1414\ Following the publication 
of NHTSA's light weighting studies, peer reviewers and manufacturers 
commented that many components such as drive axles, engine cradles, and 
radiator engine support that are considered to be non-powertrain 
secondary mass reduction opportunities cannot be downsized, as the same 
components are used across many vehicles with different powertrain 
options. Even though some of these components may provide opportunities 
for additional mass reduction, NHTSA agreed with peer reviewers and 
manufacturers that retaining a common design for all powertrain options 
provides for cost reductions due to economies of scale.
---------------------------------------------------------------------------

    \1414\ Table 6-37 and Table 6-40 in PRIA.
---------------------------------------------------------------------------

    Commenters faulted the agencies for a perceived lack of accounting 
for the cost decreases from secondary mass reduction. ICCT commented 
although the agencies relied on the Honda Accord study, which 
considered cost savings from downsizing the powertrain, in the NPRM 
only glider weight reduction was ever considered without the cost-
offsetting engine downsizing.\1415\ ICCT stated that this omission had 
two effects, first that accounting for associated powertrain weight 
reductions would have allowed for more mass reduction, thus allowing 
for greater efficiency benefits at a lower cost, and second, that 
vehicle performance was erroneously improved, contrary to the agencies' 
assertion that the analysis assumed a level of performance neutrality. 
ICCT concluded that it was unclear if and how costs were reduced for 
powertrain downsizing, as well as the precise changes to fuel 
efficiency.
---------------------------------------------------------------------------

    \1415\ NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    CARB faulted the agencies for not including secondary mass 
reduction in the NPRM analysis, and stated that by failing to account 
for secondary mass reduction as was done in the Draft TAR, the agencies 
inflated the costs for mass reduction as well as the amount of mass 
reduction that is feasible and cost-effective leading to an 
overestimate in the technology costs needed to meet the existing 
standards.
    The agencies note that the cost curves used for the NPRM and this 
final rule do in fact include secondary mass reduction. The cost curves 
reflect secondary mass reduction applied when there is sufficient 
primary mass reduction to implement secondary mass reduction without 
degrading function and safety. Specifically, the NHTSA studies, upon 
which the cost curves were built, first generated costs for 
lightweighting the vehicle body, chassis, interior, and other primary 
components, and then calculated costs for lightweighting secondary 
components. Accordingly, the cost curves reflect that, for example, 
secondary mass reduction for the brake system is only applied after 
there has been sufficient primary mass reduction to allow the smaller 
brake system to provide safe braking performance and to maintain 
mechanical functionality.
    In addition, CARB stated that the 2011 Honda Accord and the 2014 
Chevrolet Silverado studies had ``markedly'' lower costs than the 
proposal when secondary mass reduction is included. Again, the agencies 
believe these comments resulted from a lack of understanding about how 
the analysis considers primary and secondary mass reduction, even 
though the NPRM and PRIA explicitly stated how costs are accounted for 
separately.\1416\ Also, as discussed above, engine mass reduction 
enabled by mass reduction in the glider is accounted for separately and 
therefore not included as part of glider mass reduction technology, as 
doing so would result in double counting the impacts.
---------------------------------------------------------------------------

    \1416\ PRIA at 413.
---------------------------------------------------------------------------

(4) Summary of Final Rule Mass Reduction Costs
    For the final rule, the agencies continue to use multiple mass 
reduction technology levels and costs based on the lightweighting 
studies that were presented in PRIA.\1417\ Since the agencies have 
changed the glider share of curb weight assumption from 50 percent in 
NPRM to 71 percent in the final rule, the mass reduction costs reflect 
the updated glider share. Table VI-129 and Table VI-130 show mass 
reduction costs used in the CAFE model for passenger car and light 
trucks.
---------------------------------------------------------------------------

    \1417\ Table 6-37 and 6-40 in PRIA.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24549]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.265

[GRAPHIC] [TIFF OMITTED] TR30AP20.266

BILLING CODE 4910-59-C
5. Aerodynamics
    The energy required to overcome aerodynamic drag accounts for a 
significant portion of the energy consumed by a vehicle, and can become 
the dominant factor for a vehicle's energy consumption at high speeds. 
Reducing aerodynamic drag can, therefore, be an effective way to reduce 
fuel consumption and emissions.
    Aerodynamic drag is proportional to the frontal area (A) of the 
vehicle and coefficient of drag (Cd), such that aerodynamic 
performance is often expressed as the product of the two values, 
CdA, which is also known as the drag area of a vehicle. The 
coefficient of drag (Cd) is a dimensionless value that 
essentially represents the aerodynamic efficiency of the vehicle shape. 
The

[[Page 24550]]

frontal area (A) is the cross-sectional area of the vehicle as viewed 
from the front. It acts with the coefficient of drag as a sort of 
scaling factor, representing the relative size of the vehicle shape 
that the coefficient of drag describes. The force imposed by 
aerodynamic drag increases with the square of vehicle velocity, 
accounting for the largest contribution to road loads' higher speeds.
    Aerodynamic drag reduction can be achieved via two approaches, 
either by reducing the drag coefficient or reducing vehicle frontal 
area, with two different categories of technologies, passive and active 
aerodynamic technologies. Passive aerodynamics refers to aerodynamic 
attributes that are inherent to the shape and size of the vehicle, 
including any components of a fixed nature. Active aerodynamics refers 
to technologies that variably deploy in response to driving conditions. 
These include technologies such as active grille shutters, active air 
dams, and active ride height adjustment. It is important to note that 
manufacturers may employ both passive and active aerodynamic 
technologies to achieve aerodynamic drag values.
    The greatest opportunity for improving aerodynamic performance is 
during a vehicle redesign cycle when significant changes to the shape 
and size of the vehicle can be made. Incremental improvements may also 
be achieved during mid-cycle vehicle refresh using restyled exterior 
components and add-on devices. Some examples of potential technologies 
applied during mid-cycle refresh are restyled front and rear fascia, 
modified front air dams and rear valances, addition of rear deck lips 
and underbody panels, and low-drag exterior mirrors. While 
manufacturers may nudge the frontal area of the vehicle during 
redesigns, large changes in frontal area are typically not possible 
without impacting the utility and interior space of the vehicle. 
Similarly, manufacturers may improve Cd by changing the 
frontal shape of the vehicle or lowering the height of the vehicle, 
among other approaches, but the form drag of certain body styles and 
airflow needs for engine cooling often limit how much Cd may 
be improved.
    During the vehicle development process, manufacturers use various 
tools such as Computational Fluid Dynamics (CFD), scaled clay models, 
and full size physical prototypes for wind tunnel testing and 
measurements to determine aerodynamic drag values and to evaluate 
alternate vehicle designs to improve those values.
    The agencies presented a table in the PRIA showing aerodynamic drag 
improvements from individual technologies based on wind-tunnel testing 
for a study commissioned by Transport Canada, which is reproduced in 
Table VI-131 below.\1418\ The individual technologies are present in 
many of the 2016 and 2017 vehicles in the fleet. Table VI-131 shows the 
list of aerodynamic technologies and corresponding aero drag 
improvements.
---------------------------------------------------------------------------

    \1418\ Table 6-63 in PRIA.
    [GRAPHIC] [TIFF OMITTED] TR30AP20.267
    
    As discussed in the PRIA and further below, the agencies made 
several notable changes for modeling aerodynamic improvement 
technologies from the Draft TAR to the NPRM. First, the agencies 
revised the aerodynamic

[[Page 24551]]

improvements from two levels in the Draft TAR (10% and 20% improvement 
over the baseline) to four levels (5%, 10%, 15% and 20% aerodynamic 
drag improvement values over the baseline). This change provided the 
improved granularity to bin the vehicles with different aerodynamic 
improvements more appropriately. Next, the agencies assigned levels of 
aerodynamic technology to the MY 2016 fleet on a relative basis based 
on confidential business information submitted by the manufacturers, 
taking steps to verify information submitted by manufactures with other 
sources, and making changes particularly for vehicles that showed large 
improvements over baseline values. Third, the agencies limited the 
maximum level of aerodynamic improvements that certain body styles 
(pickup trucks, minivans) could achieve and limited the maximum level 
of improvements that cars and SUVs with more than 405 horsepower could 
achieve, based on the agencies' assessment of industry comments. 
Finally, the agencies updated the cost for aerodynamic improvements 
based on the assessment of comments that the National Academy of 
Sciences (NAS) cost estimates used in the Draft TAR underestimated the 
cost for aerodynamic improvements.
    Broadly, Ford commented in support of the approach to aerodynamic 
improvement modeling in the NPRM, stating that the rule recognized 
potential constraints like consumer needs and preferences regarding 
vehicle styling, vehicle utility, and interior space, by among other 
things, recognizing that the potential for aerodynamic drag differs 
among different vehicle body styles and vehicle classes.\1419\ Ford 
stated that these are major factors considered by customers when 
comparing competing vehicles, and the failure of a manufacturer to 
deliver in these areas can lead to the production of non-competitive, 
poor-selling vehicles.
---------------------------------------------------------------------------

    \1419\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    On the other hand, ICCT claimed that the agencies greatly limited 
the availability of many load reduction technologies (i.e., mass 
reduction improvements, aerodynamic improvements, and rolling 
resistance improvements) by pushing very large amounts of these 
technologies into the 2016 model year baseline fleet, thereby making 
the technologies unavailable for use in future years.\1420\ ICCT 
commented that these improvements in the analysis fleet would 
ostensibly amount to massive efficiency improvements, however, these 
assumed changes were not substantiated as resulting in any test-cycle 
efficiency improvements in the model year 2016 fleet versus the 2015 
fleet. ICCT concluded that the adjusted baseline had been developed and 
presented opaquely, apparently based primarily upon estimations from 
automaker-supplied data, without critical analysis, vetting, or sharing 
of the necessary data to substantiate the changes and real-world 
benefits by the agencies.
---------------------------------------------------------------------------

    \1420\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    As discussed further in Section VI.C.5.b) AERO drag analysis fleet 
assignments below, the agencies believe the updated analysis fleet 
aerodynamic technology level assignments in the NPRM analysis represent 
an improvement over the MY 2015 assignments in the Draft TAR, as the 
updated assignments are based on precise values, not estimated from 
road load coefficients, and have been corroborated by observed 
improvements on actual production vehicles. Accordingly, the agencies 
carried over the NPRM approach for determining the aerodynamic 
technology levels for the analysis fleet to the final rule.
a) Aerodynamics Drag Reduction Modeling in the CAFE Model
    The agencies summarized in the PRIA that the Draft TAR aerodynamic 
improvement levels were binned into two groups, AERO1 and AERO2. 
However, market observations showed that many vehicles had aero 
improvements from 0% to 10%, and some vehicles showed improvements from 
10% to 20%.\1421\ Based on industry feedback and market observations, 
the agencies revised the aerodynamic improvements from two levels in 
the Draft TAR (10% and 20% improvement over the baseline) to four 
levels (5%, 10%, 15% and 20% aerodynamic drag improvement values over 
the baseline). This revision provided the necessary granularity to bin 
the vehicles with different aerodynamic improvements appropriately.
---------------------------------------------------------------------------

    \1421\ PRIA at 437.
---------------------------------------------------------------------------

    ICCT commented that to model appropriately the baseline standards, 
the agencies would need to include increasing use of aerodynamic off-
cycle technology credits across all companies through 2025. ICCT stated 
that it appeared that the agencies did not use EPA's engineering 
expertise or compliance data, where EPA would be able to advise better 
based on their certification data from the off-cycle program.
    As discussed further in Sections VI.A and VI.C.8, the NPRM analysis 
carried forward manufacturers' off-cycle fuel consumption improvement 
values (FCIVs) at MY 2016 levels unless an explicitly simulated off-
cycle technology, like start-stop systems, was added to a vehicle in 
the simulation modeling. Specific to aerodynamic improvements, active 
grille shutters were assumed to be applied at the 20 percent 
aerodynamic improvement (AERO20) level. For the final rule analysis, 
based on the assessment of comments that the application of off-cycle 
technologies in the analysis was too conservative, the agencies agreed 
and increased each manufacturers' application of off-cycle technologies 
so that 10 g/mi of technology was applied by 2023, using an 
extrapolated increase in levels in MYs 2017-2023 based on EPA 
compliance data.\1422\ This approach did not assume any specific mix of 
off-cycle technologies that would be used by manufacturers to achieve 
the 10 g/mi off-cycle improvement, because manufactures currently use a 
variety of technologies, and different manufacturers likely would 
implement unique combinations of technologies. It is expected that 
aerodynamic off-cycle technologies would be included in the mix of off-
cycle technologies.
---------------------------------------------------------------------------

    \1422\ The 2018 EPA Automotive Trends Report, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends.
---------------------------------------------------------------------------

    Table VI-132 and Table VI-133 show aerodynamic technologies that 
could be used to achieve 5%, 10%, 15% and 20% aero improvements in 
passenger cars, SUVs, and pickup trucks.\1423\ The agencies developed 
these potential combinations of technologies using aerodynamic data 
from a National Research Council (NRC) of Canada sponsored wind tunnel 
testing program that included an extensive review of production 
vehicles utilizing these technologies, and industry 
comments.1424 1425 These technology combinations are 
intended to show a potential way for a manufacturer to achieve each 
aerodynamic improvement level; however, in the real world, 
manufacturers may implement different combinations of aerodynamic 
technologies to achieve a percentage

[[Page 24552]]

improvement over their baseline vehicles.
---------------------------------------------------------------------------

    \1423\ Table 6-67 and Table 6-68 in PRIA.
    \1424\ Larose, G., Belluz, L., Whittal, I., Belzile, M. et al., 
``Evaluation of the Aerodynamics of Drag Reduction Technologies for 
Light-duty Vehicles--a Comprehensive Wind Tunnel Study,'' SAE Int. 
J. Passeng. Cars--Mech. Syst. 9(2):772-784, 2016, https://doi.org/10.4271/2016-01-1613.
    \1425\ Larose, Guy & Belluz, Leanna & Whittal, Ian & Belzile, 
Marc & Klomp, Ryan & Schmitt, Andreas. (2016). Evaluation of the 
Aerodynamics of Drag Reduction Technologies for Light-duty 
Vehicles--a Comprehensive Wind Tunnel Study. SAE International 
Journal of Passenger Cars--Mechanical Systems. 9. 10.4271/2016-01-
1613.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.268


[[Page 24553]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.269

BILLING CODE 4910-59-C
b) Aerodynamic Drag Reduction Analysis Fleet Assignments
    The agencies described in the PRIA that for the 2015 analysis fleet 
used in the Draft TAR, the agencies received Cd values for 
the MY 2015 vehicles' baseline assignments from manufacturers, or used 
estimated Cd values. In response, the industry commented 
that Cd values often varied by measurement approach and, 
therefore, it was important to account for differences in the 
methodologies used to estimate those values. For instance, aerodynamic 
drag coefficients for the same vehicle often vary significantly from 
wind-tunnel to wind-tunnel, complicating cross-comparison and cross-
referencing.\1426\ The industry commented that, on average, the 
manufacturer-reported Cd values are nine percent lower than 
the values reported by USCAR.\1427\ For reference, USCAR follows the 
SAE J2881 test procedure. However, because Cd values are not 
required to be reported for compliance, manufacturers can and do choose 
different methods to estimate the Cd values. Therefore, the 
industry commented that assigning baseline aerodynamic improvement 
levels should not simply be comparing the lowest reported Cd 
value in a vehicle segment to other reported Cd values. The 
industry commented that such a comparison would not reflect the 
plausible amount of aerodynamic drag improvement that could be 
achieved. Accordingly, the industry suggested that the analysis should 
normalize manufacturer-reported Cd values using SAE J2881.
---------------------------------------------------------------------------

    \1426\ PRIA at 435.
    \1427\ Footnote in PRIA at 435: FCA Draft TAR comments. Docket 
ID: NHTSA-2016-0068-0082.
---------------------------------------------------------------------------

    The commenters stated manufacturers have the option to use other 
methods (apart from coast down testing) to estimate the Cd 
values such as wind tunnel testing, cross referencing the Cd 
value from other vehicles with similar frontal design and aero 
technologies deployed. Since manufacturers do not have to specify the 
methodology used to estimate the Cd value, the agencies have 
limited capability to make accurate comparisons of the Cd 
value estimates from different testing methods. As a result, the 
agencies determined using average(s) of the fleet provide a better 
estimate of Cd levels than using the lowest Cd 
value in the fleet to assign aerodynamic improvement levels. The 
agencies determined it is appropriate to continue to use the NPRM 
approach for the final rule.
    The NPRM and final rule analysis used a relative performance 
approach to assign the current aerodynamic technology level to a 
vehicle. Different body styles offer different utility and have varying 
levels of baseline form drag. In addition, frontal area is a major 
factor in aerodynamic forces, and the frontal area varies by vehicle. 
This analysis considered both frontal area and body style as utility 
factors affecting aerodynamic forces; therefore, the analysis assumed 
all reduction in aerodynamic drag forces come from improvement in the 
Cd. Per the process outlined in NHTSA's section of the Draft 
TAR,\1428\ the agencies computed an average Cd for each body 
style segment in the MY 2015 analysis fleet from drag coefficients 
published by manufacturers. By comparing the Cd among 
vehicles sharing body styles, this allowed the agencies to estimate the 
level of aerodynamic improvement present on specific vehicles.
---------------------------------------------------------------------------

    \1428\ Draft TAR at 4-80.
---------------------------------------------------------------------------

    While some small differences existed between the aggregate MY 2015 
and MY 2016 data, the agencies retained the NHTSA-calculated MY 2015 
average Cd as the baseline drag coefficient for nearly all 
body styles. For pickup trucks, the agencies assigned a baseline drag 
coefficient of 0.42, considering that a large portion of the pickups 
sold in MY 2015 already included aerodynamic features assumed for 
advanced levels of aero. The agencies harmonized the Autonomie 
simulation baselines with

[[Page 24554]]

the analysis fleet assignment baselines to the fullest extent 
possible.\1429\
---------------------------------------------------------------------------

    \1429\ Often, vehicles assigned to technology classes do not 
perfectly match up with simulated vehicles, but in most cases this 
analysis assumed the aerodynamic effects and other specifications 
were comparable and appropriate for use as proxies.
---------------------------------------------------------------------------

    The agencies assigned levels of aerodynamic technology to the MY 
2016 fleet based on confidential business information submitted by 
manufacturers on aerodynamic drag coefficients, and from other 
information sources such as in product release information. The 
analysis referenced manufacturer-submitted data (if that data was 
supplied), and the agencies took industry comments to Draft TAR into 
account and closely reviewed the manufacturer-submitted Cd 
data. In the few cases that manufacturers did not submit Cd 
values as confidential business information, the agencies estimated the 
Cd based vehicle attributes, design, and aero technologies 
applied to that vehicle. The agencies noted that the Cd 
values reported by some manufacturers showed high levels of improvement 
relative to the previous model year or previous generation. In some 
cases, the agencies contacted the manufacturers to further discuss 
differences in Cd estimation methodologies. Where 
appropriate, the agencies adjusted MY 2016 fleet Cd values 
after consultation with the manufacturers and used these values to 
assign baseline technology levels for each vehicle in the NPRM CAFE 
model simulation.
    The Alliance commented that the NPRM analysis fleet had more 
appropriately assigned aerodynamic technology levels, and the 
assignments were more accurate than the Draft TAR, where vehicles were 
generally considered to have little aerodynamic improvement technology, 
and the CAFE model would add aerodynamic improvement despite the fact 
that manufacturers had already made significant improvements and there 
was little opportunity remaining for more.\1430\ The Alliance concluded 
that the Draft TAR approach ultimately led the CAFE model to under-
predict how much powertrain technology was required for compliance. The 
Alliance also commented that it is possible to estimate aerodynamic 
features of a vehicle using road load coefficients, but the process 
requires various assumptions and is not very accurate. The Alliance 
concluded that the agencies' use of CBI to assign initial aerodynamic 
improvement values is an accurate and practical solution to support 
correct baseline assignments.
---------------------------------------------------------------------------

    \1430\ NHTSA-2018-0067-12039 at 136.
---------------------------------------------------------------------------

    Ford commented that the use of actual data, like manufacturer 
confidential information or other sources, to characterize better the 
aerodynamic improvements already incorporated into the baseline fleet 
is a substantial improvement over previous analyses that either assumed 
no aero improvement due to insufficient data, or attempted to infer 
Cd from the road load coefficients.\1431\ Ford stated that 
attempting to infer Cd from road load coefficients is not 
sufficiently accurate for a vehicle-level determination since the 
aerodynamic component of the road load coefficients is inextricably 
confounded with tire, transmission, and other parasitic losses. As part 
of its comments that the proposed rule analysis recognized constraints 
like consumer needs and preferences regarding vehicle styling and 
utility, Ford stated that the baseline Cd for pickup trucks 
properly recognized that these vehicles already include many advanced-
level aerodynamic technologies. Ford concluded that an accurate 
assessment of the current technological state of the baseline fleet is 
critical to ensuring that the benefits of technological improvements 
are not ``double-counted'' in the modeling.
---------------------------------------------------------------------------

    \1431\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    On the other hand, ICCT commented that the agencies artificially 
limited the availability of aerodynamic technologies in the CAFE model 
in future years by assigning approximately three times as many 
aerodynamic technology packages in the 2016 analysis fleet as they did 
in the 2015 baseline fleet used in the Draft TAR.\1432\ ICCT noted that 
the 2015 Draft TAR fleet had about 8 percent vehicles with one of the 
aerodynamic packages, whereas the NPRM's 2016 fleet had about 53 
percent, and argued that the agencies did not justify the increase with 
data to show that automakers actually deployed the technology. ICCT 
pointed to the agencies' introduction of intermediate aerodynamic 
improvement steps as the justification for the change, which ICCT 
argued ``redistributes the baseline fleet into more advanced 
aerodynamic levels without observing or verifying real-world 
aerodynamic improvements.''
---------------------------------------------------------------------------

    \1432\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    ICCT argued that if an improvement of this magnitude were true, it 
would be evident in fleet level miles-per-gallon and CO2 
levels (e.g., in EPA's Trends and Manufacturer Performance reports), 
but none of the quantifiable mpg or CO2 benefits that would 
be associated with these additional aerodynamic improvements were 
reflected in any real-world evidence in the model year 2016 fleet. ICCT 
stated that to show the automakers deployed this level of aerodynamic 
improvements, the agencies needed to show data on how these 
improvements are evident in the fleet and delivering benefits. 
Specifically, ICCT stated that the agencies must share the basis for 
any aerodynamic calculation and exact estimated percent improvement 
(rather than binned percentage categories) for each vehicle make and 
model in the baseline and future modeled fleet, and their technical 
justification for each value, arguing that not doing so would obscure 
the agencies' methods. In addition, ICCT stated that the agencies must 
conduct two sensitivity analysis cases that assume that every baseline 
make and model is set to 0 percent aerodynamic improvement and set to 
the previous baseline aerodynamic levels (i.e., from TAR) to 
demonstrate how much the agencies' decision to load up more baseline 
technology affects the compliance scenarios. ICCT concluded that 
because changes in aerodynamic improvement assumptions ``are opaquely 
buried in the agencies' datafiles and unexplained,'' the agencies must 
issue a new regulatory analysis and allow an additional comment period 
for review of the methods and analysis.
    ACEEE asserted, as part of its comments that the MY 2016 analysis 
fleet assignments appeared to contain errors, that the assignment of 
AERO10 for the MY 2016 Toyota Tundra pickup truck was in error.\1433\ 
ACEEE stated that Tundra pickup trucks have had similar specs from MY 
2011 to today, and the Cd for all Tundra models has been 
0.37 or 0.38 for 2WD and 4WD, respectively, since MY 2011. ACEEE noted 
that this is higher than the AERO10 Cd cut off value of 
0.355 for pickups, as shown in the 2016 Draft TAR and referenced in the 
PRIA.
---------------------------------------------------------------------------

    \1433\ NHTSA-2018-0067-12122, at 6.
---------------------------------------------------------------------------

    As described above, the agencies assigned levels of aerodynamic 
technology to the NPRM MY 2016 analysis fleet on a relative basis based 
on confidential business information submitted by the manufacturers on 
aerodynamic drag coefficients and other information sources such as in 
product release information. In addition, based on the Draft TAR 
comments, the agencies verified wherever possible the information 
submitted by manufactures with other sources (product release 
information and cross referencing with vehicles with similar design 
features and aero technologies), and made

[[Page 24555]]

changes particularly for vehicles which showed large improvements over 
baseline values. Figure 6-175 in PRIA presented the distribution of 
different levels of aerodynamic drag improvements in MY 2016 vehicle 
fleet in NPRM relative to MY 2015 vehicle fleet used in Draft TAR. The 
distribution shows that 46 percent of the MY 2016 vehicle fleet was 
assigned AERO0 (0 percent improvement), 31 percent of the fleet was 
assigned AERO5 (5% improvement), and 15 percent of the vehicle fleet 
was assigned AERO10 (10 percent improvement). This distribution clearly 
shows that there is substantial opportunity for additional aerodynamic 
drag improvements in the vehicle fleet.
    Regarding comments by ACEEE on Toyota Tundra pickup trucks, as just 
stated, the agencies used manufacturer submitted information and other 
available information to assign aerodynamic technology levels and the 
agencies applied the same process for all of the manufacturers for the 
NPRM and for the final rule. The agencies did assign AERO10 for some 
Toyota Tundra pickups, but not for all as asserted by ACEEE. Some of 
the Toyota Tundra pickups with 2WD and short bed and crew cab or double 
cab were assigned AERO5 and other configurations were assigned 
AER10.\1434\ For reference, the baseline Cd value used in 
the NPRM for pickups is 0.395; a 5 percent improvement in Cd value is 
0.375 and 10 percent improvement in Cd value is 0.355. The agencies 
considered the ACEEE comment and available information and determined 
the aerodynamic assignments for the Toyota Tundra were reasonable for 
the final rule analysis.
---------------------------------------------------------------------------

    \1434\ The variations could be from coast down testing with 
different powertrains and with different pickup bed length and crew 
cab configurations.
---------------------------------------------------------------------------

    Table VI-134 below shows the percentage aerodynamic drag 
improvement assigned to the MY 2015 (Draft TAR), MY 2016 (NPRM) and MY 
2017 (final rule) analysis fleets. It is clear from this table that 
there is natural progression of aero technologies being adopted and the 
vast majority of the MY 2017 vehicle fleet is at or below AERO10 
(81percent).
[GRAPHIC] [TIFF OMITTED] TR30AP20.270

    Moreover, notable aerodynamic improvements have actually been 
observed on production vehicles. As described in PRIA, EPA observed 76 
vehicles at the 2015 North American International Auto Show in Detroit 
(2015 NAIAS).\1435\ EPA's observations showed that manufacturers have 
widely deployed both active and passive aerodynamic drag reduction 
technologies with significant opportunity remaining to apply aero 
technologies further in more optimized fashion as vehicles enter 
redesign cycles in the future.\1436\ Although EPA did not identify the 
aerodynamic drag coefficient values for these vehicles, Figure 6-167 in 
PRIA showed the distribution of some aero technologies identified by 
EPA during this informal survey.
---------------------------------------------------------------------------

    \1435\ PRIA at 432. See also Docket No. EPA-HQ-OAR-2015-0827.
    \1436\ Draft TAR at 5-363.
---------------------------------------------------------------------------

    The survey showed that wheel dams and underbody panels are the most 
widely used aero technologies, followed by front bumper air dams and 
active grill shutters. Since this survey, many pickup trucks and 
passenger cars have active grill shutters installed to improve 
aerodynamic drag, and to get off-cycle credit. Table 6-67 in PRIA shows 
the ``active grill shutter'' by itself will improve aerodynamic drag 
reduction improvement by 3 percent. Combined with other aero 
technologies, this can improve the aerodynamic drag reduction values 
significantly in pickup trucks and SUVs. As a result, there has been 
overall fleet wide aerodynamic drag reduction improvement; however, the 
above Table VI-134 shows that only 19 percent (13 percent from AERO10, 
5 percent from AERO15 and 1 percent from AERO20) of the MY 2017 vehicle 
fleet has aerodynamic drag reduction improvement greater than 10 
percent. This shows that there is significant opportunity for the 
vehicle fleet to improve aero technologies by MY 2025.
    The agencies also described examples of how production vehicles in 
different technology classes improved aerodynamic drag reduction values 
relative to their previous generation model since the 2012 final 
rule.\1437\ The PRIA described how aerodynamic technologies were being 
deployed on production vehicles, using the MY 2015 Nissan Murano and MY 
2015 Ford F150 as examples. For example, MY 2015 Ford F150 has the 
passive and active aerodynamic technologies as shown in Table VI-135.
---------------------------------------------------------------------------

    \1437\ PRIA at 433.
---------------------------------------------------------------------------

    The air curtain technology in the MY 2015 F150 guides the air flow 
across the front wheels to reduce wind turbulence.\1438\ For reference, 
the wind tunnel testing by NRC of the MY 2015 Ford F150 showed a drag 
coefficient value of 0.37 while the coast down testing by EPA pegged 
the drag coefficient value between 0.35 to 0.40 depending on the type 
of powertrain, cab and cargo box combination. The prior generation F150 
was released in 2008 as a MY 2009 and this vehicle had

[[Page 24556]]

very few aerodynamic technologies applied. The agencies do not have the 
MY 2009 Cd value to estimate the percentage improvement. 
Since the F150 also included significant light weighting and powertrain 
improvements including a downsized turbocharged engine, the 
effectiveness improvement attributable to aerodynamic technologies is 
uncertain.
---------------------------------------------------------------------------

    \1438\ Ford, How Air Curtains on F-150 Help Reduce Aerodynamic 
Drag and Aid Fuel Efficiency (July 15, 2015), https://media.ford.com/content/fordmedia/fna/us/en/news/2015/07/15/how-air-curtains-on-f-150-help-reduce-aerodynamic-drag.html.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.271

    The Nissan Murano is an example of a mid-size SUV with greater than 
fifteen percent improvement in aerodynamic drag values compared to the 
previous generation. The SAE paper published in 2015 outlines the 
specifics of aerodynamics in the Nissan Murano,\1439\ and they include 
those listed in Table VI-136 below.
---------------------------------------------------------------------------

    \1439\ Arai, M., Tone, K., Taniguchi, K., Murakami, M. et al., 
``Development of the Aerodynamics of the New Nissan Murano,'' SAE 
Technical Paper 2015-01-1542, 2015, https://doi.org/10.4271/2015-01-1542.
---------------------------------------------------------------------------

    The exterior of this vehicle was completely redesigned from the MY 
2013-2014 generation with the goal of minimizing aerodynamic drag by 
combining passive aerodynamic devices with an optimized vehicle shape. 
The primary passive devices employed include optimization of the rear 
end shape to reduce rear end drag, and addition of a large front 
spoiler to reduce underbody air flow and redirect it toward the roof of 
the vehicle, thus augmenting the rear end drag improvements. Other 
passive improvements include plastic fillet moldings at the wheel 
arches, raising the rear edge of the hood, shaping the windshield 
molding and front pillars, engine under-cover and floor cover, and air 
deflectors at the rear wheel wells. An active lower grille shutter also 
redirects air over the body when closed. Together, these measures for 
the MY 2015 model achieved a drag coefficient of 0.31, representing a 
16 to 17 percent improvement over the 0.37 Cd of the 
previous model.

[[Page 24557]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.272


[[Page 24558]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.273

BILLING CODE 4910-59-C
    A combination of a slightly lighter MY 2015 Nissan Murano (on 
average lighter by 94 lbs. considering all trim levels), relative to 
the previous generation, and engine improvements (comparing 3.5L V6 in 
MY 2014 to 3.5L V6 in MY 2015), and transmission improvements resulted 
in an overall improvement in fuel economy.\1440\ Accordingly, the real-
world fuel economy improvement directly attributable to the package of 
aerodynamic technologies included on either vehicle is uncertain, as 
each vehicle included other fuel economy improving technologies along 
with the improvements in aerodynamic technologies.
---------------------------------------------------------------------------

    \1440\ https://www.fueleconomy.gov/feg/Find.do?action=sbs&id=34457&id=37198 (last visited 12.12.2019) shows 
20 mpg (combined) in MY2014 Nissan Murano (3.5L VQ35DE V6 with 
Variable gear ratio transmission) and 24 mpg (combined in MY2015 
Nissan Murano (3.5L VQ35DE V6 with Automatic AV S7 transmission)).
---------------------------------------------------------------------------

    The agencies considered a sensitivity case that assumed no mass 
reduction, rolling resistance, or aerodynamic improvements had been 
made to the MY 2017 fleet (i.e., setting all vehicle road levels to 
zero--MRO, AERO and ROLL0), in response to ICCT's comment. While this 
is an unrealistic characterization of the initial fleet, the agencies 
conducted a sensitivity analysis to understand any affect it may have 
on technology penetration along other paths (e.g., engine and hybrid 
technology). Under the CAFE program, the sensitivity analysis shows a 
slight decrease in reliance on engine technologies (HCR engines, 
turbocharge engines, and engines utilizing cylinder deactivation) and 
hybridization (strong hybrids and plug-in hybrids) in the baseline 
(relative to the central analysis). The consequence of this shift to 
reliance on lower-level road load technologies is a reduction in 
compliance cost in the baseline of about $300 per vehicle (in MY 2026). 
As a result, cost savings in the preferred alternative are reduced by 
about $200 per vehicle. Under the CO2 program, the general 
trend in technology shift is less dramatic (though the change in BEVs 
is larger) than the CAFE results. The cost change is also comparable, 
but slightly smaller ($200 per vehicle in the baseline) than the CAFE 
program results. Cost savings under the preferred alternative are 
further reduced by about $100. With the lower technology costs in all 
cases, the consumer payback periods decreased as well. These results 
are consistent with the approach taken by manufacturers who have 
already deployed many of the low-level road load reduction 
opportunities to improve fuel economy.
    Second, as discussed above, EPA's baseline aerodynamic levels in 
the Draft TAR were based on road load coefficients, leading to baseline 
assignments that were not accurate. In the NPRM, the agencies discussed 
in the tradeoffs between building the analysis fleet using confidential 
information from manufacturers and publicly available data on the 
vehicles.\1441\ In the case of drag coefficient values, which cannot be 
gleaned from publicly available information, except in cases where a 
manufacturer chooses to publicly release that data, or by simply 
observing a vehicle, the agencies decided that the improved accuracy 
associated with using manufacturer-provided Cd values 
outweighed the benefits of using publicly releasable Cd 
estimates based on road load coefficients, especially as manufacturer-
provided Cd values are only used to assign initial 
aerodynamic improvement levels relative to Cd values for 
each body style segment in the analysis fleet.
---------------------------------------------------------------------------

    \1441\ 83 FR 43004.
---------------------------------------------------------------------------

    In addition, manufacturers had submitted comments that the Draft 
TAR approach to baseline fleet assignments had underestimated 
technology already present on vehicles, leading the analysis to apply 
more aerodynamic drag reduction technology than could be applied in the 
real world. In response to those comments, as described in the Proposed 
Determination TSD, EPA stated that they ``agree[ ] with the commenters 
that it is appropriate to account for aerodynamic drag reductions 
already present in the baseline fleet in order to avoid overestimating 
the amount of additional improvement that can be achieved at a given 
cost.'' \1442\ Accordingly, EPA ``applied some level of aerodynamic 
drag reduction to a significant portion of the MY2015 baseline fleet.'' 
\1443\ Consequently, the agencies believe that ICCT's statement that if 
aerodynamic improvements between the MY 2015 analysis fleet used in the 
Draft TAR and the MY 2016 analysis fleet were true it would be evident 
in the fleet is incorrect. It is inappropriate to compare the Draft TAR 
MY 2015 analysis fleet, which notably included too few aerodynamic 
technology assignments, with the fleet's achieved fuel economy in the 
real world. The agencies disagree

[[Page 24559]]

with ICCT that the availability of aerodynamic technologies was 
artificially limited by appropriately assigning baseline aerodynamic 
technology levels in the analysis fleet.
---------------------------------------------------------------------------

    \1442\ Proposed Determination TSD at 2-406.
    \1443\ Proposed Determination TSD at 2-408.
---------------------------------------------------------------------------

    This also relates to ICCT's comment that the agencies must share 
the basis for any aerodynamic calculation and exact estimated percent 
improvement (rather than binned percentage categories) for each vehicle 
make and model in the baseline and future modeled fleet, and their 
technical justification for each value. As discussed above, the 
agencies shared the relative performance approach methodology for 
assigning baseline aerodynamic levels to vehicles in the analysis fleet 
in detail in the PRIA,\1444\ and this approach was the basis for the 
aerodynamic calculation performed for every vehicle make and model in 
the analysis fleet. The agencies provided the summary of aerodynamic 
drag coefficients (including averages for MY 2016 vehicles) by vehicle 
body style,\1445\ and the baseline aerodynamic improvement assignments 
for each vehicle model were included in the 
2018_NPRM_market_inputs_ref.xlsx. In addition, because aerodynamic drag 
information from manufacturers is provided as confidential business 
information, the agencies are unable to disclose that specific 
information. However, as discussed above, the agencies are closely 
examining the data provided and comparing it to other available 
information to assess the best estimate for aerodynamic technology for 
each vehicle in the analysis fleet.
---------------------------------------------------------------------------

    \1444\ PRIA at 441.
    \1445\ PRIA at 443.
---------------------------------------------------------------------------

    For these reasons, the agencies continued to use the NPRM 
methodology to assign aerodynamic drag reduction improvements for the 
MY 2017 vehicle fleet for this final rule.
c) Aerodynamic Drag Technology Adoption Features
    As discussed above, the agencies used a relative performance 
approach to assign current aerodynamic technology level to a vehicle. 
For some body styles with different utility, such as pickup trucks, 
SUVs and minivans, frontal area can vary, and this can affect the 
overall aerodynamic drag forces. In order to maintain vehicle utility 
and functionality related to passenger space and cargo space, the 
agencies assumed all technologies that improve aerodynamic drag forces 
would do so through reducing the Cd while maintaining 
frontal area.
    In the NPRM, the agencies noted that the Proposed Determination 
analysis assumed that some vehicles from all body styles could (and 
would) reduce aerodynamic forces by 20 percent, which in some cases led 
to future pickup trucks having aerodynamic drag coefficients better 
than some of today's typical cars, if frontal area were held constant 
in order to preserve interior space and cargo space. The agencies 
further noted that for some vehicle types, there was limited practical 
capability to significantly improve aerodynamic drag coefficients over 
baseline levels. In those cases, the agencies deemed the most advanced 
levels of aerodynamic drag simulated as not technically practicable 
given the need to maintain vehicle functionality and utility, such as 
interior volume, cargo area, and ground clearance.
    The industry had also commented in response to EPA's Proposed 
Determination on the difficulty to achieve AERO20 improvements for 
certain body styles. In the NPRM, the agencies considered the industry 
comments along with the observations made in the MY 2016 fleet, and 
tentatively determined the maximum feasible improvement in 
Cd that could be achieved for pickup trucks is AERO15.\1446\ 
Similarly, the agencies determined the maximum feasible improvement in 
Cd that could be achieved for minivans is AERO10. Next, the 
NPRM analysis did not apply 15 percent or 20 percent aerodynamic drag 
coefficient reduction to cars and SUVs with more than 405 horsepower. 
The agencies noted that many high-performance vehicles already include 
advanced aerodynamic features despite middling aerodynamic drag 
coefficients. In these high-performance vehicle cases, the agencies 
recognized that manufacturers tune aerodynamic features to provide 
desirable downforce at high speeds and to provide sufficient cooling 
for the powertrain, and, therefore, manufacturers may have limited 
ability to improve aerodynamic drag coefficients for high performance 
vehicles with internal combustion engines without reducing horsepower. 
Accordingly, the agencies did not allow application of AERO15 and 
AERO20 technology for all vehicles with more than 405 HP. Approximately 
400,000 units of volume in the MY 2016 market data file included 
limited application of aerodynamic technologies because of vehicle 
performance. The agencies sought comment on limiting the Cd 
improvement in these circumstances.
---------------------------------------------------------------------------

    \1446\ The agencies noted in the NPRM that although ANL created 
full-vehicle simulations for trucks with 20 percent drag reduction, 
those simulations were not used in the CAFE modeling. The agencies 
concluded that level of drag reduction was likely not 
technologically feasible with today's technology, and the analysis 
accordingly restricted the application of advanced levels of 
aerodynamics in some instances, such as in that case, due to 
bodystyle form drag limitations.
---------------------------------------------------------------------------

    Ford commented in support of the agencies' decision to limit the 
application of AERO20 on pickup trucks, noting that limiting AERO20 on 
pickups is appropriate given the high inherent form drag associated 
with pickups' aerodynamic profile.\1447\
---------------------------------------------------------------------------

    \1447\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    CARB commented that the agencies excluded AERO20 inconsistently 
across the fleet, noting that while some of the restrictions may be 
valid, the broad rule the agencies used resulted in technology being 
inappropriately excluded from too many vehicles.\1448\ Specifically, 
CARB took issue with the majority of luxury sedans and SUVs being 
excluded from AERO20 because they had high horsepower engines, while 
the agencies did assign AERO20 to vehicles like the Tesla Model S and 
Model X SUVs, which have horsepower in excess of 405. CARB stated that 
while electrification provides a higher motivation to minimize road 
load through technologies such as aerodynamic reductions, implementing 
AERO20 reductions on high horsepower sedans and SUVs is clearly 
feasible and should not be artificially restricted in the CAFE model.
---------------------------------------------------------------------------

    \1448\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    In addressing these comments, the agencies considered the relative 
cooling requirements for all electric powertrains and for high 
performance internal combustion engine powertrains since airflow 
diverted for cooling adversely impacts a vehicle's Cd. The 
peak heat rejection and engine cooling needs for high performance 
internal combustion engines is significantly higher than for all 
electric powertrains. Internal combustion engines convert a lower 
percentage of energy contained in gasoline into mechanical work (and 
other useful work, such as lighting and sound), and the energy not 
converted into mechanical work (or other useful work) is converted into 
heat. A significant amount of the waste heat must be handled by the 
cooling systems. Battery electric vehicles convert most of the 
electrical energy stored in the battery into mechanical work and other 
useful work, and therefore convert less energy into heat that must be 
handled by the cooling system. Also, electric powertrains can provide a 
degree of electric braking, whereas internal combustion engines 
exclusively use friction braking, which generates heat and requires 
greater cooling,

[[Page 24560]]

particularly on vehicles with substantial braking performance 
capabilities. In the case of high-performance BEVs, since the cooling 
needs are not as demanding as with high-performance vehicles that use 
internal combustion engines, manufacturers can (and do, as can be 
observed in the fleet) apply higher levels of aerodynamic technologies. 
The agencies believe it is appropriate to account for these differences 
in considering the amount of aerodynamic improvement that can be 
implemented, and determined there are valid technical reasons for 
allowing BEVs with greater than 405 horsepower to adopt AERO20 
technology.
d) Aerodynamic Drag Technology Effectiveness
    The NPRM analysis included four levels of aerodynamic improvements, 
AERO5, AERO10, AERO15, and AERO20, representing 5, 10, 15, and 20 
percent Cd improvements, respectively. Notably, the NPRM 
analysis assumed that aerodynamic drag reduction could only come from 
reduction in the aerodynamic drag coefficient and not from reduction of 
frontal area, to maintain vehicle functionality and utility, such as 
passenger space, ingress/egress ergonomics, and cargo space.\1449\
---------------------------------------------------------------------------

    \1449\ 83 FR 43047.
---------------------------------------------------------------------------

    Ford commented in support of the agencies' decision to consider the 
frontal area and body style as ``utility factors'' and requiring that 
aerodynamic improvements come from reductions in Coefficient of Drag 
(Cd) and not from reductions in frontal area.\1450\
---------------------------------------------------------------------------

    \1450\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    CBD commented that EPA staff had critiqued NHTSA's characterization 
of research on aerodynamic drag coefficients and the NPRM did not 
appear to incorporate or respond to this input.1451 1452 
Specifically, CBD stated that EPA staff had commented in response to 
the characterization that ``[f]or some bodystyles, the agencies have no 
evidence that manufacturers may be able to achieve 15 percent or 20 
percent aerodynamic drag coefficient reduction relative to baseline 
(for instance, with pickup trucks'' and noted that ``[i]n the past, EPA 
has assigned aero tech in the baseline relative to a ``Null'' and then 
applied drag reduction level against that Null in order to ensure that 
the maximum aero level (i.e., 15 or 20 percent) would always be 
achievable for all body styles.'' This comment reflects deliberative, 
in-process input from EPA staff. In fact, the NPRM text was developed 
by the agencies with the benefit of this and other input from EPA 
staff, and the NPRM clarified that reducing frontal area would likely 
degrade other utility features like interior volume or ingress/egress.
---------------------------------------------------------------------------

    \1451\ NHTSA-2018-0067-12000, at 188.
    \1452\ Docket No. EPA-HQ-OAR-2018-0283-0453, June 29, 2018 
Comments at 93.
---------------------------------------------------------------------------

    CARB commented, as part of its broader comments, that the agencies' 
effectiveness values were reduced relative to what EPA's LPM 
calculated, that the benefits of aerodynamic improvements were 
underestimated.\1453\ Specifically, CARB cited the H-D Systems 
comparison of LPM benefits for AERO10 and AERO20 of 2.1 percent and 4.3 
percent, respectively, compared with Autonomie benefits of 1.51 percent 
and 3.03 percent, respectively, and stated that the agencies' analysis 
provided no description or cited any new data or evidence as to why 
they reduced the projected assumptions compared to what EPA's Lumped 
Parameter Model calculated.
---------------------------------------------------------------------------

    \1453\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    HDS also commented that the Autonomie modeling assumed no engine 
change when aerodynamic drag and rolling resistance reductions were 
implemented, as well as no changes to the transmission gear ratios and 
axle ratios, which vary by transmission type but not by the tractive 
load.\1454\ HDS stated that the EPA ALPHA model adjusted for this 
effect, which accounted for the difference in technology effectiveness 
estimates that HDS characterized between the Draft TAR and NPRM. HDS 
provided a ``correct estimate'' for AERO20 effectiveness improvements 
of 4.3 percent, with the justification that there was no gear/axle 
ratio adjustment in the Autonomie analysis.
---------------------------------------------------------------------------

    \1454\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    In response to HDS's comment, the Alliance submitted supplemental 
comments questioning the extent to which aerodynamics (and changes in 
top gear ratio) affect performance metrics held constant in the 
analysis, like low- and high-speed acceleration performance and 
gradeability.\1455\ The Alliance cited a study for the proposition that 
vehicle acceleration is most influenced by engine power and weight, and 
also that bodystyle differences have a lesser impact on acceleration 
performance. The Alliance further commented that ``[r]egarding changes 
in top gear ratios in response to aerodynamic changes, the Alliance is 
not aware of any examples in which a top gear ratio was changed solely 
due to aerodynamic improvements. There may be examples where a 
vehicle's top gear ratio was changed at the same time aerodynamic 
changes were made, but such changes would be made in response to the 
cumulative changes across the entire vehicle, not just aerodynamic 
improvements.'' The Alliance concluded that ``[t]here are also 
practical manufacturing and investment constraints which limit the 
potential for applying engine changes in response to improved vehicle 
aerodynamics,'' citing the agencies decision to only resize engines 
with significant design changes, to account for product complexity and 
economies of scale.
---------------------------------------------------------------------------

    \1455\ NHTSA-2018-0067-12385, at 31-32.
---------------------------------------------------------------------------

    In response to the Alliance's supplemental comment, HDS submitted 
supplemental comments stating that ``[d]rag reduction is usually 
accomplished when a vehicle body is redesigned, so gear and axle ratios 
are typically re-optimized for the entire set of changes, but these 
changes include the drag reduction.'' \1456\ HDS commented that the 
Alliance's comments acknowledged that calibration changes are made in 
response to tractive load changes, while the Autonomie analysis 
recalibrates the powertrain in response only to large mass reduction 
improvements, and not any other vehicle changes that reduce tractive 
load, like aerodynamic improvements, even when those changes would 
result in a greater tractive load reduction than a 10 percent mass 
reduction. HDS reiterated its statement that ``[i]n the real world (and 
as captured in EPA's prior ALPHA model), automakers typically alter 
many vehicle attributes affecting tractive load simultaneously, 
including aerodynamics,'' and the Autonomie outputs underrepresent the 
benefit of tractive load reduction strategies by not optimizing engine 
efficiency after most changes in tractive load because the model 
employees fixed shift points, gear ratios, and axle ratios when drag or 
tire rolling resistance is reduced.
---------------------------------------------------------------------------

    \1456\ NHTSA-2018-0067-12395, at 4-5.
---------------------------------------------------------------------------

    Regarding the first set of comments that the aerodynamic 
effectiveness values were reduced from EPA's values presented in the 
Draft TAR, that results from differences in the two modeling 
approaches. As discussed above, for this analysis the agencies decided 
that aerodynamic drag reduction could only come from reduction in the 
aerodynamic drag coefficient, and not from a reduction in vehicle 
frontal area, at least without reducing other attributes of the 
vehicle. EPA's process for assigning road load technologies to baseline 
vehicles used road load coefficients from coast downs, which aggregated 
individual aero, mass and tire reduction technologies. In contrast,

[[Page 24561]]

the CAFE Model and Autonomie used individually assigned road load 
technologies for each vehicle to appropriately assign initial road load 
and to appropriately capture benefits of subsequent individual road 
load technologies. The differences in using road load coefficients from 
coast downs and individually isolating the improvements from existing 
and future road load technologies in the Autonomie modeling resulted in 
the differences noted by commenters. And so, the resulting 
effectiveness from the incremental adoption of individual technologies 
to a newer analysis fleet will have different result than what was 
estimated by the previous analyses. For further discussion of the 
analysis fleet see Section VI.B.1.
    In Section VI.B.3 Tech Effectiveness and Modeling and Section 
VI.C.2 Transmissions, the agencies provide a full discussion of the 
issues associated with assuming the engine and transmission can be 
optimized for every combination of technologies. It would be 
unreasonable and unaffordable to resize powertrains, including engines 
and transmission and axle ratios, for every unique combination of 
technologies, and exceedingly so for every unique combination 
technologies across every vehicle model due to the extreme 
manufacturing complexity that would be required to do so. Product 
complexity and economies of scale are real, and in the NPRM, engine 
resizing was limited to specific incremental technology changes that 
would typically be associated with a major vehicle or engine 
redesign.\1457\ As noted by HDS, the EPA Draft TAR and Proposed 
Determination analyses adjusted the effectiveness of every technology 
combination, including for aerodynamics technologies, assuming 
performance could be held constant for every combination. However, 
those analyses did not recognize or account for the extreme complexity 
nor the associated costs for that impractical assumption. The NPRM and 
final rule analyses account for these real-world practicalities and 
constraints, and doing so explains some of the effectiveness and cost 
differences between the Draft TAR/Proposed Determination and the NPRM/
final rule. The agencies believe the NPRM and the final rule approach 
appropriately resizes powertrain components for specific incremental 
technology changes that would typically be associated with a major 
vehicle or engine redesign.
---------------------------------------------------------------------------

    \1457\ See 83 FR 43027 (Aug. 24, 2018).
---------------------------------------------------------------------------

    For the NPRM, and carried into the final rule analysis, Autonomie 
simulates all road load conditions (e.g., MR, AERO, and ROLL technology 
levels) for each engine and transmission combination. In addition, 
engines are resized for appropriate specific technology changes that 
would be associated with a major vehicle or engine redesign. Also, as 
discussed further in Section VI.C.2 Transmissions, many commenters 
seemed to conflate the practice in the analysis of using a common 
(same) gear set across vehicle configurations (to address manufacturing 
complexity) with using the same shift maps. As commenters stated, they 
assumed the same shift maps were applied across vehicle models. 
However, the shift initializer routine was run for every unique 
Autonomie full vehicle model configuration and generated customized 
shifting maps. The algorithms' optimization was designed to balance 
minimization of energy consumption and vehicle performance. This 
balance was necessary to achieve the best fuel efficiency while 
maintaining customer acceptability by meeting performance neutrality 
requirements. The agencies believe the level of optimization of engine 
size, transmissions, gear ratios and shift schedules reasonably 
approximate what is achievable and what manufacturers actually do.
    Figure VI-47 below shows the range effectiveness used for AERO 
technologies for the NPRM analysis.
BILLING CODE 4910-59-P

[[Page 24562]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.274

    Figure VI-48 below shows the range of aero effectiveness used for 
the final rule analysis.

[[Page 24563]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.276

BILLING CODE 4910-59-C
e) Aerodynamic Drag Technology Cost
    For the Draft TAR, the agencies relied on the 2015 NAS report to 
estimate the cost of AERO1 and AERO2 levels of aerodynamic drag 
coefficient improvements. The agencies received several comments 
related to the cost assumptions used in the Draft TAR, mainly that they 
were too low to meet AERO1 and AERO2 levels. The industry submitted 
confidential business information on the costs of passive aerodynamic 
technologies needed to achieve AERO1 (10 percent improvement in drag 
improvement), which showed a significantly higher estimated costs than 
assumed for the Draft TAR. Similarly, the industry submitted 
confidential business information on the costs of active aerodynamic 
technologies, including some high cost technologies. The industry also 
commented that some active aerodynamic technologies could only be 
implemented during vehicle redesigns and not during a mid-cycle vehicle 
refresh.
    The agencies considered these comments and performed additional 
research to assess the costs for passive and active aerodynamic 
technologies. The agencies revised the cost estimates for the NPRM, 
based in part on confidential information from the automotive industry, 
and from the agencies' own assessment of manufacturing costs for 
specific aerodynamic technologies from available sources. In general, 
the NPRM cost estimates were higher than Draft TAR cost estimates. The 
agencies included a high-level discussion in the PRIA that the cost to 
achieve AERO5 is relatively low, as most of the improvements can be 
made through body styling changes. The cost to achieve AERO10 is higher 
than AERO5, due to the addition of several passive aero technologies, 
and the cost to achieve AERO15 and AERO20 is higher than AERO10 due to 
use of both passive and active aero technologies.
    The agencies did not receive any comments on the costs of 
aerodynamic improvements, and accordingly, for the final rule, as shown 
in Table VI-137 and Table VI-138 below, the agencies used the same 
aerodynamic improvement costs presented in NPRM.

[[Page 24564]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.277

[GRAPHIC] [TIFF OMITTED] TR30AP20.278

6. Tire Rolling Resistance
    Tire rolling resistance is a road load force that arises primarily 
from the energy dissipated by elastic deformation of the tires as they 
roll. Tire design characteristics (for example, materials, 
construction, and tread design) have a strong influence on the amount 
and type of deformation and the energy it dissipates. Designers can 
select these characteristics to minimize rolling resistance. However, 
these characteristics may also influence other performance attributes, 
such as durability, wet and dry traction, handling, and ride comfort.
    Low rolling resistance tires are increasingly specified by OEMs in 
new vehicles and are also increasingly available from aftermarket tire 
vendors. They commonly include attributes such as higher inflation 
pressure, material changes, tire construction optimized for lower 
hysteresis, geometry changes (e.g., reduced aspect ratios), and reduced 
sidewall and tread deflection. These changes are commonly accompanied 
by additional changes to vehicle suspension tuning and/or suspension 
design to mitigate any potential impact on other performance attributes 
of the vehicle.
    Lower-rolling-resistance tires have characteristics that reduce 
frictional losses associated with the energy dissipated mainly in the 
deformation of the tires under load, thereby improving fuel economy and 
reducing CO2 emissions. The agencies considered two levels 
of improvement for low rolling resistance tires in the analysis: The 
first level of low rolling resistance tires considered reduced rolling 
resistance 10 percent from an industry-average baseline, while the 
second level reduced rolling resistance 20 percent from the baseline.
    Walter Kreucher commented that the agencies should eliminate low 
rolling resistance tires from the list of viable technologies, in 
recognition of the safety impacts of low rolling resistance tires in 
relation to stopping distance and accident rates.\1458\ Separately, Mr. 
Kreucher argued that the model should reflect the safety impact of low 
rolling resistance tires.
---------------------------------------------------------------------------

    \1458\ NHTSA-2018-0067-0444.
---------------------------------------------------------------------------

    The agencies have been following the industry developments and 
trends in application of rolling resistance technologies to light duty 
vehicles. As stated in the NAP special report on Tires and Passenger 
Vehicle Fuel Economy,\1459\ cited by Mr. Kreucher, national crash data 
does not provide data about tire structural failures specifically 
related to tire rolling resistance, because the rolling resistance of a 
tire at a crash scene cannot be determined. However, other metrics like 
brake performance compliance test data

[[Page 24565]]

are helpful to show trends like that stopping distance has not changed 
in the last ten years,\1460\ during which time many manufacturers have 
installed low rolling resistance tires in their fleet--meaning that 
manufacturers were successful in improving rolling resistance while 
maintaining stopping distances through tire design, tire materials, 
and/or braking system improvements. In addition, NHTSA has addressed 
other tire-related issues through rulemaking,\1461\ and continues to 
research tire problems such as blowouts, flat tires, tire or wheel 
deficiency, tire or wheel failure, and tire degradation.\1462\ However, 
there are currently no data connecting low rolling resistance tires to 
accident or fatality rates.
---------------------------------------------------------------------------

    \1459\ Tires and Passenger Vehicle Fuel Economy: Informing 
Consumers, Improving Performance--Special Report 286 (2006), 
available at https://www.nap.edu/read/11620/chapter/6.
    \1460\ https://one.nhtsa.gov/cars/problems/comply/index.cfm.
    \1461\ 49 CFR 571.138, Tire pressure monitoring systems.
    \1462\ Tire-Related Factors in the Pre-Crash Phase, DOT HS 811 
617 (April 2012), available at https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811617.
---------------------------------------------------------------------------

    With better tire design, tire compound formulations and improved 
tread design, tire manufacturers have tools to balance stopping 
distance and reduced rolling resistance. As stated in one article 
referenced by Mr. Kreucher, tire manufacturers can use ``higher 
performance materials in the tread compound, more silica as reinforcing 
fillers and advanced tread design features'' to mitigate issues related 
to stopping distance.\1463\ The agencies do not believe that there is 
sufficient data or other information to support removing low rolling 
resistance tires as a viable technology considered in the CAFE and 
CO2 analysis at this time.
---------------------------------------------------------------------------

    \1463\ Jesse Snyder, A big fuel saver: Easy-rolling tires (but 
watch braking) (July 21, 2008), https://www.autonews.com/article/20080721/OEM01/307219960/a-big-fuel-saver-easy-rolling-tires-but-watch-braking. Last visited December 3, 2019.
---------------------------------------------------------------------------

    HDS argued, as discussed further below, that based on available 
data on current vehicle models and the likely possibility that there 
would be additional tire improvements over the next decade, the 
agencies should consider ROLL30 technology, or a 30 percent reduction 
of tire rolling resistance over the baseline.\1464\
---------------------------------------------------------------------------

    \1464\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    As stated in Joint TSD for the 2017-2025 final rule, tire 
technologies that enable rolling resistance improvements of 10 and 20 
percent have been in existence for many years.\1465\ Achieving 
improvements of up to 20 percent involves optimizing and integrating 
multiple technologies, with a primary contributor being the adoption of 
a silica tread technology. Tire suppliers have indicated that 
additional innovations are necessary to achieve the next level of low 
rolling resistance technology on a commercial basis, such as 
improvements in material to retain tire pressure, tread design to 
manage both stopping distance and wet traction, and development of 
carbon black material for low rolling resistance without the use of 
silica to reduce cost and weight.\1466\ The agencies are continuously 
monitoring these and other tire technology improvements. The agencies 
believe that the tire industry is in the process of moving automotive 
manufacturers towards the first level of low rolling resistance 
technology across the vehicle fleet (10 percent reduction in rolling 
resistance), and that 20 percent improvement is achievable in the 
rulemaking timeframe. However, the agencies believe that at this time, 
the emerging tire technologies that would achieve 30 percent 
improvement in rolling resistance, like changing tire profile, 
strengthening tire walls, or adopting improved tires along with active 
chassis control,\1467\ among other technologies, will not be available 
for commercial adoption in the fleet during the rulemaking timeframe. 
As a result, the agencies decided not to incorporate 30 percent 
reduction in rolling resistance technology for this final rule.
---------------------------------------------------------------------------

    \1465\ EPA-420-R-12-901, at page 3-210.
    \1466\ Assessment of Fuel Economy Technologies for Light-Duty 
Vehicles (2011) at page 103.
    \1467\ Mohammad Mehdi Davari, Rolling resistance and energy loss 
in tyres (May 20, 2015), available at https://www.sveafordon.com/media/42060/SVEA-Presentation_Davari_public.pdf. Last visited 
December 30, 2019.
---------------------------------------------------------------------------

a) Rolling Resistance Modeling in the CAFE Model
    The two levels of rolling resistance technology considered in the 
analysis include ROLL10 and ROLL20, which represent a 10 percent and 20 
percent rolling resistance reduction from the baseline (ROLL0), 
respectively.
    To understand the following discussions about rolling resistance 
analysis fleet assignments and effectiveness values, it is important to 
understand how the agencies developed the baseline value (ROLL0) used 
in prior analyses, and how the agencies developed the baseline value 
used in the NPRM and final rule. In the Draft TAR, the agencies used 
unique baseline rolling resistance coefficients for each vehicle class. 
Specifically, the compact car class value was 0.0075, the midsize car 
value was 0.008, the small SUV value was 0.0084, the midsize SUV value 
was 0.0084, and the pickup truck value was 0.009. The PRIA described 
that since the Draft TAR, the agencies had reassessed rolling 
resistance values for contemporary tires through discussions with 
vehicle manufacturers, tire manufactures, and independent bench 
testing. Based on a thorough review of confidential business 
information submitted by industry, and a review of other literature, 
including the CARB/CONTROLTEC study mentioned below, the baseline 
rolling resistance coefficient for all vehicle classes was updated to 
0.009 for the NPRM analysis. The agencies concluded that the updated 
baseline value brought the NPRM simulations into better alignment with 
tires in the MY 2016 analysis fleet. The agencies also discussed that 
updated value was consistent with the findings of the CONTROLTEC study 
on vehicle road loads, sponsored by CARB.\1468\ The following figure 
shows the distribution of estimated tire rolling resistance coefficient 
values for the 1,358 MY 2014 vehicles evaluated in the CONTROLTEC/CARB 
study.
---------------------------------------------------------------------------

    \1468\ Technical Analysis of Vehicle Load Reduction Potential 
for Advanced Clean Cars, https://www.arb.ca.gov/research/apr/past/13-313.pdf, page 39.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24566]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.279

BILLING CODE 4910-59-C
    ICCT commented that it was ``quite confusing and perhaps 
troubling'' that the agencies adopted a higher average rolling 
resistance coefficient than that of the Draft TAR, ``as it would imply 
that the fleet rolling resistance got worse, but the agencies are 
deciding to provide baseline credit as if there was more rolling 
resistance technology deployed.'' \1469\ ICCT stated that the change 
appeared to be attributed to the agencies' use of CBI on tire rolling 
resistance received since the Draft TAR.
---------------------------------------------------------------------------

    \1469\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    As described in the PRIA, the values used in the Draft TAR 
represented the ``Best in Class'' values in each of the vehicle classes 
and this did not necessarily reflect the average ``Rolling Resistance 
Coefficient'' (RRC) of the fleet. For the Draft TAR, the agencies did 
not have access to manufacturer confidential business information and 
relied on estimates from CONTROLTEC. As stated earlier, Figure VI-49 
shows the distribution of the estimated RRC for 1,358 vehicles models. 
The average RRC from the CONTROLTEC study (0.009) aligned with the NPRM 
estimate which was based in part on manufacturer submitted confidential 
business information. CONTROLTEC compared the estimated RRC data with 
the values provided by Rubber Manufacturers Association (renamed as 
USTMA-U.S. Tire Manufacturers Association) for original equipment 
tires. The average RRC from the data provided by RMA was 0.0092,\1470\ 
compared to average of 0.009 from CONTROLTEC. CONTROLTEC attributed the 
difference due to analysis assumption, tire loading during coast down 
vs. load during tire testing, inflation pressure during coast down vs. 
inflation pressure during tire testing, coast down test reporting 
issues, tire types represented in the sample, tire break-in, and 
advancement in tire rolling resistance since the time RMA collected the 
data.
---------------------------------------------------------------------------

    \1470\ Technical Analysis of Vehicle Load Reduction by 
CONTROLTEC for California Air Resources Board (April 29, 2015) at 
page 40.
---------------------------------------------------------------------------

    CONTROLTEC also stated that RRC values for some vehicles fell below 
the average RRC (indicating better performance) due to estimation 
assumptions for vehicles where manufacturer data was not available, and 
coast down test reporting issues.\1471\ Further, CONTROLTEC performed a 
sensitivity study by mathematically removing aerodynamic contribution 
from the coast down coefficients. It was observed that the average RRC 
without the aerodynamic contribution is around 0.011. Accordingly, the 
agencies believe that it was reasonable to use 0.009 as the average RRC 
for the fleet for the NPRM and to continue to use that value for the 
final rule, based on the latest available data from manufacturers and 
alignment with the average RRC to the CONTROLTEC study estimate.
---------------------------------------------------------------------------

    \1471\ Technical Analysis of Vehicle Load Reduction by 
CONTROLTEC for California Air Resources Board (April 29, 2015) at 
page 38.
---------------------------------------------------------------------------

    H-D Systems (HDS) commented that the CONTROLTEC/CARB study showed 
that there is a very significant fraction of the fleet with tire 
rolling resistance coefficients above 10kg/1000 kg, and a small 
percentage of vehicles with rolling resistance coefficients already at 
0.05 or 0.06. HDS stated that NHTSA's baseline of 0.09 appeared ``a 
little low but may be appropriate if the distribution was sales 
weighted.'' HDS argued that a number of vehicle models already have 
tires below 0.07, and the likelihood that there would be additional 
tire improvements over the next decade are likely, meaning that ROLL30 
technology--or a 30 percent reduction of the tire rolling resistance 
coefficient to 0.063--is possible and appropriate for MY 2025.
    Roush commented that rolling resistance is erroneously assumed to 
be the same across different vehicle classes, and that rolling 
resistance would vary depending upon the vehicle size, power, 
acceleration and performance package.\1472\
---------------------------------------------------------------------------

    \1472\ NHTSA-2018-0067-11984.
---------------------------------------------------------------------------

    As explained earlier, the RRC values used in the CONTROLTEC study 
were a combination of manufacturer information, estimates from coast 
down tests for some vehicles, and application of tire RRC values across 
other vehicles on the same platform. CONTROLTEC stated that some RRC 
values were below the estimated average (showing significant 
improvement from the baseline) due to assumptions that were

[[Page 24567]]

applied to some vehicles when manufacturer data was not available. 
Further, some of the RRC estimates were based on vehicle coast down 
tests which had errors.\1473\ As a result, some of the RRC values used 
in the Draft TAR showed significant improvements (30 percent reduction 
in rolling resistance relative to baseline), as observed by HDS. Based 
on a review of manufacturer-submitted confidential business information 
and other sources, the agencies are unaware of any tires in production 
which have 30 percent reduction in rolling resistance relative to 
baseline values.
---------------------------------------------------------------------------

    \1473\ Technical Analysis of Vehicle Load Reduction by 
CONTROLTEC for California Air Resources Board (April 29, 2015) at 
page 38.
---------------------------------------------------------------------------

    As stated earlier, the baseline values used for the Draft TAR 
analysis were ``Best in Class'' values from the estimates developed by 
CONTROLTEC and not representative of the average of the fleet or 
average for the vehicle classes. For the NPRM, the agencies revisited 
the ROLL technology assignments based on the RRC values provided by 
manufacturers, and the average RRC for each of the vehicle class was 
near the fleet average (RRC = 0.009). As shown in Figure VI-50, a vast 
majority of the vehicles in the fleet are in the ROLL0 bin across the 
different vehicle class, vehicle size, power, acceleration and 
performance configurations. For these reasons, the agencies will 
continue to use the fleet average of RRC = 0.009 as the baseline value 
to assess ROLL technology improvements.
b) Rolling Resistance Analysis Fleet Assignments
    As discussed above, NHTSA's Draft TAR analysis showed little 
rolling resistance technology in the baseline fleet for three reasons: 
the simulations used baseline values already reflecting best-in-class 
tire rolling resistance, credible tire rolling resistance values for 
all vehicles from bench data were not available to the agencies at the 
time of Draft TAR, and few manufacturers submitted rolling resistance 
values for the Draft TAR analysis.
    For the NPRM, baseline (ROLL0) rolling resistance values were 
updated to 0.009, and any better rolling resistance values were 
assigned based on whether information indicated that vehicle had 
technology at least 10 percent better than baseline (.0081 or better 
for ROLL10), or at least 20 percent better than baseline (.0072 or 
better for ROLL20). The agencies used confidential business information 
provided by manufacturers to assign initial rolling resistance values 
for each vehicle make and model.
    The Alliance commented that the NPRM MY 2016 analysis fleet had 
been updated with appropriate ratings of rolling resistance 
improvements, compared to the Draft TAR where vehicles were generally 
considered to have unimproved tires (meaning the Draft TAR assumed 
additional improvements were more achievable than in reality).\1474\ 
The Alliance noted that the Draft TAR approach led to the CAFE model 
adding additional tire rolling resistance improvements even though 
manufacturers had already made significant improvements with that 
technology. This meant that the real-world fleet had little remaining 
opportunity for additional tire-related improvements, ultimately 
leading to the Draft TAR analysis underpredicting the amount of 
powertrain technology required for compliance.
---------------------------------------------------------------------------

    \1474\ NHTSA-2018-006712039 at 136.
---------------------------------------------------------------------------

    The Alliance noted that it is possible to estimate rolling 
resistance features of a vehicle using road load coefficients, but the 
process requires various assumptions and is not very accurate. The 
Alliance concluded that the agencies' use of CBI to assign baseline 
technology levels correctly was an accurate and practical solution. 
Similarly, Ford commented in support of the agencies' low rolling 
resistance tire assignments in the baseline fleet, stating that the 
accuracy of the baseline fleet assessment had been considerably 
improved using actual tire rolling resistance data.\1475\
---------------------------------------------------------------------------

    \1475\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    HDS commented that the analysis fleet ``accounts for the 
distribution of tires below 0.09 as 19% of vehicles in MY 2016 are 
modeled as having used ROLL10 and 25% of vehicles as having used ROLL20 
in the base year, but there is no accounting for the ~25% of vehicles 
having RRC values 10 to 20% above the 0.09 RRC average.'' \1476\ HDS 
concluded that ``[a] stricter accounting of the baseline and, possibly 
setting specific lower limits for 2025 RRC by vehicle type (as done for 
aero drag in the PRIA) will show significant additional fleetwide 
effectiveness from RRC reduction which is a very cost-effective 
technology.''
---------------------------------------------------------------------------

    \1476\ NHTSA-2018-0067-11985 at 49.
---------------------------------------------------------------------------

    ICCT commented that the agencies made a ``dramatic and 
unjustified'' shift in baseline tire rolling resistance assignments 
from the 2015 fleet used in the Draft TAR to the 2016 fleet used in the 
NPRM.\1477\ ICCT noted that per the agencies' updated baseline value, 
nearly 20 percent of all vehicles in the MY 2016 analysis fleet 
achieved 0.0081 (or better) rolling resistance value, and more than 26 
percent achieve 0.0072 (or better). ICCT argued that rather than 
changing the definition of rolling resistance technology to include 
improvements beyond the baseline, the agencies instead redefined the 
technology available, reducing the number of vehicles that can use tire 
improvements in future compliance years within the modeling framework, 
which artificially forced companies to use other, more expensive 
technologies.
---------------------------------------------------------------------------

    \1477\ NHTSA-2018-0067-11741 full comments.
---------------------------------------------------------------------------

    ICCT stated that to substantiate the baseline rolling resistance 
assignments, the agencies need to show data on how these improvements 
are evident in the fleet and delivering benefits. ICCT alleged that if 
an improvement of that magnitude were true, it would be evident in 
fleet level miles-per-gallon and CO2 levels; however, ``none 
of the quantifiable mpg or CO2 benefits that would be 
associated with these additional rolling resistance improvements were 
reflected with any real-world evidence in the model year 2016 fleet.'' 
ICCT stated this seemed to be a case of the agencies ``artificially 
burying efficiency technology in the baseline, rendering it unusable in 
the post model year 2016 compliance scenarios.''
    ICCT also stated that the agencies must share absolute road load 
coefficients for each vehicle make and model in the baseline fleet, and 
the technical justification for each value, in addition to conducting 
two sensitivity analysis cases ``assum[ing] that every baseline make 
and model is set to 0% rolling resistance improvement and set to the 
previous baseline rolling resistance (from the Draft TAR) to 
demonstrate how much the agencies' decision to load up more baseline 
technology affects the compliance scenarios, as it appears that the 
agencies may have made a unsupportable and non-rigorous assumption 
about rolling resistance technology across the models.'' ICCT concluded 
that because the changes were buried in the datafiles and unexplained, 
the agencies must issue a new regulatory analysis and allow an 
additional comment period for review of the methods and analysis.
    Based on the comments from HDS and ICCT, the agencies reexamined 
available tire rolling resistance data. The assignment of ROLL20 
technology was revised for some vehicle models based on information on 
the use of common tires across vehicles that shared a platform. As a 
consequence, for the final rule, only 20 percent of the MY2017 vehicle 
fleet is assigned ROLL20. The

[[Page 24568]]

agencies will continue to investigate additional methods to improve the 
accuracy of this method, however as the Alliance and Ford noted, the 
accuracy of the baseline levels had been significantly improved over 
prior analyses by using actual tire RRC data. The agencies approach is 
consistent with the NAS recommendation to have two ROLL technology 
levels. The agencies determined that 30 percent rolling resistance 
improvement while maintaining other tire characteristics is unlikely to 
be available in the rulemaking timeframe.
    The agencies considered a sensitivity case that assumed no mass 
reduction, rolling resistance, or aerodynamic improvements had been 
made to the MY 2017 fleet (i.e., setting all vehicle road levels to 
zero--MRO, AERO and ROLL0), in response to ICCT's comment. While this 
is an unrealistic characterization of the initial fleet, the agencies 
conducted a sensitivity analysis to understand any affect it may have 
on technology penetration along other paths (e.g. engine and hybrid 
technology). Under the CAFE program, the sensitivity analysis shows a 
slight decrease in reliance on engine technologies (HCR engines, 
turbocharge engines, and engines utilizing cylinder deactivation) and 
hybridization (strong hybrids and plug-in hybrids) in the baseline 
(relative to the central analysis). The consequence of this shift to 
reliance on lower-level road load technologies is a reduction in 
compliance cost in the baseline of about $300 per vehicle (in MY 2026). 
As a result, cost savings in the preferred alternative are reduced by 
about $200 per vehicle. Under the CO2 program, the general 
trend in technology shift is less dramatic (though the change in BEVs 
is larger) than the CAFE results. The cost change is also comparable, 
but slightly smaller ($200 per vehicle in the baseline) than the CAFE 
program results. Cost savings under the preferred alternative are 
further reduced by about $100. With the lower technology costs in all 
cases, the consumer payback periods decreased as well. These results 
are consistent with the approach taken by manufacturers who have 
already deployed many of the low-level road load reduction 
opportunities to improve fuel economy.
    Figure VI-50 shows the distribution of ROLL technology for the 
Draft TAR, NPRM and final rule. For the NPRM, 64 percent of the MY 2016 
vehicle fleet was assigned ROLL0 and for the final rule, 59 percent of 
the MY2017 vehicle fleet is assigned ROLL0. This shows that the 
majority of the fleet is still at the ROLL0 technology level and there 
is still significant opportunity for the vehicle fleet to improve ROLL 
technology.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.280

BILLING CODE 4910-59-C
c) Rolling Resistance Adoption Features
    In some cases, low rolling resistance tires can affect traction, 
which may adversely impact acceleration, braking and handling 
characteristics for some high-performance vehicles. Similar to past 
rulemakings, the agencies recognized in the NPRM that to maintain 
performance, braking and handling functionality, some high-performance 
vehicles would not adopt low rolling resistance tire technology. For 
cars and SUVs with more than 405 horsepower (hp), the agencies 
restricted the application of ROLL20. For cars and SUVs with more than 
500 hp, the agencies restricted the application of any additional 
rolling resistance technology (ROLL10 or ROLL20). The agencies 
developed these cutoffs based on a review of confidential business 
information and the distribution of rolling resistance values in the 
fleet.
    Ford commented that the NPRM analysis appropriately limited the 
application of ROLL technology where it would be infeasible or would be 
at odds with the vehicles' intended function, characterizing that the 
decision to restrict application of ROLL10 and ROLL20 for high 
performance vehicles as reasonable.\1478\
---------------------------------------------------------------------------

    \1478\ NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    Accordingly, the agencies continued with the NPRM methodology of 
restricting certain ROLL technology for high performance vehicles. In 
the final rule, the agencies restricted the ROLL technology to ROLL0 
and ROLL10 for vehicles with greater than 405 hp and below 505hp. For 
vehicles greater than 505hp, the agencies restricted the ROLL 
technology to ROLL0.
d) Rolling Resistance Effectiveness Modeling and Resulting 
Effectiveness Values
    As discussed above, the agencies updated the baseline rolling 
resistance value to 0.009, based on a thorough review of confidential 
business information submitted by industry, and a review of other 
literature. To achieve ROLL10 in the NPRM and for the final rule 
analysis, the tire rolling resistance must be at least 10 percent 
better than baseline (.0081 or better). To achieve ROLL20, the tire 
rolling resistance must be at least 20 percent better than baseline 
(.0072 or better).
    HDS commented that the Autonomie modeling assumed no engine change 
when drag and rolling resistance reductions were implemented, as well 
as no change to the transmission gear ratios and axle ratios, which 
vary by transmission type but not by the tractive

[[Page 24569]]

load.\1479\ HDS stated that ``reduction in rolling resistance is 
accompanied by axle ratio adjustments so that the engine operates at 
about the same load but at lower RPM. The EPA ALPHA model adjusts for 
this effect, which accounts for the difference in benefit estimates'' 
between Autonomie and the ALPHA model simulations.
---------------------------------------------------------------------------

    \1479\ NHTSA-2018-0067-11985.
---------------------------------------------------------------------------

    As stated in Section VI.B.3 Tech Effectiveness and Modeling, 
Autonomie builds performance-neutral vehicle models by resizing 
engines, electric machines, and hybrid electric vehicle battery packs 
only at specific incremental technology steps. To address product 
complexity and economies of scale, engine resizing is limited to 
specific incremental technology changes that would typically be 
associated with a major vehicle or engine redesign.\1480\ Manufacturers 
have repeatedly told the agencies that the high costs for redesign and 
the increased manufacturing complexity that would result from resizing 
engines for small technology changes preclude them from doing so. It 
would be unreasonable and unaffordable to resize powertrains for every 
unique combination of technologies, and exceedingly so for every unique 
combination technologies across every vehicle model due to the extreme 
manufacturing complexity that would be required to do so. The agencies 
explained in the NPRM that the analysis should not include engine 
resizing with the application of every technology or for combinations 
of technologies that drive small performance changes to reflect better 
what is feasible for manufacturers.\1481\
---------------------------------------------------------------------------

    \1480\ See 83 FR 43027 (Aug. 24, 2018).
    \1481\ For instance, a vehicle would not get a modestly bigger 
engine if the vehicle comes with floor mats, nor would the vehicle 
get a modestly smaller engine without floor mats. This example 
demonstrates small levels of mass reduction. If manufacturers 
resized engines for small changes, manufacturers would have 
dramatically more part complexity, losing economies of scale.
---------------------------------------------------------------------------

    Compliance modeling in the CAFE model also accounts for the 
industry practice of platform, engine, and transmission sharing to 
manage component complexity and associated costs.\1482\ At a vehicle 
refresh cycle, a vehicle may inherit an already resized powertrain from 
another vehicle within the same engine-sharing platform that adopted 
the powertrain in an earlier model year. In the Autonomie modeling, 
when a new vehicle adopts fuel saving technologies (such as ROLL 
technology) that are inherited, the engine is not resized (the 
properties from the baseline reference vehicle are used directly and 
unchanged) and there may be a small change in vehicle performance.
---------------------------------------------------------------------------

    \1482\ Ford EcoBoost Engines are shared across ten different 
models in MY 2019. https://www.ford.com/powertrains/ecoboost/. Last 
accessed Nov. 05, 2019.
---------------------------------------------------------------------------

    Regarding customizing transmission gear ratios as rolling 
resistance changes are implemented, the agencies explained in Section 
VI.C.2 Transmissions that it is an observable practice in industry to 
use a common gear set across multiple platforms and applications. The 
most recent example is the GM 10L90, a 10-speed automatic transmission 
that used the same gear set in both pick-up truck and passenger car 
applications.\1483\ In Autonomie, optimization of transmission 
performance is achieved through shift control logic rather than 
customized hardware (e.g., gear ratios) for each vehicle line. The 
shift initializer routine was run for every unique Autonomie full 
vehicle model configuration to generate customized shifting maps. The 
algorithms' optimization was designed to balance minimization of energy 
consumption against vehicle performance.\1484\ This balance was 
necessary to achieve the best fuel efficiency while maintaining 
customer acceptability by meeting performance neutrality requirements. 
See Section VI.B.3.a)(6) Performance Neutrality for more details. If 
the systems were over-optimized for the agencies' modeling, such as 
applying a unique gear set for each individual vehicle configuration, 
the analysis would likely over-predict the reasonably achievable fuel 
economy improvement for the technology. Over-prediction would be 
exaggerated when applied under real-world large-scale manufacturing 
constraints necessary to achieve the estimated costs for the 
transmission technologies.
---------------------------------------------------------------------------

    \1483\ ``GM Global Propulsion Systems--USA Information Guide 
Model Year 2018'' (PDF). General Motors Powertrain. Retrieved 
September 26, 2019. https://www.gmpowertrain.com/assets/docs/2018R_F3F_Information_Guide_031918.pdf.
    \1484\ See ANL model documentation for final rule.
---------------------------------------------------------------------------

    As HDS noted, the EPA Draft TAR and Proposed Determination analyses 
performed using the ALPHA model adjusted the effectiveness of every 
technology combination assuming performance could be held constant for 
every combination, and did not recognize or account for the extreme 
complexity nor the associated costs for that impractical assumption. 
The NPRM and final rule analyses account for real-world practicalities 
and constraints related to both engine adoption and transmission 
adoption when other vehicle technologies are implemented, which 
explains some of the effectiveness and cost differences between the 
Draft TAR/Proposed Determination and the NPRM/final rule.
    Figure VI-51 below shows the range of effectiveness used for the 
NPRM analysis for ROLL technologies.
BILLING CODE 4910-59-P

[[Page 24570]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.281

BILLING CODE 4910-59-C
    Figure VI-52 below shows the range of effectiveness values used for 
the final rule analysis.
BILLING CODE 4910-59-P

[[Page 24571]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.282

BILLING CODE 4910-59-C
e) Rolling Resistance Cost
    For the NPRM, the analysis used DMC for ROLL technology from the 
Draft TAR and updated the values to reflect 2016$ dollars. The agencies 
continued to use the same cost assumptions presented in the NPRM for 
the final rule, and updated the values to 2018$ dollars. Table VI-139 
and Figure VI-53 show the different levels of tire rolling resistance 
technology cost.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.283


[[Page 24572]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.284

BILLING CODE 4910-59-C
7. Other Vehicle Technologies
    Four other vehicle technologies were included in the analysis--
electric power steering (EPS), improved accessory devices (IACC), low 
drag brakes (LDB), and secondary axle disconnect (SAX) (which may only 
be applied to vehicles with all-wheel-drive or four-wheel-drive). The 
effectiveness of these technologies was applied directly by the CAFE 
model, with unique effectiveness values for each technology and for 
each technology class. This methodology was used in these four cases 
because the effectiveness of these technologies varies little with 
combinations of other technologies. Also, applying these technologies 
directly in the CAFE model significantly reduces the number of 
Autonomie simulations that are needed.
a) Electric Power Steering (EPS)
    Electric power steering reduces fuel consumption and CO2 
emissions by reducing load on the engine. Specifically, it reduces or 
eliminates the parasitic losses associated with engine-driven power 
steering pumps, which pump hydraulic fluid continuously through the 
steering actuation system even when no steering input is present. By 
selectively powering the electric assist only when steering input is 
applied, the power consumption of the system is reduced in comparison 
to the traditional ``always-on'' hydraulic steering system. Power 
steering may be electrified on light duty vehicles with standard 12V 
electrical systems and is also an enabler for vehicle electrification 
because it provides power steering when the engine is off (or when no 
combustion engine is present).
    Power steering systems can be electrified in two ways. 
Manufacturers may choose to eliminate the hydraulic portion of the 
steering system and provide electric-only power steering (EPS) driven 
by an independent electric motor, or they may choose to move the 
hydraulic pump from a belt-driven configuration to a stand-alone 
electrically driven hydraulic pump. The latter system is commonly 
referred to as electro-hydraulic power steering (EHPS). As discussed in 
the NPRM, manufacturers have informed the agencies that full EPS 
systems are being developed for all types of light-duty vehicles, 
including large trucks.
    EPS is also discussed in Section VI.C.3.a) Electrification Modeling 
in the CAFE model.
b) Improved Accessories (IACC)
    Engine accessories typically include the alternator, coolant pump, 
cooling fan, and oil pump, and are traditionally mechanically-driven 
via belts, gears, or directly by other rotating engine components such 
as camshafts or the crankshaft. These can be replaced with improved 
accessories (IACC) which may include high efficiency alternators, 
electrically driven (i.e., on-demand) coolant pumps, electric cooling 
fans, variable geometry oil pumps, and a mild regeneration 
strategy.\1485\ Replacing lower-efficiency and/or mechanically-driven 
components with these improved accessories results in a reduction in 
fuel consumption, as the improved accessories can conserve energy by 
being turned on/off ``on demand'' in some cases, driven at partial load 
as needed, or by operating more efficiently.
---------------------------------------------------------------------------

    \1485\ IACC in this analysis excludes other electrical 
accessories such as electric oil pumps and electrically driven air 
conditioner compressors.
---------------------------------------------------------------------------

    For example, electric coolant pumps and electric powertrain cooling 
fans provide better control of engine cooling. Flow from an electric 
coolant pump can be varied, and the cooling fan can be shut off during 
engine warm-up or cold ambient temperature conditions,

[[Page 24573]]

reducing warm-up time, fuel enrichment requirements, and, ultimately 
reducing parasitic losses.
    IACC is also discussed in Section VI.C.3.a) Electrification 
Modeling in the CAFE model.
c) Low Drag Brakes (LDB)
    Low or zero drag brakes reduce or eliminate brake drag force by 
separating the brake pad from the rotor, either by mechanical or 
electric methods. Conventional disc brake systems are designed such 
that the brake pad is in contact with the brake rotor at all times. 
This is true even when the brakes are not being applied, and although 
the contact pressure is light in this case, this still produces some 
drag force on the vehicle.
    LDBs have historically employed a caliper and rotor system that 
allows the piston in the caliper to retract,\1486\ in turn pulling the 
brake pads away from the rotor. However, if pads are allowed to move 
too far away from the rotor, the first pedal application made by the 
vehicle operator can feel spongy and have excessive travel. This can 
lead to customer dissatisfaction regarding braking performance and 
pedal feel. For this reason, in conventional hydraulic-only brake 
systems, manufacturers are limited by how much they can allow pads to 
move away from the rotor.
---------------------------------------------------------------------------

    \1486\ The brake caliper pistons are used to push the brake pad 
against the brake rotor, or disc.
---------------------------------------------------------------------------

    Recent developments in braking systems have resulted in brakes with 
the potential for zero drag. In these systems, the pedal feel is 
separated from hydraulics by a pedal simulator. This system is similar 
to the brake systems designed for hybrid and electric vehicles, where 
some of the primary braking is done through the recuperation of kinetic 
energy in the drive system. However, the pedal feel and the 
deceleration the operator experiences is tuned to provide a braking 
experience equivalent to that of a conventional hydraulic brake system. 
These ``brake-by-wire'' systems have highly tuned pedal simulators that 
feel like typical hydraulic brakes and seamlessly transition to a 
conventional system as required by different braking conditions. The 
application of a pedal simulator and brake-by-wire system is new to 
non-electrified vehicle applications. By using this type of system, 
vehicle manufacturers can allow brake pads to move farther away from 
the rotor and still maintain the initial pedal feel and deceleration 
associated with a conventional brake system.
    In addition to reducing brake drag, the zero drag brake system 
provides ancillary benefits. It allows for a faster brake application 
and greater deceleration than is normally applied by the average 
vehicle operator. It also allows manufacturers to tune the braking for 
different customer preferences within the same vehicle. This means 
manufacturers can provide a ``sport'' mode, which provides greater 
deceleration with less pedal displacement and a ``normal'' mode, which 
might be more appropriate for day-to-day driving.
    The zero drag brake system also eliminates the need for a brake 
booster. This saves cost and weight in the system. Elimination of the 
conventional vacuum brake booster could also improve the effectiveness 
of stop-start systems. Typical stop-start systems need to restart the 
engine if the brake pedal is cycled because the action drains the 
vacuum stored in the booster. Because the zero drag brake system 
provides braking assistance electrically, there is no need to 
supplement lost vacuum during an engine off event.
    Finally, many engine technologies being considered to improve 
efficiency also reduce pumping losses through reduced throttling, and 
in turn there is less engine vacuum available to power-assist a 
conventional brake system. The reduction in throttling could require a 
supplemental vacuum pump to provide vacuum for a conventional brake 
system. This is the situation in many diesel-powered vehicles. Diesel 
engines have no throttling and require a supplemental vacuum for 
conventional brake systems. A zero drag brake system both eliminates 
brake drag and avoids the need for a supplemental vacuum pump.
d) Secondary Axle Disconnect (SAX)
    All-wheel drive (AWD) and four-wheel drive (4WD) vehicles provide 
improved traction by delivering torque to the front and rear axles, 
rather than just one axle. When a second axle is rotating, it tends to 
consume more energy because of additional losses related to lubricant 
churning, seal friction, bearing friction, and gear train 
inefficiencies.\1487\ \1488\ Some of these losses may be reduced by 
providing a secondary axle disconnect function that disconnects one of 
the axles when driving conditions do not call for torque to be 
delivered to both.
---------------------------------------------------------------------------

    \1487\ Phelps, P. ``EcoTrac Disconnecting AWD System,'' 
presented at 7th International CTI Symposium North America 2013, 
Rochester MI.
    \1488\ Pilot Systems, ``AWD Component Analysis,'' Project 
Report, performed for Transport Canada, Contract T8080-150132, May 
31, 2016.
---------------------------------------------------------------------------

    The terms AWD and 4WD are often used interchangeably, although they 
have also developed a colloquial distinction, and are two separate 
systems. The term AWD has come to be associated with light-duty 
passenger vehicles providing variable operation of one or both axles on 
ordinary roads. The term 4WD is often associated with larger truck-
based vehicle platforms providing a locked driveline configuration and/
or a low range gearing meant primarily for off-road use.
    Many 4WD vehicles provide for a single-axle (or two-wheel) drive 
mode that may be manually selected by the user. In this mode, a primary 
axle (usually the rear axle) will be powered, while the other axle 
(known as the secondary axle) is not. However, even though the 
secondary axle and associated driveline components are not receiving 
engine power, they are still connected to the non-driven wheels and 
will rotate when the vehicle is in motion. This unnecessary rotation 
consumes energy,\1489\ and leads to increased fuel consumption and CO2 
emissions that could be avoided if the secondary axle components were 
completely disconnected and not rotating.
---------------------------------------------------------------------------

    \1489\ Any time a drivetrain component spins it consumes some 
energy, primarily to overcome frictional forces.
---------------------------------------------------------------------------

    Light-duty AWD systems are often designed to divide variably torque 
between the front and rear axles in normal driving to optimize traction 
and handling in response to driving conditions. However, even when the 
secondary axle is not necessary for enhanced traction or handling, in 
traditional AWD systems it typically remains engaged with the driveline 
and continues to generate losses that could be avoided if the axle was 
instead disconnected. The SAX technology observed in the marketplace 
disengages one axle (typically the rear axle) for 2WD operation, but 
detects changes in driving conditions and automatically engages AWD 
mode when it is necessary. The operation in 2WD can result in reduced 
fuel consumption. For example, Chrysler has estimated the secondary 
axle disconnect feature in the Jeep Cherokee reduces friction and drag 
attributable to the secondary axle by 80% when in disconnect 
mode.\1490\
---------------------------------------------------------------------------

    \1490\ Brooke, L. ``Systems Engineering a new 4x4 benchmark,'' 
SAE Automotive Engineering, June 2, 2014.
---------------------------------------------------------------------------

e) Analysis Fleet Assignments for Other Vehicle Technologies
    The agencies described in the PRIA that the aforementioned 
technologies have been applied, to some extent, in the MY 2016 fleet. 
However, these technologies are difficult to observe and

[[Page 24574]]

assign to the analysis fleet, and the agencies relied heavily on 
industry engagement and feedback to assign the technologies properly to 
the NPRM analysis fleet vehicles. In the NPRM, the agencies noted that 
the Draft TAR analysis did not properly account for the presence of 
these technologies in the analysis fleet, and far too few were 
assigned. Accordingly, the NPRM analysis reflected higher EPS and IACC 
application rates than the Draft TAR analysis.
    The agencies received a handful of comments stating that the 
additional technologies were incorrectly applied to the analysis fleet. 
ICCT stated that the inclusion of EPS, IACC, and LDB in the analysis 
fleet was unsubstantiated, and removed the technologies from potential 
use during the subsequent simulated years.\1491\ ACEEE commented that 
IACC should not have been applied to certain vehicles in the analysis 
fleet because those vehicles do not in actuality display the fuel 
consumption reduction that would confirm the presence of these 
additional technologies.\1492\ In addition, ACEEE commented that the 
CAFE model assumes significant baseline SAX penetration that they could 
not corroborate from Ford F-150 product information brochures.\1493\ 
HDS compared the available levels of IACC improvements from the Draft 
TAR to the NPRM analysis, noting that the NPRM only employed one level 
of improved accessory technologies.\1494\ HDS stated that this implied 
the effectiveness of what was previously considered IACC1 (the first 
level of IACC technology improvement available in the Draft TAR) was 
completely used up in the 2016 analysis fleet for this rule.
---------------------------------------------------------------------------

    \1491\ International Council on Clean Transportation, Attachment 
3, Docket No. NHTSA-2018-0067-11741, at I-37.
    \1492\ American Council for an Energy-Efficient Economy, 
Attachment 6, Docket No. NHTSA-2018-0067-12122, at 6.
    \1493\ American Council for an Energy-Efficient Economy, 
Attachment 6, Docket No. NHTSA-2018-0067-12122, at 7.
    \1494\ H-D Systems, ``HDS final report,'' Docket No. NHTSA-2018-
0067-11985, at 21.
---------------------------------------------------------------------------

    As the agencies stated in the PRIA, in part because of the 
difficulty in observing EPS, IACC, LDB, and SAX on actual vehicles, far 
too few of those technologies were assigned to vehicles in the Draft 
TAR analysis fleets. For the final rule, each vehicle in the MY 2017 
analysis fleet was studied using confidential and publicly available 
information to determine whether, as commenters suggested, the agencies 
had improperly applied any of these additional vehicle technologies. 
This resulted in some adjustments in the application of the 
technologies in the analysis fleet. In regard to ACEEE's comment on SAX 
penetration in the analysis fleet, for the NPRM and final rule 
analysis, the agencies considered all 4WD vehicles to have the 
capability manually to disconnect either the front or rear wheel axle 
and associated rotating components, thus shifting to a 2WD mode. When 
4WD operation is required for safety and utility, the consumer can 
enable this feature. As stated above, this capacity to shift between 
2WD and 4WD modes is another form of SAX. For AWD vehicles, publicly 
available manufacturer information was reviewed to identify the 
specific vehicles that have SAX technology. Based on market 
observations and feedback from OEMs, the entire analysis fleet for NPRM 
and the final rule was considered to have a basic level of improved 
accessories (comparable to what Draft TAR referred to as IACC1). The 
application of IACC in the NPRM and final rule analysis fleets 
represents further improvements to accessories such as electric water 
pumps and higher efficiency alternators with mild regeneration 
capacity.
    The following distribution of technologies in the analysis fleet 
from the NPRM to the final rule analysis shows a slight decrease in the 
portion of total vehicles produced that have EPS and IACC, a very 
slight increase in the portion of total vehicle production that have 
LDB, and a slight increase in the portion of 4WD/AWD vehicles with SAX 
technology.
BILLING CODE 4910-59-P

[[Page 24575]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.285

BILLING CODE 4910-59-C
f) Effectiveness Estimates for Other Vehicle Technologies
    The effectiveness estimates for these four technologies rely on 
previous work published as part of the rulemaking process, both for the 
2012 rule for MYs 2017-2025 and the Draft TAR. The effectiveness values 
are unchanged from the Draft TAR.
    The effectiveness of both EPS and EHPS is derived from the 
decoupling of the pump from the crankshaft, and is considered to be 
practically the same for both. Thus, a single effectiveness value is 
assigned to all vehicles in the analysis fleet that possess either EPS 
or EHPS, and the ``EPS'' designation is applied.
    For the Draft TAR analysis, two levels of IACC were offered as a 
technology path (a low improvement level and a high improvement level). 
Since much of the market has incorporated some of these technologies in 
the baseline MY 2016 and 2017 fleets, the NPRM and final rule analyses 
assumed all vehicles have incorporated what was previously the low 
level, so only the high level remained as an option for vehicles. The 
figure above shows the distribution of IACC for NPRM and FRM, which is 
the equivalent type of technology as the high-level IACC in the DRAFT 
TAR.
    The NPRM analysis carried forward work on the effectiveness of SAX 
systems conducted in the Draft TAR and EPA Proposed Determination. This 
work involved gathering information by monitoring press reports, 
holding meetings with suppliers and OEMs, and attending industry 
technical conferences. The resulting effectiveness estimates used in 
the Draft TAR, NPRM, and this final rule are shown below.
BILLING CODE 4910-59-P

[[Page 24576]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.286

BILLING CODE 4910-59-C
g) Cost Estimates and Learning Rates for Other Vehicle Technologies
    The cost estimates for these technologies rely on previous work 
published as part of the rulemaking process, both for the 2012 rule for 
MYs 2017-2027 and the Draft TAR. The cost values are from the same 
sources as the Draft TAR and were updated to 2016 dollars for the NPRM 
and 2018 dollars for the final rule analysis. Learning rates for these 
technologies are also unchanged since the NPRM, and can be seen in 
Section VI.B.4.d)(4) Cost Learning as Applied in the CAFE Model.
    CARB noted that the IACC costs in Tables 6-32 and 6-33 of the PRIA 
did not align with the Technologies central analysis input file.\1495\ 
HDS commented, as part of its comparison of IACC penetration in the 
analysis fleet from the Draft TAR to NPRM, that IACC costs were based 
on the difference between IACC1 and IACC2 costs and this appeared to be 
inconsistent with the cost of accessory electrification which is more 
expensive.\1496\
---------------------------------------------------------------------------

    \1495\ CARB, Docket No. NHTSA-2018-0067-12428, at 21.
    \1496\ H-D Systems, ``HDS final report,'' Docket No. NHTSA-2018-
0067-11985, at 21.
---------------------------------------------------------------------------

    In the PRIA, the cost of IACC was reported in some tables as an 
absolute cost (the cost of adding IACC to a base vehicle), while the 
NPRM Technologies central analysis input file showed IACC cost 
incremental to EPS. This was necessary in the model input file because 
the accounting method of the NPRM CAFE model utilized incremental 
costs. In contrast, a change in the CAFE model accounting method for 
this final rule allows all costs in the input file to be reported as 
absolute costs, incremental to a base vehicle. It was assumed that EPS 
must be present on a vehicle in order for it to adopt IACC, and as such 
the cost of IACC includes the cost of EPS. For further detail on the 
use of absolute costs in place of incremental costs, see Section 
VI.C.7.g). Although HDS commented that accessory electrification has a 
higher cost than what is being used in the analysis, no specific 
additional input was given; the cost of IACC, as was done for Draft TAR 
(where it was referred to as IACC2), was taken from the 2015 NAS 
Report.\1497\
---------------------------------------------------------------------------

    \1497\ National Research Council. 2015. Cost, Effectiveness, and 
Deployment of Fuel Economy Technologies for Light-Duty Vehicles. 
Washington, DC--The National Academies Press, Table 8A.2a, available 
at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles.
---------------------------------------------------------------------------

    Table VI-141 below shows the absolute costs for these technologies 
for select model years. The FRM Technologies central analysis input 
file shows the costs for all model years.

[[Page 24577]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.287

8. Simulating Off-Cycle and A/C Efficiency Technology Adjustments
    Off-cycle and air conditioning (A/C) efficiency technologies can 
provide fuel economy improvements in real-world vehicle operation, but 
that benefit cannot be adequately captured by the 2-cycle test 
procedures used to demonstrate compliance with fuel economy and 
CO2 emissions standards.\1498\ Off-cycle technologies 
include technologies like high efficiency alternators and high 
efficiency exterior lighting.\1499\ A/C efficiency technologies operate 
mainly by reducing the operation of the compressor, which pumps A/C 
refrigerant around the system loop. The less the compressor operates or 
the more efficiently it operates, the less load the compressor places 
on the engine, resulting in better fuel efficiency and lower 
CO2 emissions.
---------------------------------------------------------------------------

    \1498\ See 49 U.S.C 32904(c) (``The Administrator shall measure 
fuel economy for each model and calculate average fuel economy for a 
manufacturer under testing and calculation procedures prescribed by 
the Administrator. . . . the Administrator shall use the same 
procedures for passenger automobiles the Administrator used for 
model year 1975 (weighted 55 percent urban cycle and 45 percent 
highway cycle), or procedures that give comparable results.'').
    \1499\ See 83 FR 43057. A partial list of off-cycle technologies 
is included in Tables II-21 and II-22 of the NPRM.
---------------------------------------------------------------------------

    Vehicle manufacturers have the option to generate credits for off-
cycle technologies and improved A/C systems under the EPA's 
CO2 program and receive a fuel consumption improvement value 
(FCIV) equal to the value of the benefit not captured on the 2-cycle 
test under NHTSA's CAFE program. The FCIV is not a credit in the NHTSA 
CAFE program, but the FCIVs increase the reported fuel economy of a 
manufacturer's fleet, which is used to determine compliance. EPA 
applies FCIVs during determination of a fleet's final average fuel 
economy reported to NHTSA.\1500\ FCIVs are only calculated and applied 
at a fleet level for a manufacturer and are based on the volume of the 
manufacturer's fleet that contain qualifying technologies.\1501\
---------------------------------------------------------------------------

    \1500\ 49 U.S.C. 32904(c)-(e). EPCA granted EPA authority to 
establish fuel economy testing and calculation procedures. See 
Section IX for more information.
    \1501\ 40 CFR 600.510-12(c)
---------------------------------------------------------------------------

    As discussed further in Section IX.D Compliance Issues that Affect 
Both the CO2 and CAFE Programs, three pathways can be used 
to determine the value of A/C efficiency and off-cycle adjustments. 
First, manufacturers can use a predetermined list or ``menu'' of credit 
values established by EPA for specific off-cycle technologies.\1502\ 
Second, manufacturers can use 5-cycle testing to demonstrate and 
justify off-cycle CO2 credits; \1503\ the additional tests 
allow emission benefits to be demonstrated over some elements of real-
world driving not captured by the 2-cycle compliance tests, including 
high speeds, rapid accelerations, and cold temperatures. Third, 
manufacturers can seek EPA approval, through a notice and comment 
process, to use an alternative methodology other than the menu or 5-
cycle methodology for determining the off-cycle technology improvement 
values.\1504\
---------------------------------------------------------------------------

    \1502\ See 40 CFR 86.1869-12(b). The Technical Support Document 
(TSD) for the 2012 final rule for MYs 2017 and beyond provides 
technology examples and guidance with respect to the potential 
pathways to achieve the desired physical impact of a specific off-
cycle technology from the menu and provides the foundation for the 
analysis justifying the credits provided by the menu. The 
expectation is that manufacturers will use the information in the 
TSD to design and implement off-cycle technologies that meet or 
exceed those expectations in order to achieve the real-world 
benefits of off-cycle technologies from the menu.
    \1503\ See 40 CFR 86.1869-12(c). EPA proposed a correction for 
the 5-cycle pathway in a separate technical amendments rulemaking. 
See 83 FR 49344 (Oct. 1, 2019). EPA is not approving credits based 
on the 5-cycle pathway pending the finalization of the technical 
amendments rule.
    \1504\ See 40 CFR 86.1869-12(d).
---------------------------------------------------------------------------

    The agencies have been collecting data on the application of these 
technologies since implementing the programs.\1505\ Most manufacturers 
are generating A/C efficiency and off-cycle credits; in MY 2017, 15 
manufacturers generated A/C efficiency credits and 15 manufacturers 
generated off-cycle credits, through the level of deployment varies by 
manufacturer.\1506\
---------------------------------------------------------------------------

    \1505\ See 77 FR at 62832, 62839 (Oct. 15, 2012). EPA introduced 
A/C and off-cycle technology credits for the CO2 program 
in the MYs 2012-2016 rule and revised the program in the MY 2017-
2025 rule and NHTSA adopted equivalent provisions for MYs 2017 and 
later in the MY 2017-2025 rule.
    \1506\ The 2018 EPA Automotive Trends Report, EPA-420-R-19-002, 
March 2019 at Chapter 5.B., Figures 5.10 and 5.11.
---------------------------------------------------------------------------

a) A/C and Off-Cycle Effectiveness Modeling
    The NPRM analysis used the off-cycle FCIVs and credits earned by 
each manufacturer in MY 2016 and carried these forward at the same 
levels for future years for the CO2 analysis and beginning 
in MY 2017 for the CAFE analysis. The 2016 values for off-cycle FCIVs 
for each manufacturer and fleet, denominated in grams CO2 
per mile,\1507\ are provided in Table VI-142.\1508\ Additional off-
cycle FCIVs were added in future years if a manufacturer applied a 
technology that was explicitly simulated in the analysis and also was 
an off-cycle technology listed on the predefined menu.\1509\ 
Technologies explicitly simulated in the analysis that are also on the 
off-cycle menu include start-stop systems that reduce fuel consumption 
during idle and active grille shutters that improve aerodynamic drag at 
highway speeds,

[[Page 24578]]

among others. Any off-cycle adjustments that accrued as the result of 
applying these technologies were calculated dynamically in each model 
year the technology was applied, with adjustments accumulating up to 
the 10 g/mi cap. As a practical matter, most of the adjustments for 
which manufacturers can claim off-cycle FCIVs exist outside of the CAFE 
model technology tree so the off-cycle menu cap was rarely reached for 
the NPRM analysis.
---------------------------------------------------------------------------

    \1507\ For the purpose of estimating their contribution to CAFE 
compliance, the grams CO2/mile values in Table I-1 are 
converted to gallons/mile and applied to a manufacturer's 2-cycle 
CAFE performance. When calculating compliance with EPA's 
CO2 program, there is no conversion necessary (as 
standards are also denominated in grams/mile).
    \1508\ 2016 GHG Manufacturer Performance Report. EPA-420-R-18-
002. January 2018. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100TGIA.pdf. Last Accessed Nov. 14, 2019. 2016 
Report Tables for the GHG Manufacturer Performance Report. January 
2018. https://www.epa.gov/sites/production/files/2018-01/ghg-report-2016-data-tables.xlsx. Last Accessed Nov. 14, 2019.
    \1509\ For more details, see Section IX.D Compliance Issues that 
Affect Both the CO2 and CAFE Programs and Section IX.D.3 
Flexibilities for Off-Cycle Technologies.
---------------------------------------------------------------------------

    The agencies sought comment on both the A/C and off-cycle data that 
was used for the NPRM analysis as well as the assumptions for applying 
those technologies.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.288

BILLING CODE 4910-59-C
    Universally, stakeholders believed the application of off-cycle 
adjustments in the analysis was too conservative. Stakeholders believed 
the A/C and off-cycle technologies would be rapidly deployed and 
manufacturers would reach the cap values within the rulemaking 
timeframe.
---------------------------------------------------------------------------

    \1510\ See 83 FR 43159-60 (``. . . this analysis uses the off-
cycle credits submitted by each manufacturer for MY 2017 compliance 
and carries these forward to future years with a few exceptions.'').
---------------------------------------------------------------------------

    The Institute for Policy Integrity (IPI) questioned the position 
the agencies assumed in the NPRM analysis, and suggested the agencies 
``assume that manufacturers will efficiently deploy all cost-saving 
offset opportunities, especially in the face of increasingly stringent 
standards.'' \1511\
---------------------------------------------------------------------------

    \1511\ Comments from Institute from Policy Integrity, Attachment 
1, NPRM Docket No. NHTSA-2018-0067-12213, at 20-21.
---------------------------------------------------------------------------

    ICCT stated ``far greater use of the off-cycle provisions will 
occur by 2025'' and emphasized that off-cycle technologies are ``highly 
cost-effective and being deployed in greater sales penetrations than 
many of the test-cycle efficiency technologies that the agencies are 
analyzing.'' \1512\ ICCT supported manufacturers maximizing the use of 
off-cycle technologies, and supported the analysis estimating 
``fleetwide off-cycle credit use at over 10 g/mile by 2020,'' and 
further suggested fleetwide achievement of 15 g/mile by 2025.\1513\
---------------------------------------------------------------------------

    \1512\ Comments from ICCT, Attachment 1, NPRM Docket No. NHTSA-
2018-0067-11741, at I40--I41.
    \1513\ Note there is a regulatory ``cap'' on menu technologies 
of 10 g/mi (see Section IX for further discussion of the cap), 
however a manufacturer can receive additional off-cycle credit/FCIV 
by using the pathways described above to petition for off-menu 
technologies. ICCT's comment suggests that manufacturers will reach 
the regulatory menu cap and apply additional technologies to get an 
additional 5 g/mi credit above the menu cap.
---------------------------------------------------------------------------

    FCA, General Motors and the Auto Alliance all provided similar 
observations, stating ``[m]anufacturers have rapidly deployed 
technology in response to this all new regulatory

[[Page 24579]]

mechanism.'' Each of the commenters provided support for an argument of 
rapid off-cycle technology adoption, stating ``[i]n the MY2021-2026 
timeframe of the proposed rule, it is likely that manufacturers will 
hit the existing 10 g/mi cap.'' \1514\
---------------------------------------------------------------------------

    \1514\ Comments from Automotive Alliance, Appendix 1, NPRM 
Docket No. NHTSA-2018-0067-12073, at 92; Comments from Fiat Chrysler 
Automobiles, Attachment1, NPRM Docket No. NHTSA-2018-0067-11943, at 
8; Comments from General Motors, Appendix 4--Comments to Technical 
Issues, NPRM Docket No. NHTSA-2018-0067-11858, at 1.
---------------------------------------------------------------------------

    The DENSO Corporation further supported the increased use of off-
cycle technologies, commenting that ``[a]vailable data on OEM off-cycle 
technology credit utilization within the past few years demonstrates 
that the use of off-cycle technologies is expected to grow--
particularly technologies on the credit menus.'' \1515\
---------------------------------------------------------------------------

    \1515\ Comments from DENSO Corporation, Attachment 1, NPRM 
Docket No. NHTSA-2018-0067-11880, at 6.
---------------------------------------------------------------------------

    However, Toyota Motors North America asked for constraints on 
considerations of off-cycle technology in the analysis.\1516\ Toyota 
expressed concern for over-reliance on off-cycle technologies to 
provide flexibilities for compliance, as ``most of the technologies 
provide little tangible value proposition for customers.'' In 
additional comments, Toyota repeated the concern noting, ``most of 
these technologies lack consumer demand.'' Finally, Toyota specifically 
cautioned against overusing off-cycle technologies in the analysis, 
stating ``[t]he suggested pursuit of maximum credits overlooks the 
associated costs and market acceptance challenge for certain off-cycle 
technologies.'' Toyota listed costs versus risk of customer acceptance 
and agency approval as factors that ``introduce a high level of 
uncertainty for an auto manufacturer's planning and make investments in 
off-cycle technologies risky and less appealing.''
---------------------------------------------------------------------------

    \1516\ Comments from Toyota Motors North America, Attachment 1, 
NHTSA Docket No. NHTSA-2018-0067-130798, at 9-10; Supplemental 
Comments from Toyota Motors North America, Attachment 1, NHTSA 
Docket No. NHTSA-2018-0067-12150, at 24; Supplemental Comments from 
Toyota Motors North America, Attachment 1, NHTSA Docket No. NHTSA-
2018-0067-12376, at 4-5.
---------------------------------------------------------------------------

    After carefully considering the comments, the agencies agree that 
A/C and off-cycle technologies are likely to be more broadly applied by 
manufacturers within the rulemaking timeframe. The final rule analysis 
has been updated to reflect an increased application of the 
technologies. Similar to the NPRM, the final rule analysis used the A/C 
and off-cycle FCIVs earned by each manufacturer in the baseline fleet 
(MY 2017 for the final rule analysis) as a starting point. However, the 
final rule analysis increased these values in subsequent model years. 
In addition to the dynamic application of off-cycle FCIVs, as in the 
NPRM, each manufacturer's fleet FCIVs were increased by extrapolating 
the manufacturers' historical rate of FCIV application through 
2017.\1517\ In line with most commenters, the agencies increased the 
FCIVs for each manufacturer such that the maximum value of 10 g/mi will 
be reached by MY 2023. For manufacturers who did not reach maximum 
values prior to 2023 through data extrapolation, a linear increase to 
the cap was assumed. The agencies believe this approach balances a 
greater application of FCIV technologies across the fleet, while 
avoiding uncertain over-reliance on flexibilities for the analysis.
---------------------------------------------------------------------------

    \1517\ The 2018 EPA Automotive Trends Report, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends. Accessed Aug 23, 2019.
---------------------------------------------------------------------------

    The agencies disagreed with the proposal to model the application 
of 15 g/mi of FCIVs universally in the rulemaking timeframe. Based on 
historical data and industry comments from both manufacturers and 
suppliers, the agencies expect there will be an increase in off-cycle 
technology application. However, there are two issues with assuming 
manufacturers will exceed the existing off-cycle caps. First, only a 
few manufacturers approached the cap limit in MY 2018, and the fleet 
average menu credit was 4.7 grams/mile, less than half the cap 
value.\1518\ Second, new off-cycle technologies may address the same 
inefficiencies as menu technologies, rather than work in conjunction. 
Accordingly, the agencies believe there is a reasonable basis for 
assuming manufacturers could, and would only achieve 10 g/mi on average 
by MY 2023, and used that assumption for the final rule analysis.
---------------------------------------------------------------------------

    \1518\ The 2018 EPA Automotive Trends Report, Greenhouse Gas 
Emissions, Fuel Economy, and Technology since 1975, EPA-420-R-19-002 
(Mar. 2019).
---------------------------------------------------------------------------

    Table VI-143 shows passenger car values for FCIVs and Table VI-144 
shows light truck values for FCIVs applied for the final rule analysis.
BILLING CODE 4910-59-P

[[Page 24580]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.289


[[Page 24581]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.290

[GRAPHIC] [TIFF OMITTED] TR30AP20.291


[[Page 24582]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.292


[[Page 24583]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.293

BILLING CODE 4910-59-C

[[Page 24584]]

A/C Efficiency, A/C Leakage and Off-Cycle Costs
    As discussed above, the only A/C efficiency and off-cycle 
technologies applied dynamically in the NPRM analysis were explicitly 
simulated technologies like stop-start systems and active aerodynamic 
technologies. The NPRM analysis fully accounted for both the 
effectiveness and cost of these technologies and therefore separate 
cost accounting was not needed. For example, when stop-start or active 
aerodynamics technology was added by the model to a vehicle, the 
corresponding off-cycle FCIVs were applied and the technology costs 
were captured the same as every other technology on the decision trees.
    For the final rule analysis, A/C and off-cycle technologies are 
applied independently of the decision trees using the extrapolated 
values, so it is necessary to account for the costs of those 
technologies independently. Table VI-145 shows the costs used for A/C 
and off-cycle FCIVs the final rule analysis. The costs are shown in 
dollars per gram of CO2 per mile ($ per g/mile). The A/C 
costs and off-cycle technology costs are the same costs used in the EPA 
Proposed Determination and described in the EPA Proposed Determination 
TSD.\1519\
---------------------------------------------------------------------------

    \1519\ EPA PD TSD. EPA-420-R-16-021. November 2016. At 2-423-2-
245. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf. Last 
accessed Nov.14, 2019.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.294

BILLING CODE 4910-59-C

D. Impacts that Result From Simulating Manufacturer Compliance with 
Regulatory Alternatives

1. Simulating Economic Impacts of Regulatory Alternatives
a) What Economic Impacts Occur When Vehicle Manufacturers Comply With 
Different CAFE and CO2 Standards?
1) The NPRM Framework for Analyzing Economic Impacts
    In the proposed rule, the agencies noted the importance of 
identifying the mechanisms by which vehicle manufacturers' compliance 
with different CAFE and CO2 standards generated impacts on 
manufacturers, owners of new and used vehicles, and the remainder of 
the U.S. The agencies organized the analysis of alternative standards 
using a framework that clarified the economic impacts on vehicle 
producers, illustrated how costs were transmitted to buyers of new 
vehicles, highlighted the collateral economic effects on owners of used 
vehicles, and identified how these responses created various indirect 
costs and benefits. Throughout the analysis, the agencies stressed the 
distinction between the proposal's economic consequences for private 
businesses and households, and its ``external'' economic impacts--those 
ultimately borne by the rest of the U.S. economy.
    To clarify the framework used in the proposal, the agencies used 
Table VI-146 below (which is based on Tables II-25 to II-28 from the 
NPRM) \1520\ to report costs and benefits and to trace how they pass 
through the economy. As the table shows, the economic impacts of 
standards initially fall on vehicle manufactures, but ultimately are 
borne by consumers who purchase and drive new models. Smaller, indirect 
economic effects of the proposal would be borne by owners of used cars 
and light trucks (vehicles produced during model years prior to those 
affected by the proposal, but still in use) as well as by the general 
public and government agencies. On balance, the agencies projected that 
most of the proposal's economic effects would fall on private 
businesses and households, with the remainder of the U.S. economy 
bearing much smaller impacts.
---------------------------------------------------------------------------

    \1520\ See 83 FR at 43062-66.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24585]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.295

BILLING CODE 4910-59-C
    More specifically, the agencies' analysis showed that the proposal 
would initially have saved manufacturers the costs of adding the 
technologies that would otherwise have been necessary to enable their 
new cars and light trucks to comply with the baseline fuel economy and 
CO2 emissions regulations, with the estimated dollar value 
of those savings shown in line 1 of Table VI-146. The proposal also 
enabled some manufacturers to make lower civil penalty payments for 
failing to comply with the more demanding standards that were 
supplanted (line 2), although these savings would have been exactly 
offset by lower civil penalty revenue to the

[[Page 24586]]

Federal Government (line 16). The analysis assumed that manufacturers 
would have the ability, in a competitive market, to pass their savings 
in technology costs and any reduction in civil penalties paid on to 
buyers, by charging lower prices for new vehicles. Although lower 
prices reduced their revenues (line 3), on balance, their savings in 
compliance costs, reduced civil penalty payments, and lower sales 
revenue were assumed to leave manufacturers financially unaffected 
(shown by the zero entry in line 4 of the table).
    Under the proposal, the analysis showed that buyers of new cars and 
light trucks benefited directly from those vehicles' lower purchase 
prices and financing costs (line 5). They also avoided the increased 
risk of crash-related injuries that would have resulted from reductions 
in the weight of some new models, as manufacturers attempted to improve 
fuel economy to comply with the baseline standards. The economic value 
of this reduction in risk represented an additional benefit from the 
proposal to reducing the stringency of the standards vis-[agrave]-vis 
the baseline (line 6).
    At the same time, however, the lower fuel economy that some new 
cars and light trucks were expected to offer with less stringent 
standards in place would have imposed various additional costs on their 
buyers and users. Drivers experienced higher fuel costs as a 
consequence of new vehicles' increased fuel consumption (line 7), as 
well as the added time and inconvenience of having to make more 
frequent refueling stops required by reduced driving range (line 8). 
They also forfeited some mobility benefits as they drove newly-
purchased cars and light trucks less in response to their higher fuel 
costs (line 9). On balance, the agencies' analysis of the proposal 
showed that buyers of new cars and light trucks produced during the 
model years it affected would experience significant economic benefits 
(line 10).
    A novel feature of the agencies' evaluation of the proposal showed 
that lowering prices for new cars and light trucks, some owners of used 
vehicles retired them from service earlier than they otherwise would 
have done. In combination with increased sales of new models, this 
transferred some driving that would have occurred with used cars and 
light trucks to newer and safer models, thus reducing the total costs 
of fatalities and injuries sustained in motor vehicle crashes.\1521\ In 
the proposal, this reduction in injury risks provided benefits to 
owners and drivers of older cars and light trucks that had not been 
recognized or quantified in its analyses of previous CAFE and 
CO2 standards (line 11).
---------------------------------------------------------------------------

    \1521\ This improvement in safety resulted from the fact that 
cars and light trucks have become progressively more protective in 
crashes over time (and also slightly less prone to certain types of 
crashes, such as rollovers). Thus, shifting some travel from older 
to newer models reduced injuries and damages sustained by drivers 
and passengers because they were traveling in inherently safer 
vehicles, rather than because of changes to driver risk profiles.
---------------------------------------------------------------------------

    Table VI-146 also showed that the changes in fuel consumption and 
vehicle use resulting from the proposal would in turn generate both 
benefits and costs to the remainder of the U.S. economy. The analysis 
described these as ``external'' effects, in the sense that they were 
by-products of households' choices among new vehicle models, decisions 
about keeping older cars and light trucks in service, and allocations 
of driving across the fleet that were experienced broadly throughout 
the U.S. economy, rather than by the individuals making such decisions. 
The largest of these was additional refining and consumption of 
petroleum-based fuel and the associated increases in emissions of 
carbon dioxide and other gases, which were projected to increase the 
cost of economic damages inflicted on the U.S. economy by future 
changes in the global climate (line 13). Added fuel production and use 
under the proposal also led to higher emissions of localized air 
pollutants, and the resulting increase in the U.S. population's 
exposure and its adverse effects on health imposed additional external 
costs (line 14).
    Increased consumption of petroleum-derived fuel also imposed higher 
external costs on the U.S. economy, in the form of potential losses in 
economic output and costs to businesses and households for adjusting to 
any sudden changes in energy prices (line 15 of the table). Reduced 
driving by buyers of new cars and light trucks in response to their 
higher operating costs also reduced the external costs from their 
contributions to traffic delays and noise, benefits that were expected 
to be experienced throughout the U.S. economy (line 17). Finally, some 
of the higher fuel costs to buyers of new cars and light trucks will 
consist of increased fuel taxes; this increase in revenue was projected 
to enable Federal and State government agencies to improve upkeep of 
roads and highways, fund increases in other services, or reduce other 
tax burdens (line 18).\1522\
---------------------------------------------------------------------------

    \1522\ In some States, levies on gasoline include both general 
sales taxes as well as excise taxes, and not all proceeds are 
dedicated to transportation purposes.
---------------------------------------------------------------------------

    The net economic effect (line 22) of the proposal consisted of the 
benefits and costs imposed directly on car and light truck 
manufacturers, accompanying indirect effects on buyers of new vehicles 
and owners of used ones, external costs driving decisions generated 
throughout the U.S. economy, and changes in revenue to government 
agencies. The agencies' organization was intended to convey the causal 
connections among these impacts, by highlighting how the proposed 
change in fuel economy standards faced by manufacturers would set in 
motion the sequence of behavioral responses that determined its 
economy-wide costs and benefits. This contrasted with the way benefits 
and costs of previous proposals to establish CAFE and CO2 
standards were analyzed and presented, which obscured their sequence 
and causal connections.
    In those previous analyses, most economic effects other than 
manufacturers' costs to comply with proposed standards and anticipated 
changes in fuel consumption were grouped together and reported as ``co-
benefits.'' This obscured how these various consequences arose from the 
proposed standards, providing no information about who would ultimately 
experience the costs of complying with the standards, or who would 
experience their direct and indirect benefits. In contrast, the recent 
analysis spelled out how each category of benefits and costs resulted 
from the proposed change in standards, identified the mechanisms that 
translated direct economic impacts into indirect costs and benefits, 
and distinguished between those arising from changes in fuel 
consumption, and safety consequences of changes in vehicle use. The 
proposal's framework also clarified who would bear each category of 
impacts, distinguishing between the proposal's economic impacts on 
private actors--vehicle manufacturers, new car and light truck buyers, 
and owners of used vehicles--and the external economic consequences for 
the general public and government agencies that stem indirectly from 
such private impacts.
2) Final Rule Framework
    While the agencies received several comments about which economic 
effects are included in the analysis, the agencies received no comments 
about the specific structure of the framework. Substantive comments 
about individual

[[Page 24587]]

effects are addressed over the next several sections.
    The agencies have expanded the accounting framework for benefits 
and costs shown in Table VI-146 above to include two additional 
entries, as well as to distinguish financial impacts on government 
agencies from externalities borne broadly across the remainder of the 
U.S. economy. The revised accounting framework for costs and benefits 
is shown in Table VI-147, below. Line 6 of the revised table reports 
the change in consumer surplus experienced by buyers of new cars and 
light trucks when prices and sales of those vehicles adjust in response 
to changes in CAFE and CO2 standards. The gain in consumer 
surplus that occurs when production costs and prices for vehicles fall 
and sales increase in response represents a benefit to buyers, while 
any loss in consumer surplus that occurs when more stringent standards 
increase costs and prices and cause sales to decline appears as a loss 
to new car and light truck buyers.
    Line 7 of Table VI-147 reports the estimated value of changes to 
attributes of new cars and light trucks other than fuel economy that 
their manufacturers make to comply with changes in CAFE and 
CO2 standards. In the case where standards are less 
stringent, manufacturers are able to employ many of the same resources 
they would have deployed to increase fuel economy for the alternative 
purpose of improving other attributes of vehicles that their potential 
buyers value more highly than the forgone improvements in fuel economy. 
This response provides an additional benefit to purchasers of new cars 
and light trucks that was not recognized in the agencies' analysis of 
the proposal, but is included in the analysis of this final rule. Of 
course, if CAFE and CO2 standards are made more stringent, 
manufacturers employ those technologies to increase fuel economy, thus 
sacrificing potential improvements in competing attributes--those that 
entail tradeoffs with higher fuel economy--and the value of 
improvements in those other attributes that is sacrificed or forgone 
represents an opportunity cost to those buyers. This implicit 
opportunity cost is analyzed in a sensitivity analysis and is not 
included in the primary analysis.
    Finally, the agencies revised the framework for reporting costs and 
benefits of changes in CAFE and CO2 standards to identify 
government agencies separately from the entry previously labeled ``Rest 
of U.S Economy.'' This minor revision is intended to distinguish more 
clearly between changes in external costs imposed by externalities that 
result from fuel production and use, and the revenue effects on 
government agencies from changes in tax and civil penalty payments. 
While both effects ultimately result from manufacturers' compliance 
with revised standards and the resulting changes in fuel consumption, 
externalities represent real economic costs; in contrast, changes in 
tax revenues received by government agencies are financial transfers, 
whose offsetting effects on manufacturers and vehicle buyers are also 
recognized elsewhere in the accounting framework.
BILLING CODE 4910-59-P

[[Page 24588]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.296

BILLING CODE 4910-59-C
b) Economic Assumptions
    The agencies' analysis of CAFE and CO2 standards for the 
model years covered by this final rule rely on a range of forecast 
information, estimates of economic, safety, and environmental 
variables, and input parameters. While the analysis accompanying the 
proposal largely resembled previous CAFE and CO2 analyses, 
the agencies updated many of the underlying inputs and assumptions--
based on the most up-to-date data--and expanded the central analysis to 
account for changes in new vehicle sales and the retirement of older 
vehicles.
    EDF, UCS, CARB and others commented that the agencies acted 
arbitrarily and capriciously by changing inputs and assumptions from 
previous analyses, and argued that the agencies failed to provide 
``good reasons'' for the changes.\1523\ In the following sections, the 
agencies will respond directly to these comments. However, the agencies 
note that it would be uncommon to retain inputs and assumptions from 
prior analyses--which are typically informed by transitory empirical 
observations--on the basis of precedent. The agencies are ``neither 
required nor

[[Page 24589]]

supposed to regulate the present and the future within the inflexible 
limits of yesterday.'' \1524\
---------------------------------------------------------------------------

    \1523\ See, e.g,. IPI, Appendix, NHTSA-2018-0067-12213, at 99-
100.
    \1524\ American Trucking Associations v. Atchison, 387 U.S. 397, 
416 (1967).
---------------------------------------------------------------------------

    The agencies also received a number of comments focused on the 
agencies' attempt to incorporate the effects of changes in new vehicle 
prices on new vehicle sales, retirement rates of used vehicles, and the 
resulting ``turnover'' of the vehicle fleet. Some comments endorsed the 
agencies' more comprehensive analysis, although many of those same 
commenters later disagreed with aspects of the results. For example, 
RFF noted that ``Incorporating sales and scrappage effects represents a 
step in the right direction for modeling the effects of the 
regulation.'' \1525\ Similarly, NRDC stated that ``it is reasonable and 
appropriate to develop a mechanism for estimating future vehicle 
populations, and the NPRM documents appropriately present considerable 
discussion on the topic and the derivation of the utilized algorithm.'' 
\1526\ One commenter explicitly recognized that the narrower analysis 
utilized in previous rules likely led to incorrectly estimating costs 
and benefits, and endorsed the broader approach used by the proposal. 
Specifically, American Fuel & Petrochemical Manufacturers stated that 
the absence of scrappage in prior rules ``likely led to a significant 
overestimation of the existing standard's benefits with respect to fuel 
and air pollutant emission reductions and an underestimation of safety 
risks and societal costs.'' FCA also expressed general support for the 
agency's expanded analysis.\1527\
---------------------------------------------------------------------------

    \1525\ Resources for the Future, NHTSA-2018-0067-11789, at 2.
    \1526\ Meszler Engineering Services & Baum and Associates, on 
behalf of Natural Resources Defense Council, NHTSA-2018-0067-11943-
43, NHTSA-2018-0067-11723.
    \1527\ FCA, NHTSA-2018-0067-12078.
---------------------------------------------------------------------------

    In contrast, some commenters objected to the inclusion of `new' 
impacts, including the effect of fuel economy regulations on new 
vehicle prices, the resulting changes in their sales, and retirement 
rates for used cars. Workhorse Group, Inc. noted that the agencies 
``made novel assumptions about the safety impacts of consumers delaying 
vehicle purchases due to the increased costs of fuel economy 
improvements that contradicts the analytical approach NHTSA has 
followed in all prior safety and CAFE rulemakings.'' \1528\ Honda 
agreed ``that significantly higher-priced new vehicles have the 
potential to depress the new vehicle market and thus increase the fleet 
of used vehicles, with concomitant increased safety risks associated 
with driving greater numbers of older vehicles in lieu of newer ones,'' 
but found it ``premature and ill-advised'' to model the impact of fleet 
turnover.\1529\ CBD et. al. argued that the sales and scrappage effects 
were too uncertain to include in the analysis and cited EPA's 2016 
proposed determination as stating, ``a reasonable qualitative 
assessment is preferable to a quantitative estimate lacking sufficient 
basis, or (due to uncertainties like those here) having such an 
enormous range as to be without substantial value.'' \1530\
---------------------------------------------------------------------------

    \1528\ Workhorse Group, Inc., NHTSA-2018-0067-12215.
    \1529\ American Honda Motor Company, Inc., NHTSA-2018-0067-
11818.
    \1530\ Environmental group coalition, Appendix A, NHTSA-2018-
0067-12000, at 174.
---------------------------------------------------------------------------

    As was done repeatedly throughout the proposal, the agencies 
acknowledge that dynamically modeling fleet turnover is new for this 
rulemaking; however, the agencies disagree that the analysis relied on 
`novel' assumptions or contradicted previous analyses. The agencies 
have described the sales and scrappage responses similarly in prior 
rulemakings,\1531\ and have expressed an interest in quantitatively 
measuring them.\1532\ The agencies agree with commenters that--like 
many of the effects included in today's analysis--there remains a 
degree of uncertainty about the magnitude of the sales and scrappage 
responses. However, CBD v. NHTSA stressed that a variable should not be 
excluded from the analysis simply because it is uncertain when the 
effect is quantifiable, ``certainly not zero,'' and the analysis 
``monetize[s] other uncertain benefits.'' \1533\ As discussed in the 
coming sections, the agencies are confident that (a) changes in new 
vehicle prices impact the volume of new vehicle sales and rate of 
retirement of older vehicle, (b) of the direction of those effects, and 
(c) their ability to reasonably estimate the impacts. As such, the 
agencies strongly believe that including the sales and scrappage 
responses improves the thoroughness of the analysis, is consistent with 
case law, and is necessary to comprehensively analyze the cost-benefits 
of the rule.
---------------------------------------------------------------------------

    \1531\ See, e.g., 76 FR 75153.
    \1532\ See, e.g., 77 FR 61971.
    \1533\ 538 F.3d 1172, 1200-02 (2008).
---------------------------------------------------------------------------

    The following subsections briefly describes the sources of the 
agencies' estimates of each of the economic, environmental, and safety 
estimates. In reviewing these variables and the agencies' estimates of 
their values for purposes of this final rule, NHTSA and EPA considered 
comments received in response to the proposed rule and, in response, 
made several changes to the economic assumptions used for the final 
analysis.
1) Macroeconomic Assumptions That Affect the Agencies' Analysis
    As the proposed rule noted, the more comprehensive economic impact 
analysis of CAFE and CO2 included in this rulemaking 
requires a more detailed and explicit explanation of the macroeconomic 
context in which regulatory alternatives are evaluated. The agencies 
continued to rely on projections of future fuel prices to evaluate 
manufacturers' use of fuel-saving technologies, the resulting changes 
in fuel consumption, and various other benefits. Furthermore, the 
agencies expanded the scope of their analysis to include projecting 
future sales of new cars and light trucks, as well as the retirement of 
used vehicles under each regulatory alternative. In addition to 
projections of future fuel prices, constructing these forecasts 
requires explicit projections of macroeconomic variables, including 
U.S. Gross Domestic Product (GDP), labor force participation (the 
number of persons employed or actively seeking employment), and 
bellwether interest rates, which are likely to vary according to 
roughly the same pattern as interest rates on new car loans.
    The analysis presented in the proposal as well as the accompanying 
RIA and EIS employed forecasts of future fuel prices developed by the 
agencies using the U.S. Energy Information Administration's (EIA's) 
National Energy Model System (NEMS). An agency within the U.S. 
Department of Energy (DOE), EIA collects, analyzes, and disseminates 
independent and impartial energy information to promote sound 
policymaking, efficient markets, and public understanding of energy and 
its interaction with the economy and the environment. EIA uses NEMS to 
produce its Annual Energy Outlook (AEO), which presents forecasts of 
future fuel prices, among many other energy-related variables. AEO 
projections of energy prices and other variables are not intended as 
predictions of what will happen; rather, they are projections of the 
likely course of these variables that reflect their past relationships, 
specific assumptions about future developments in global energy 
markets, and the forecasting methodologies incorporated in NEMS. Each 
AEO includes a ``Reference'' case as well as a range of alternative 
scenarios that each incorporate

[[Page 24590]]

somewhat different assumptions from those underlying the Reference 
Case.
    For the proposal, the agencies used the AEO2017 version of NEMS, as 
this was the most current version of the model that was available at 
the time. Using this version of NEMS, the agencies reevaluated the 
``Reference,'' ``Low Oil Price,'' and ``High Oil Price'' cases 
described in AEO2017, by setting aside their assumption that mandates 
by California and other States to sell ``Zero Emission Vehicles'' 
(ZEVs) would be enforced. The agencies used the resulting modified 
Reference case fuel prices as inputs to the proposal's central case 
results, and used the modified ``Low Oil Price'' and ``High Oil Price'' 
case fuel prices, which were generated using NEMS, as inputs to several 
of the sensitivity analysis cases that were presented in the proposal. 
The sensitivity analysis also included a case that applied the 
Reference case fuel prices from the then recently issued AEO2018, which 
did not reflect the modification of EIA's forecasting model to set 
aside state mandates for ZEV sales.\1534\
---------------------------------------------------------------------------

    \1534\ The results of these and other sensitivity analyses were 
reported in NHTSA and EPA, ``Notice of Proposed Rulemaking: The 
Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model Years 
2021-2026 Passenger Cars and Light Trucks,'' Federal Register Vol. 
83, No. 165, August 24, 2018, Tables Vii-90 to Vii-98, pp. 43353-69.
---------------------------------------------------------------------------

    The analysis supporting the proposed rule simulated the economic 
impacts of car and light truck manufacturers' compliance with 
alternative CAFE and CO2 standards through model year 2032, 
and in doing so estimated the number of vehicles originally produced 
and sold in each model year that would remain in service during each 
year of their useful lives (assumed to extend for a maximum of 40 
years), as well as their usage, fuel consumption, and safety 
performance. This required the forecasts of macroeconomic variables 
that affect vehicle sales, use, and retirement rates, which include 
U.S. Gross Domestic Product (GDP), the size of the domestic labor 
force, and key interest rates, to extend well beyond calendar year 
2050. One of the few sources that provides forecasts of these variables 
spanning such a long time horizon was the 2017 OASDI Trustees Report 
from the U.S. Social Security Administration, and the analysis 
supporting the proposed rule relied on this source for forecasts of 
these key macroeconomic measures.\1535\
---------------------------------------------------------------------------

    \1535\ Social Security Administration, The 2017 Annual Report of 
the Board of Trustees of the Federal Old-Age and Survivors Insurance 
and Federal Disability Insurance Trust Funds, available at https://www.ssa.gov/OACT/TR/2017/.
---------------------------------------------------------------------------

(a) Comments on the Fuel Price Forecasts and Macroeconomic Assumptions 
Used in the NPRM Analysis
    The agencies received relatively few comments on the projections of 
fuel prices and macroeconomic variables that were used in their 
analysis supporting the proposed rule, virtually all of them focused on 
the fuel price projections the agencies employed. While only one 
comment questioned the agencies' use of price projections that rely on 
EIA's methodology and assumptions, a few commenters called attention to 
the unreliability of price projections reported in earlier editions of 
AEO. Other comments noted the importance of updating projections used 
to analyze the proposal to reflect more recent developments in energy 
markets, without necessarily questioning the reliability of EIA's fuel 
price projections. Several comments emphasized the implications for the 
agencies' analysis of the wide variation in alternative fuel price 
projections reported in both EIA's 2017 and 2018 Annual Energy 
Outlooks, with most stressing the possibility that future prices might 
be above even those projected in their High Oil Price cases. Only a 
single comment identified a potential alternative source of fuel price 
projections, but noted that it was within the range of projections the 
agencies considered.
    One commenter claimed that AEO's projections of fuel prices are 
``inappropriate'' for the agencies to employ in analyzing the 
consequences of CAFE and CO2 standards; because EIA ``does 
not speculate on changes in international policy or geopolitics,'' 
which contribute to the uncertainty surrounding future prices.\1536\ 
However, this commenter did not identify an alternative source for fuel 
price projections that reflect such considerations; and because 
projections of fuel prices are a central element in the agencies' 
evaluation of alternative future standards, the observation that EIA's 
projections do not incorporate some sources of uncertainty is unhelpful 
by itself.
---------------------------------------------------------------------------

    \1536\ NHTSA-2018-0067-11837, Alliance to Save Energy, p. 2 
(``EIA takes a transparently conservative approach in modeling 
future oil prices, and does not speculate on changes in 
international policy or geopolitics. As a result, their projections 
are an inappropriate measure of future fuel prices.'').
---------------------------------------------------------------------------

    Some commenters asserted that by relying on the AEO2017 Reference 
Case projections of fuel prices in their central analysis of the 
proposed rule while considering the significantly higher fuel prices 
projection in the AEO High Oil Price scenario only in the accompanying 
sensitivity analyses, the agencies inadequately considered the possible 
effect of higher fuel prices on the estimated economic benefits from 
alternatives that would have relaxed the augural standards, including 
the preferred alternative.\1537\ Surprisingly, none of these comments 
acknowledged that the fuel price projections reported in the High Oil 
Price cases accompanying past editions of the Annual Energy Outlook 
have so far proven to be significantly above actual prices, or that EIA 
has consistently lowered its fuel price projections in more recent 
editions of the AEO. In any case, supplemental material included in the 
NPRM regulatory docket showed that the ranking of regulatory 
alternatives by their estimated net economic benefits remained 
unchanged from the central analysis in the sensitivity analysis that 
substituted the AEO2017 High Oil Price case projection of fuel prices.
---------------------------------------------------------------------------

    \1537\ See e.g., Securing America's Future Energy (SAFE), NHTSA-
2018-0067-11981, pp. 12 & 30 and Institute for Policy Integrity, 
NHTSA-2018-0067-12213, p. 31.
---------------------------------------------------------------------------

    None of the commenters who argued that the agencies inadequately 
considered the possibility of higher fuel prices observed that the 
agencies' analogous use of lower fuel price projections from the 
AEO2017 Low Oil Price case only in their sensitivity analyses 
inadequately considered the possibility that future fuel prices might 
prove to be lower than projected in the AEO2017 Reference Case, and its 
potential effect on the proposal's estimated benefits. Nor did any of 
the commenters offer substantive guidance about how the agencies might 
revise their analysis to accord greater emphasis to fuel price 
projections above (or below) those from the AEO Reference Case.\1538\
---------------------------------------------------------------------------

    \1538\ One commenter did refer to guidance to EPA contained in a 
National Research Council report on incorporating and conveying 
uncertainty about key inputs directly into that agency's estimates 
of benefits from reducing air pollution, rather than simply 
recognizing it in supplemental sensitivity analyses. This was 
presumably intended as potential guidance to the agencies about how 
they might do so in their evaluations of fuel economy and 
CO2 standards, although that was not stated explicitly. 
See American Fuel & Petrochemical Manufacturers, NHTSA-2018-0067-
12078, p. 19, citing National Research Council (2002), Estimating 
the Public Health Benefits of Proposed Air Pollution Regulations, 
2002, available at https://www.nap.edu/catalog/10511/estimating-the-public-health-benefits-of-proposed-air-pollution-regulations.
---------------------------------------------------------------------------

    Other comments stressed the fact that EIA's current projections of 
future fuel prices are significantly lower than those the agencies 
relied on when they established CAFE standards through

[[Page 24591]]

model year 2021 and introduced the augural standards for subsequent 
model years in the rulemaking they conducted in 2012, citing this as 
support for the agencies' reconsideration of the augural standards in 
the current rulemaking.\1539\
---------------------------------------------------------------------------

    \1539\ For example, Fiat Chrysler Automobiles (FCA) pointed out 
that the AEO 2017 Reference Case forecast of gasoline prices through 
2025 is approximately 36% lower than that in the AEO 2012 Reference 
Case, which the agencies relied on in the analysis supporting that 
earlier rulemaking; see NHTSA-2018-0067-11943, p. 33.
---------------------------------------------------------------------------

    One comment compared the range of fuel price projections spanned by 
the High and Low Oil Price cases from AEO2017 and AEO2018 to the range 
of future prices spanned by another widely-recognized and relied-upon 
projection, concluding that the alternative scenarios included in 
AEO2017 incorporated an even wider range of uncertainty about future 
prices, and noted that the net economic benefits of the preferred 
alternative were positive over this entire range of alternative future 
fuel prices. This same commenter noted that by combining high and low 
fuel price projections with alternative assumptions about other key 
economic variables (such as GDP growth) and parameter assumptions 
(principally payback period), the agencies' sensitivity analyses 
captured potentially important interactions between uncertainty 
regarding fuel prices and other key economic inputs.\1540\
---------------------------------------------------------------------------

    \1540\ See Alliance of Automobile Manufacturers, NHTSA-2018-
0067-1207, p. 108.
---------------------------------------------------------------------------

(b) Macroeconomic Assumptions Used To Analyze Economic Consequences of 
the Final Rule
    After considering these comments, the agencies have concluded that 
there is no convincing reason to rely on sources other than EIA's NEMS 
model to project future energy prices, or to rely on alternatives to 
the Reference Case scenario in the current edition of AEO as their 
basis for using NEMS. The agencies agree that the resulting projections 
will be uncertain, but note that EIA regularly publishes retrospective 
analyses comparing past Reference case projections to subsequent market 
price outcomes, thus enabling an assessment of this uncertainty. 
Although EIA does not identify its Reference case as a ``most likely'' 
outcome, in the agencies' judgment that case's design--which assumes 
future trends are consistent with historical and current market 
behavior--makes it a reasonable and appropriate basis for projecting 
fuel prices to use in the agencies' central analysis of alternative 
CAFE and CO2 standards.
    The agencies also conclude that the wide range of uncertainty about 
future petroleum prices encompassed in EIA's ``Low Oil Price'' and 
``High Oil Price'' cases means that including them in the accompanying 
sensitivity analyses provides a meaningful basis for assessing the 
potential economic consequences of future energy prices that prove to 
be considerably lower or higher than those reflected in the Reference 
case. Although these alternative cases do not incorporate unbridled 
speculation regarding hypothetical changes in ``international policy or 
geopolitics,'' the agencies believe that this restraint means that 
relying on them produces a more, rather than less, meaningful test of 
the effect of the inherent uncertainty surrounding projections of fuel 
prices.
    For today's final rule, the agencies have therefore used the 
AEO2019 version of NEMS to develop projections of future prices for 
transportation fuels, as this was the most current version available 
when this analysis was conducted. Using this version of NEMS, the 
agencies modified EIA's AEO2019 Reference case by (1) setting aside 
presumed enforcement by California and other States of any mandates to 
sell ``Zero Emission Vehicles'' (ZEVs), (2) setting aside post-2020 
increases in the stringency of CAFE and CO2 standards, and 
(3) modifying inputs regarding battery costs, in order to bring those 
costs down to levels more consistent with battery cost estimates 
applied in the CAFE model analysis.\1541\ All other NEMS inputs used to 
develop the AEO2019 Reference case were left unchanged in this 
analysis.
---------------------------------------------------------------------------

    \1541\ These inputs are all contained in the ``trnldvx.xlsx'' 
NEMS input file. The input file utilized for today's analysis is 
available in regulatory docket NHTSA-2018-0067, https://www.regulations.gov/docket?D=NHTSA-2018-0067 (see Supporting 
Documents), as is the corresponding output file from which reference 
case fuel and electricity prices were obtained to be used as inputs 
to the CAFE model. The version of NEMS utilized for today's analysis 
is available at https://www.eia.gov/outlooks/aeo/info_nems_archive.php.
---------------------------------------------------------------------------

    Setting aside enforcement of state mandates to sell ZEVs makes the 
supporting analysis consistent with the agencies' recent One National 
Program Action,\1542\ under which EPA withdrew aspects of a Clean Air 
Act Preemption waiver previously granted to California, and NHTSA 
concluded that EPCA expressly and implied preempted State ZEV mandates. 
Setting aside the post-2020 increase in the stringency of CAFE and 
CO2 standards ensures that the fuel prices used in the 
agencies' analysis are at least as high as those that would prevail 
under the least stringent regulatory alternative considered, since that 
alternative produces the highest level of fuel consumption and thus the 
highest fuel prices.
---------------------------------------------------------------------------

    \1542\ 84 FR 51310.
---------------------------------------------------------------------------

    Figure VI-55 and Figure VI-56 below show the resulting modified 
projections of BEV prices and sales, and compare them to the 
projections reported in EIA's AEO2019 Reference case. As they 
illustrate, the combination of these modifications led NEMS to project 
significantly lower BEV prices and correspondingly higher BEV sales 
volumes. Figure VI-57 and Figure VI-58 show the modified projections of 
gasoline and electricity prices, and again compare these to the 
projections reported in EIA's AEO2019 Reference case. As those figures 
indicate, the agencies' modifications to NEMS did not significantly 
affect its projections of future prices for transportation fuels.
BILLING CODE 4910-59-P

[[Page 24592]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.297

BILLING CODE 4910-59-C
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.298

BILLING CODE 4910-59-C
BILLING CODE 4910-59-P

[[Page 24593]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.299

BILLING CODE 4910-59-C
    The agencies used the resulting Reference case fuel prices as 
inputs to the rule's central analysis. The agencies also used the as-
published (by EIA) ``Low Oil Price'' and ``High Oil Price'' case fuel 
prices as inputs to several of the cases included in the sensitivity 
analysis presented in the accompanying RIA.
    For the projections of macroeconomic variables used in the analysis 
supporting this rule, the agencies elected to rely on different sources 
from those that informed their analysis of the proposed rule. 
Specifically, the agencies rely on projections of future growth in U.S. 
GDP reported in AEO2019 to support their central analyses of the final 
rule's impacts on new car and light truck sales and the retirement of 
used vehicles. These incorporate underlying projections generated using 
the IHS Markit Global Insight long-term macroeconomic model, as 
modified via this model's interaction with NEMS' representation of 
global energy markets and their future outcomes. The alternative 
projections of future growth in GDP used in the agencies' accompanying 
sensitivity analyses are drawn from the AEO2019 High Economic Growth 
and Low Economic Growth cases. These reflect alternative future trends 
in U.S. labor force and productivity growth, and are also consistent 
with the energy market outcomes projected by NEMS under the resulting 
future performance of the U.S. economy.
[GRAPHIC] [TIFF OMITTED] TR30AP20.300


[[Page 24594]]


    For estimates of the number of U.S. households during future years, 
which influence the projections of new car and light truck sales used 
in the analysis, the agencies rely on projections of new household 
formation developed the Harvard University Joint Center for Housing 
Studies.\1543\ These are consistent with the most recent projections of 
future growth in the nation's population prepared by the U.S. Bureau of 
the Census.\1544\
---------------------------------------------------------------------------

    \1543\ See Harvard University Joint Center for Housing Studies, 
Updated Household Growth Projections: 2018-2028 and 2028-2038, 
December 18, 2018, available at https://www.jchs.harvard.edu/sites/default/files/Harvard_JCHS_McCue_Household_Projections_Rev010319.pdf.
    \1544\ Ibid., pp. 2-5.
---------------------------------------------------------------------------

(2) Approach To Estimating Sales Response Under Different Standards
    Prior to the NPRM, all previous CAFE and CO2 rulemaking 
analyses used static fleet forecasts that were based on a combination 
of manufacturer compliance data, public data sources, and proprietary 
forecasts (or product plans submitted by manufacturers). When 
simulating compliance with regulatory alternatives, those analyses 
projected identical sales across the alternatives, for each 
manufacturer down to the make/model level--where the exact same number 
of each model variant was assumed to be sold in a given model year 
under both the least stringent alternative (typically the baseline) and 
the most stringent alternative considered (intended to represent 
``maximum technology'' scenarios in some cases). To the extent that an 
alternative matched the assumptions made in the production of the 
proprietary forecast, using a static fleet based upon those assumptions 
may have been warranted. However, a sales forecast is unlikely to be 
representative of a broad set of regulatory alternatives with 
significant variation in the cost of new vehicles. A number of 
commenters on previous regulatory actions encouraged consideration of 
the potential impact of fuel efficiency standards on new vehicle prices 
and sales, and the changes to compliance strategies that those shifts 
could necessitate.\1545\ In particular, the continued growth of the 
utility vehicle segment creates compliance challenges within some 
manufacturers' fleets as sales volumes shift from one region of the 
footprint curve to another, or as mass is added to increase the ride 
height of a vehicle on a sedan platform to create a crossover utility 
vehicle, which exists on the same place of the footprint curve as the 
sedan upon which it might be based.
---------------------------------------------------------------------------

    \1545\ See, e.g., Alliance of Automobile Manufacturers, Comment, 
EPA-HQ-OAR-2015-0827-4089, at 115-16.
---------------------------------------------------------------------------

    However, some NPRM commenters referenced the agencies' previous 
omission of this effect as justification to continue ignoring this 
issue in the current rulemaking. EDF commented,\1546\ ``use of a sales 
response model constitutes an unexplained reversal in the agency's 
position on the feasibility of doing so.'' To say that the agencies 
never used a model is a misrepresentation. Assuming that sales never 
change in any model year, even at the individual nameplate level, 
regardless of the stringency of fuel economy regulations or the 
technology costs required to comply with those regulations, is, itself, 
a model. It is a model that implicitly asserts that, while fuel economy 
regulation impacts vehicle prices, such regulations have no impact on 
the quantity or mix of new vehicle sold, regardless of stringency. This 
is an implicit argument that new vehicle demand is perfectly 
inelastic--and that no change in vehicle prices can impact the number 
of cars consumers will buy. Logically, however, there must exist a 
level of stringency that would have a negative impact on new sales. 
Picking an extreme example to prove the point, if the agencies set 
standards at an extraordinarily stringent level that forced all 
vehicles into battery electric propulsion systems next year, sales 
would obviously be impacted. The increase in new vehicle price or 
changes to other relevant attributes like range, refueling time, or 
operating cost would surely affect the decisions of some buyers. But, 
by arguing that the agencies should continue to model new vehicle sales 
as if they are entirely unaffected by standards, commenters are 
effectively asking the agencies to assume that the alternatives 
considered in this rule are insufficiently stringent to affect the 
market. By endorsing the approach from the 2012 final rule, which 
assumed no impact on the new vehicle market from standards as stringent 
as 7 percent increase, year-over-year, beginning in 2017, commenters 
are suggesting that even those standards would have no impact on new 
vehicle sales. Manufacturers have asserted in their comments that fuel 
economy regulations change both the cost of producing new vehicles and 
consumer demand for them. In the recent peer review of the NPRM release 
of the CAFE model, all reviewers encouraged the inclusion of a sales 
response to fuel economy regulations (albeit not necessarily the 
version of the response model that appeared in the NPRM).\1547\ Based 
on earlier comments and the agencies' own analysis, the agencies were 
persuaded to include a sales response mechanism in the NPRM, and do so 
again in this final rule.
---------------------------------------------------------------------------

    \1546\ EDF, Appendix B, NHTSA-2018-0067-12108, at 37-38.
    \1547\ CAFE Model Peer Review, DOT HS 812 590, Revised (July 
2019), available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
---------------------------------------------------------------------------

    While several commenters (CARB, NCAT, CBD, Aluminum Association) 
discouraged the agencies from attempting to account for the effect of 
regulations on new vehicle sales, other commenters stated that the NPRM 
analysis was improved by explicitly considering this effect (RFF, 
Toyota, the Alliance of Automobile Manufacturers). CBD cited EPA's 2016 
proposed determination, stating ``[a] reasonable qualitative assessment 
is preferable to a quantitative estimate lacking sufficient basis, or 
(due to uncertainties like those here) having such an enormous range as 
to be without substantial value.'' \1548\ However, RFF supported the 
inclusion of the effect (with caveats about the specific 
implementation, for which they suggested alternative approaches), 
stating ``[i]ncorporating sales and scrappage effects represents a step 
in the right direction for modeling the effects of the 
regulation.\1549\ It is reasonable to conclude that regulations as 
transformative as fuel economy standards will impact the market for new 
vehicles, and excluding the effect (as CBD and others suggested) is 
equivalent to stating that it does not exist.
---------------------------------------------------------------------------

    \1548\ Environmental group coalition, Appendix A, NHTSA-2018-
0067-12000, at 174.
    \1549\ RFF, Comments, NHTSA-2018-0067-11789, at 3.
---------------------------------------------------------------------------

    The NPRM version of the sales response relied on differences in the 
average price of new vehicles to produce sales differences between 
regulatory alternatives. Some commenters (ACEEE, IPI, CBD, UCS, 
Aluminum Association, and Alliance to Save Energy) argued that new 
vehicle prices do not increase with the addition of technology required 
to comply with fuel economy regulations. Some argued that manufacturers 
will choose not to ``pass through'' the full incremental cost of fuel 
saving technologies to consumers, instead absorbing those costs into 
their profit margin.\1550\ The question of cost pass-through is one 
that academic and industry researchers have considered for decades--and 
two of the

[[Page 24595]]

agencies' recent peer reviewers addressed this issue in their comments.
---------------------------------------------------------------------------

    \1550\ E.g. IPI, Appendix, NHTSA-2018-0067-12213, 28-29; CBD et 
al., Attachment 1, NHTSA-2018-0067-12123, at 23-24.
---------------------------------------------------------------------------

    Dr. John D. Graham, one of the peer reviewers, argued that the 
assumption of complete cost pass-through is defensible, and more likely 
in the long-run than the short-run.\1551\ The reviewer also suggested 
that changes to the CAFE (and subsequent CO2) program that 
base a manufacturer's standard on the mix of vehicle footprints in each 
fleet more equitably spreads the impact of the standards across the 
industry, and that industry shifts toward increasingly competitive 
market models (rather than the oligopolistic models that existed 
earlier in the last century) both act to increase the likelihood that 
manufacturers will pass regulatory costs through to consumers. In 
particular, this reviewer stated: \1552\
---------------------------------------------------------------------------

    \1551\ CAFE Model Peer Review, DOT HS 812 590, Revised (July 
2019), pp. B31-B33, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
    \1552\ Gron Anne, Swenson, Deborah L, Cost Pass-Through in the 
US Automobile Market, Review of Economics and Statistics, Vol. 82(2) 
(May 2000), at 3.

    In a classic study, Gron and Swenson (2000) examined list prices 
of automobiles at the model level in the U.S. from 1984 to 1994 
coupled with data on production, vehicle characteristics, foreign 
versus domestic firm ownership, wages of employees, exchange rates, 
imported parts content, tariffs and other variables. Although their 
work rejects the hypothesis of 100% pass through of cost to consumer 
price, they find higher rates of pass through than previous studies, 
and much of the incomplete pass through occurs when cost increases 
impact only a few models or firms. Confirming earlier studies, they 
show that U.S. auto manufacturers engage in more aggressive pass-
through pricing than Asian and European manufacturers (greater than 
100% in some specifications), possibly due to the eagerness of 
importers to enlarge market share in lieu of recovering regulatory 
costs, at least in the short run (see Dinopolous and Kreinin, 1988; 
\1553\ Froot, 1989 \1554\[hairsp]). This study helps explain why 
pass-through pricing is a more viable hypothesis in the long run 
than in the short run.
---------------------------------------------------------------------------

    \1553\ Dinopoulos, Elias, Kreinin, Mordechai, Effects of U.S.-
Japan Auto VER on European Prices and on U.S. Welfare, The Review of 
Economics and Statistics, Vol. 70(3) (1988), at 484-91.
    \1554\ Froot, Kenneth A, Klemperer, Paul D, Exchange Rate Pass-
Through When Market Share Matters, American Economic Review, Vol. 
79(4) (1989), at 637-54.
---------------------------------------------------------------------------

    The original design of the CAFE program is a contrasting case 
where pass-through pricing was difficult for some automakers. All 
auto makers, regardless of their product mix, were subject to the 
same fleet-wide average CAFE standard, such as 27.5 miles per gallon 
for cars in 1990. In practice, those standards impacted only three 
high-volume companies (General Motors, Ford and Chrysler) because 
the Big Three produced a higher proportion of large and performance-
oriented vehicles than did Japanese companies. As a result, 
manufacturers such as Toyota and Honda consistently surpassed the 
federal fleet-wide standard for cars without any regulatory cost 
(i.e., partly due to their smaller product mix). In the 1975-2007 
period, the Big Three were not able to pass on all of their 
compliance costs to consumers and thus experienced some declines in 
profitability due to CAFE (Kleit, 1990; \1555\ Kleit, 2004; \1556\ 
Jacobsen, 2013\1557\[hairsp]).
---------------------------------------------------------------------------

    \1555\ Kleit, Andrew N., The Effect of Annual Changes in 
Automobile Fuel Economy Standards, Journal of Regulatory Economics, 
Vol. 2. (1990,), at 151-72.
    \1556\ Kleit, Andrew N, Impact of Long-Range Increases in the 
Fuel Economy (CAFE) Standard, Economic Inquiry, Vol. 42(2) (2004), 
at 279-94.
    \1557\ Jacobsen, Mark R., Evaluating U.S. Fuel Economy Standards 
in a Model with Producer and Household Heterogeneity, American 
Economic Journal: Economic Policy, Vol. 5(2) (2013), at 148-87.
---------------------------------------------------------------------------

    When the CAFE program was reformed for light trucks in 2008 (and 
for cars in 2011) on the basis of vehicle size (the so-called 
``footprint'' adjustments to CAFE stringency), the, the technology 
costs of CAFE standards were spread more evenly among automakers, 
although the overall societal efficiency of the regulation 
diminished due to the removal of downsizing as a compliance 
option.\1558\ Given that the size-based fuel economy programs are 
not concentrating the costs of compliance on one or two automakers, 
it is reasonable to predict a fairly high degree of pass-through 
pricing for the 2021-2025 fuel economy standards. In related 
literature on manufacturer pricing responses to a national carbon 
tax, Bento and Jacobsen (2007) \1559\ and Bento (2013) \1560\ report 
high rates of pass-through pricing (on the order of 85%). Carbon 
taxes are more efficient than footprint-based CAFE standards, but 
both instruments are likely to impact a wide range of companies in 
the auto sector and result in a high degree of pass-through pricing 
by impacted companies.
---------------------------------------------------------------------------

    \1558\ See Ito, Koichiro, Sallee, James M., The Economics of 
Attribute-Based Regulation: Theory and Evidence from Fuel-Economy 
Standards, Review of Economics and Statistics, in press (2018).
    \1559\ Bento, Antonio M., Jacobsen, Mark R, Environmental Policy 
and the `double-dividend' hypothesis, Journal of Environmental 
Economics and Management, Vol. 53(1) (January 2007) at 17-31.
    \1560\ Bento, Antonio M. Equity Impacts of Environmental Policy, 
Annual Review of Resource Economics, Vol. 5 (May 2013), at 181-96.
---------------------------------------------------------------------------

    Also, it should be noted that the U.S. automotive industry is 
much more competitive today than it was from 1970 to 2000. The 
market share of General Motors, once the dominant, majority producer 
in the U.S. market, has declined dramatically, and a variety of 
Japanese and Korean companies have captured substantial market 
share. Moreover, the rise of startups (e.g., Tesla and other 
electric vehicle start-ups) and ride-sharing services (e.g., Uber) 
are adding a new competitive dimension in the U.S. industry. As a 
result, some of the most recent auto regulatory studies have given 
more emphasis to analytic results based on competitive models than 
oligopolistic models (see, e.g., Davis and Knittel (2016) 
\1561\[hairsp]).
---------------------------------------------------------------------------

    \1561\ Davis, Lucas, Knittel, Christopher R., Are Fuel Economy 
Standards Regressive? Working Paper 22925, National Bureau of 
Economic Research, Cambridge, MA (2016).

    Another peer reviewer, Dr. James Sallee, suggested that costs would 
pass through to new vehicle buyers to different degrees, depending upon 
the stringency of the standards.\1562\ The reviewer argued that more 
stringent standards, which result in larger increases to the cost of 
production, are likely to induce greater degrees of pass-through than 
less stringent standards, which automakers may, as some commenters have 
suggested, be able to absorb in the form of lost profit. If the degree 
of cost pass-through should vary by the stringency of the alternative, 
the agencies are underestimating the difference in price between the 
most and least stringent alternatives--which would favor alternatives 
with higher stringency.
---------------------------------------------------------------------------

    \1562\ CAFE Model Peer Review, DOT HS 812 590, Revised (July 
2019), pp. B54-B75, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
---------------------------------------------------------------------------

    Other commenters argued that manufacturers are able to compensate 
fully for the costs of fuel economy standards by increasing the prices 
of luxury vehicles--which would increase the average new vehicle price, 
but leave large sections of the market unaffected by the increased cost 
of producing fleets that comply with the standards. While it seems 
likely that manufacturers employ pricing strategies that push 
regulatory costs (as well as increases in costs like pension 
obligations and health care costs for employees) into the prices of 
models and segments with less elastic demand, the extent to which any 
OEM is able to succeed at this is unknown by the agencies. At some 
point, however, price increases on even luxury models will merely price 
more and more purchasers out of the market, and make competition with 
other manufacturers and market segments that much more difficult. And 
the more that avoided price increases for lower ends of the vehicle 
market are subsidized by luxury vehicles, the more either prices for 
luxury models would need to be increased, or (if moderately increasing 
prices) more of those luxury models would need to be sold. It is worth 
noting that luxury vehicles tend to be more powerful and content-rich, 
and often have fuel economy levels below (or CO2 levels 
above) their targets on the curves--so that selling more of them to 
compensate for lost profit elsewhere

[[Page 24596]]

further erodes the compliance levels of the fleets in which they 
reside.
    While manufacturers could conceivably push some small cost 
increases into the prices of their vehicle segments that have less 
elastic demand to cover accordingly small increases in stringency, 
larger stringency increases would exhaust the ability of such segments 
to absorb additional costs. In addition, the agencies do not attempt to 
adjust the mix of vehicle models based on their own price elasticity of 
demand; doing so would require a pricing model that takes the 
compliance cost for each manufacturer (which the agencies' model 
estimates dynamically) and apportions that cost to the prices of 
individual nameplates and trim levels. The agencies have experimented 
with pricing models (when integrating vehicle choice models, pricing 
models are a necessity), but each manufacturer almost certainly has a 
unique pricing strategy that is unknown to the agencies, and involves 
both strategic decisions about competitive position within a segment 
and the volumes needed fully to amortize fixed costs associated with 
production. To the extent that the agencies assume all regulatory costs 
are passed through and affect the average regulatory cost of each 
vehicle instead of being priced in a fashion to minimize the impact on 
aggregate sales, the agencies note that--more stringent alternatives 
are provided an artificial analytical advantage because manufacturers 
are better positioned to incorporate smaller price adjustments into 
their current strategic pricing models. The agencies opted to take the 
conservative approach instead of speculating on manufacturer's private 
business models.
    Finally, some commenters have argued that, even if regulations do 
increase the cost of producing vehicles and those costs are passed on 
to new vehicle buyers, it does not matter because sales have increased 
in recent years under both rising standards and rising prices. EDF, 
CARB, Aluminum Association, SAFE, CBD, and CA et al. and Oakland et 
al., all make some version of this argument in their comments.\1563\ 
The commenters are confusing correlation with causation and failing to 
consider the counterfactual case. Higher prices of new vehicles 
certainly did not cause sales to increase since 2012. Sales increased 
over that period, in large part, as a result of economic expansion 
following the great recession.\1564\ The statistical model used in the 
NPRM attempted to isolate the effect of average price on new vehicle 
sales, independent of the overall health of the US economy which plays 
an obviously important role. That model showed a negative relationship 
between sales and price (albeit a modest one), and positive 
relationships with GDP and employment. Even under the most stringent 
alternative in the NPRM, sales increased over time. However, in other 
alternatives, where the same macroeconomic conditions prevailed but 
average new vehicle prices were lower, sales increased relative to the 
baseline. That is the counterfactual case that is relevant for 
regulatory analysis--it attempts to answer the question, ``would sales 
have been even higher if average prices had been lower?''
---------------------------------------------------------------------------

    \1563\ See, e.g. EDF, Appendix B, NHTSA-2018-0067-12108, at 37; 
CARB, Detailed Comments, NHTSA-2018-0067-11873, at 198-204; Aluminum 
Association, Comments, NHTSA-2018-0067-11952, at 19-21; SAFE, 
Comments, NHTSA-2018-0067-11981 at 36; CBD et al., Attachment 1, 
NHTSA-2018-0067-12123, at 20. States and Cities, Detailed Comments, 
NHTSA-2018-0067-11735, at 87-89.
    \1564\ Table VI-148 below shows a large and statistically 
significant effect of GDP on sales.
---------------------------------------------------------------------------

    As discussed below, identifying the independent contribution of 
price to new vehicle sales is econometrically challenging. In the NPRM, 
the agencies stated that the simultaneous nature of price and sales--
where transaction prices are higher in periods of higher demand, 
because the market will bear them, and lower in periods of lower 
demand, because the market will not, for an otherwise identical 
vehicle--creates a form of reverse causality. As commenters suggested, 
in recent years sales have increased along with average transaction 
price increases--and transaction price increases will occur when 
regulation forces manufacturers to add content, and their corresponding 
costs, to the vehicles they sell. Thus, it is understandable that some 
commenters could interpret the recent increase in new vehicle sales 
following the recession as evidence that standards (and maybe prices) 
have no impact on new sales. However, that view confuses correlation 
for causation (or lack thereof, in this case).
    In response to these comments, the agencies have modified their 
approach to modeling the sales impacts of regulatory alternatives. In 
order to isolate the impact of the standards, the agencies have broken 
the sales response module into two discrete components. The first 
captures the effects of broader economic forces such as GDP growth. The 
second measures how changes in vehicle prices influence sales. As 
elaborated in more detail in the following passages, the agencies 
considered alternative approaches and specific changes suggested by 
commenters, but concluded that the comments either lacked enough 
information to implement a change, failed to remedy identified alleged 
weaknesses of the NPRM model, or created new limitations for which 
there were no practical solutions. Furthermore, the two-pronged 
approach addresses many of the concerns raised by commenters better 
than any specific modeling alteration. First, the structural changes to 
the model address many of the econometric concerns raised by 
commenters. Second, by modeling sales in the first step as a function 
of macroeconomic conditions, and then applying an independent own-price 
elasticity to estimate the change in sales across alternatives, the 
agencies are able to more clearly distinguish between demand-side and 
supply-side impacts on prices, the issue that appears to have tripped 
up some of the commenters.
Comments on the Econometric Model Used in the NPRM
    Any model of sales response must satisfy two requirements: It must 
be appropriate for use in the CAFE model, and it must be based in both 
sound economic theory and appropriate empirical analysis. The first of 
these requirements implies that forecasts of any variable used in the 
estimation of the econometric model must also be available as a 
forecast throughout the duration of the years covered by the 
simulations (this analysis explicitly simulates compliance through MY 
2050). Some values the model calculates endogenously, making them 
available in future years for sales estimation, but others must be 
known in advance of the simulation. As the CAFE model simulates 
compliance, it accumulates technology costs across the industry and 
over time. By starting with the last known average transaction price 
(associated with MY 2016, in this analysis) and adding accumulated 
regulatory costs to that value, the model is able to represent an 
estimated average selling price in each future model year, assuming 
that manufacturers are able to pass their compliance costs on to buyers 
of new vehicles. Other variables used in the estimation can be entered 
into the model as inputs prior to the start of the compliance 
simulation.
    The NPRM analysis was based on an econometric model that attempted 
to estimate the price elasticity of aggregate demand for new light-duty 
vehicles based on exogenous factors, intended to represent (1) 
macroeconomic forces that influence demand for new vehicles, and (2) 
average new vehicle price, intended

[[Page 24597]]

to represent the impact of regulation. A number of commenters voiced 
opposition to the approach. Some disagreed with the theoretical framing 
of the issue--arguing that the model of sales response should have 
acknowledged the relevance of other vehicle attributes, included 
consumer valuation of fuel savings for new vehicles, based the response 
on something other than price, and considered the effect at a lower 
level of aggregation, rather than average price across the industry.
    In the NPRM, the agencies relied upon an autoregressive distributed 
lag (ARDL) statistical model to estimate the impact of price 
differences between regulatory alternatives and to produce a time 
series of total new vehicle sales in each year of the analysis. The 
statistical model estimated new vehicle sales per year based on two 
lagged variables of new sales (new sales in the previous period, and 
the period before that), GDP and lagged GDP, and labor force 
participation and lagged labor force participation. The model used 
quarterly data and seasonally adjusted annual rates to increase the 
number of observations over the sample period for which reliable sales 
data existed (1978-2015). The ARDL model used in the NPRM was chosen to 
address sales impacts at a high level of aggregation, namely the total 
new vehicle market (across all vehicle brands and body styles), and to 
resolve the econometric issues associated with the time series data 
related to total new vehicle sales.
    Stock et al. commented at length on the econometric specification 
of the NPRM sales response model, identifying limitations and 
suggesting alternative approaches.\1565\ In particular, they argued 
that the length of the response to price shocks should dissipate faster 
than the NPRM model allows--an artifact of using quarterly data and 
seasonally adjusted annual rates to estimate the effect and 
implementing it on an annual basis in the CAFE model. The agencies 
agree that this was a flaw in the implementation of the NPRM model. 
While this approach produced the correct units (i.e., annual sales) the 
response to changes in price should have dissipated at a quarterly 
rate, rather than an annual rate. As a result, a single price shock, 
which appears in one year and disappears the next, was projected to 
have a longer impact on sales in future years than was appropriate 
given the specification. The sales response in the final rule corrects 
for this objective error and takes a more conservative approach to 
price shocks.
---------------------------------------------------------------------------

    \1565\ EPA-HQ-OAR-2018-0283 and NHTSA-2018-0067.
---------------------------------------------------------------------------

    Stock et al. commented that ``it is important to estimate the 
dynamic effect on sales of a price increase, that is, the causal effect 
on current and future demand of a price increase'' because ``it allows 
the response to an intervention--here, a one-time price increase or 
sequence of such increases--to evolve over time.'' \1566\ The comment 
suggests that the agencies should include future responses in sales to 
a one-time price increase that exists for a single period and then 
disappears. In our analytical framework, this implies that a price 
difference between any alternative and the baseline that causes a 
difference in sales in that year should also produce a difference in 
sales in the following year (and possibly subsequent years), though of 
smaller magnitude, even if the price difference only exists for a 
single period. The Stock et al. comment illustrates a quickly 
diminishing response to a single price shock. The final rule assumes 
(more conservatively) that each price shock lasts only for a single 
year, and produces no future ``ripple'' effects in the new vehicle 
market in subsequent years. Furthermore, the regulatory alternatives 
considered in this analysis do not produce single period price shocks 
(in the form of price differences between alternatives), but rather 
persistent price differences between alternatives that result from 
continued differences in stringency. The persistent nature of the price 
differences resulting from fuel economy and CO2 regulations 
further reduce the importance of capturing these multi-period effects 
caused by single-period price shocks.
---------------------------------------------------------------------------

    \1566\ Ibid.
---------------------------------------------------------------------------

    Stock et al. also objected to the use of an ARDL model to estimate 
the impact of price on new vehicle sales. In order for the estimation 
of causality to be valid in a time series model, the current price 
movements must be uncorrelated with unobserved demand shocks in the 
past, present, and future; so-called strict exogeneity. The commenters 
argue that the NPRM fails this test because actions taken in the market 
(by both buyers and sellers) can influence the response to price 
changes in the next period. They suggest the use of a vector 
autoregression (VAR) model to address the relationship between past 
demand disturbances and current prices to address the temporal 
exogeneity issues they identify. However, an important caveat is that 
this approach still does not resolve the largest econometric 
challenge--that of contemporaneous endogeneity between price and sales 
(in the same period). To address that challenge, one needs to employ 
instrumental variable methods.
    The agencies attempted several modifications to the statistical 
model developed for the NPRM based on the Stock et al. comment. The 
agencies reviewed the initial approach and attempted several 
specifications that would explicitly address the temporal endogeneity 
bias identified in the comment. In particular, the agencies addressed 
data limitations that were raised by Stock et al. (and also by EDF), 
who encouraged us to reconsider the quarterly specification and to use 
quality-adjusted price data for new vehicles in order to ensure a more 
consistent definition of the average vehicle over the time series, as 
the ``average vehicle'' has consistently improved in a myriad of ways 
over successive model years. The quarterly price series was 
statistically interpolated in the NPRM to increase the number of 
observations,\1567\ but represented a less-than-ideal solution. The 
interpolating process may have impacted the underlying quarterly data 
generating process, resulting in unreliable, or potentially biased, 
regression results. This issue was remedied by sourcing both vehicle 
sales and price data from IHS Markit, which provides these data at the 
same base frequency (quarterly) and obviates the need for any 
interpolation. In addition, the macroeconomic data used in the model 
specification were also sourced from IHS, which provides consistency 
between historical and forecast data (i.e., forecasts of sales, price, 
personal income, etc., were all based on a consistent set of input 
assumptions and modeling framework during testing).
---------------------------------------------------------------------------

    \1567\ Interpolation is the practice of adding unobserved data 
points based on observed trends to provide more observations to a 
limited data set.
---------------------------------------------------------------------------

    Historical quarterly series for new light vehicle average price and 
total sales are presented in Figure VI-59 below. Due to the lack of 
data availability for business investment in light vehicles, the 
historical series for average vehicle price begins in 1987. Average 
prices were transformed into quality adjusted real terms using the CPI 
for new motor vehicles, and both series were seasonally adjusted.\1568\ 
Quality adjusted prices have risen overtime, while total sales have 
remained relatively flat in recent years with the major exception being 
the significant economic downturn of 2008-2009. The difference in these 
trends suggests that the number of vehicles purchased per

[[Page 24598]]

household does not necessarily change, or grow, over time, as income 
grows, but rather households adjust the ``amount'' of new vehicle they 
are willing to purchase (i.e., switching from sedan to an SUV).\1569\ 
Moreover, while disposable income has steadily increased during this 
period, sales have not seen the same type of upward trend, and instead 
only returned to its pre-recession average of around 17 million annual 
sales.
---------------------------------------------------------------------------

    \1568\ Seasonal adjustment was made using X.12 in EViews.
    \1569\ Aggregate light duty vehicle sales data does not allow 
for observing the distribution of vehicles being sold, which will 
have an effect on the average price.
[GRAPHIC] [TIFF OMITTED] TR30AP20.301

    Even as real disposable income has risen since 2000, and outside of 
the great recession, new vehicle sales have remained relatively steady. 
This, in turn, suggests there are other economic, or behavioral, 
factors beyond disposable income influencing the decision to purchase a 
new vehicle. Given the significant cost to purchase a new vehicle, and 
the long multiyear timeframe over which they are typically financed, 
households' forward-looking view on the health of the economy likely 
plays a role in their willingness to purchase a new vehicle. Put 
differently, households may delay their purchasing decisions if their 
view outlook on the economy sours, regardless of income level. These 
observations are consistent with the framework of the NPRM model, and 
Figure VI-60 presents the consumer sentiment index and total new sales, 
with both series exhibiting similar trends over this period. Some 
commenters advocated that consumer sentiment (also known as consumer 
confidence) should be included in the sales forecast. For example, the 
Aluminum Association indicated that prior sales models have shown 
consumer behavior to be ``highly sensitive to macroeconomic conditions, 
consumer confidence and employment levels.'' While consumer sentiment 
was not included in the NPRM model, it was included in specifications 
that the agencies tested and considered and is a component of the 
forecasting model used in the final rule.\1570\
---------------------------------------------------------------------------

    \1570\ Commenters mentioned consumer confidence as a predictor 
of consumer behavior. For instance, the Aluminum Association 
indicated that prior sales models have shown consumer behavior to be 
``highly sensitive to macroeconomic conditions, consumer confidence 
and employment levels.'' Comments, NHTSA-2018-0067-11952, at 14.

---------------------------------------------------------------------------

[[Page 24599]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.302

    All macroeconomic data were sourced from IHS including real 
disposable income, number of US households, and the University of 
Michigan's consumer sentiment index. The summary statistics for all 
series are presented below in Table VI-148.
[GRAPHIC] [TIFF OMITTED] TR30AP20.303

    Each series was transformed into natural logarithms and tested for 
stationarity using the modified Dicky-Fuller test.\1571\ Results 
presented in Table VI-149 indicate each variable containing contained a 
unit-root, while being differenced stationary (i.e., integrated of 
order one).
---------------------------------------------------------------------------

    \1571\ Using nonstationary variables would generate unreliable 
estimates of their influence, as prior values of those variables are 
correlated with their future values, and this violates the 
assumption that values variables take on are independent over time.

---------------------------------------------------------------------------

[[Page 24600]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.304

    Two separate variables lists were then tested for the existence of 
one or more cointegrating relationships, with results from the Johansen 
test presented in Table VI-150.\1572\ In each set of variables, both 
total LDV sales and disposable income were converted to household units 
as a means to control for the growth in US households and the possible 
decision making process of buying/consuming a new unit of LDV. The 
results show that 4 out of the 5 lag length selections for both 
variable sets conclude there being one cointegrating relationship (rank 
I(1)) among them.
---------------------------------------------------------------------------

    \1572\ The number of lag lengths were also tested formally, with 
general consensus between 2 and 6 lags as being optimal. Test 
results are available upon request, however, the final lag length 
selection was determined on the full set of VAR and VECM output that 
includes satisfying time series conditions such as no presence of 
autocorrelation and plausible interpretability of the estimated 
output.
[GRAPHIC] [TIFF OMITTED] TR30AP20.305

    Taken together, these tests confirm the need to address the time 
series properties of each variable in any modeling framework. This will 
become especially important when discussing the correct modeling 
approach, as The pre-modeling tests provide evidence against running a 
simple OLS regression or VAR in first differences, because doing so 
would have the potential outcome of excluding important long-run 
information.
    Furthermore, the endogeneity between vehicle sales and price is 
another element that needs to be considered for model specification. 
The IHS historical series for average price of a new light duty vehicle 
is defined as a

[[Page 24601]]

function of business and private residential spending on light vehicles 
divided by total new light vehicle sales; from this identity, the 
average price represents the nominal price per new unit of light duty 
vehicle sold. This definition supports the existence of an endogenous 
relationship between vehicle price and sales that needs to be accounted 
for when developing an econometric estimation of the influence of new 
vehicle price on sales. This is consistent with economic theory, 
whereby vehicle sales and price are simultaneously determined in the 
market, and therefore should be included together when specifying a 
forecasting equation.\1573\ This restriction holds even if nominal 
vehicle price is transformed into a quality adjusted real dollar 
series, as some commenters (EDF, Stock et al) proposed.\1574\
---------------------------------------------------------------------------

    \1573\ Endogeneity results in correlation between an independent 
variable in a regression and the error term leading to biased 
coefficient estimates.
    \1574\ For reference on how the BLS measures quality adjustments 
in vehicles: https://www.bls.gov/cpi/factsheets/new-vehicles.htm.
---------------------------------------------------------------------------

Models
    Faced with the simultaneity problem associated with price and 
sales, several specifications were reviewed to determine the best 
method for addressing this issue. An Instrumental Variable (IV) method 
was deemed the most direct approach, with the advantage of preserving 
the initial model's autoregressive distributed lag structure. In order 
to obtain consistent estimates of the price elasticity of demand, a 
suitable instrument that is correlated with average LDV price but 
uncorrelated with the error term is needed in the first stage. A 
suitable instrument must also make economic sense and have a plausible 
causal relationship. In theory, instruments that satisfy all three 
conditions (exogeneity, causality, and non-weak correlation) should 
exist. In practice, however, it is often prohibitively difficult to 
find a viable instrument. Both Stock et al. and CARB suggested 
instrumenting to resolve the endogeneity issue in the NPRM model, but 
neither suggested specific candidates for instrumental variables.
    For the purposes of modeling vehicle sales, candidate IVs would 
reflect the price of inputs to production that are broad enough, so 
that the underlying behavior of the variable is not deterministic of 
LDV sales. Examples of candidate variables include producer price 
indices (PPIs) of auto or other related manufacturing, cost of capital 
required for production, labor market data, energy costs, technology 
changes, and exogenous shocks to price, production, labor, or policy 
changes.
    The lack of data availability and quality concerns reduced the 
primary list of candidate IVs to relatable PPIs such as for 
manufacturing and automobile primary products. Even the most 
``promising'' candidate IVs, however, proved to be poor instruments, 
with counterintuitive signs, lack of statistical significance, and poor 
overall first stage F-statistics (even by relatively lenient weak 
instrument test standards).
    The lack of reasonable results from the IV approach led to testing 
vector autoregressive (VAR) and vector error correction (VECM) models. 
Relaxing the strict exogeneity assumption needed under an ARDL 
framework is the main advantage of modeling price, sales, and 
macroeconomic variables as a system of equations where the feedback 
from previous period shocks affect both price and sales.\1575\ In 
addition, a VAR or VECM can also adequately handle the time series and 
nonstationary properties discussed above. For both the VAR and VECM, a 
parsimonious specification was preferred with either a three or four 
variable system using the variables discussed above.
---------------------------------------------------------------------------

    \1575\ Strict exogeneity requires there to be past, 
contemporaneous, and future exogeneity between the variables of 
interest.
---------------------------------------------------------------------------

    We first estimated a simple VAR using a Wold causal ordering of 
real disposable income per household, average price of new LDV, and new 
total sales of LDVs per household.\1576\ The alternative specification 
included the consumer sentiment variable in the ordering the consumer 
sentiment variable after income and before price. This ordering assumes 
that households' disposable income (and consumer sentiment) do not 
respond to shocks to auto prices and sales within the same quarter. It 
also assumes that prices are contemporaneously exogenous of sales 
(demand), since the MSRPs are set in advance. Lastly, sales are able to 
respond to unexpected changes in price in the same quarter. The 
alternative ordering of placing sales before average price was deemed 
unrealistic as it would presume sales responding independently to an 
unexpected change in prices.
---------------------------------------------------------------------------

    \1576\ The Wold causal ordering creates a lower triangular 
matrix for our shocks, so by construction these shocks are 
orthogonal to each other to allow for causal inference. This 
recursive or Wold ordering technique should be predetermined and 
based on economic theory as the causal interpretation of the impulse 
responses are dependent on the correct/plausible ordering of 
variables.
---------------------------------------------------------------------------

    In the first specification, all variables were transformed to first 
differences to ensure stationarity, while ignoring any possible long-
run information (for the moment). A combination of post-estimation 
tests for autocorrelation and stability conditions were considered 
along with impulse response functions to gauge the model performance. 
The preferred model was estimated with five lags, and the impulse 
response functions (IRF) of a 1 percent shock to price on sales for the 
two specifications are presented in Figure VI-61.

[[Page 24602]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.306

    Both figures show a similar trend of the response in sales 
oscillating from negative to positive before ultimately returning to 
zero 12 quarters out. The three variable VAR sees a positive response 
in the first few periods, while the four variable VAR manages to dip 
below zero briefly after 4 periods out. This behavior, which by 
definition is short-run due to the differencing of the variables, could 
be representing auto dealerships' attempts to pull sales back to its 
equilibrium level after the price shock pushes sales negative, implying 
some level of over compensation during this process. Nonetheless, 
despite the model showing there is some evidence of an immediate and 
negative price elasticity, the overly simplified VAR model is missing 
key long run information (as identified in the cointegration tests), 
creating some reservations about the results. It is also worth noting 
that the lagged positive response in sales from an unexpected price 
shock is persistent regardless of the lag length selection, and in many 
cases even more pronounced.
    A number of preliminary conclusions can be drawn from the IRF 
results shown in Figure VI-61. First, at least at this level of 
aggregation, any short-run and immediate effect of a price increase on 
total LDV sales is relatively small in nature. This does not suggest, 
however, that the price elasticity of demand is zero. Instead, what may 
be the case is that when faced with an unexpected change in price, 
consumers will choose to purchase a less expensive car with fewer 
features as opposed to no car at all. In other words, the level of 
aggregation being used, total car sales, removes important variation 
between the type of vehicle being sold and consumer purchasing 
decisions from the data; what is left is a clouded version of the true 
relationship between price and sales. Second, this type of VAR ignores 
and throws out any long run information that may exist, which would 
create omitted variable bias if such a cointegrated relationship 
exists.
    Based on the conclusions from the Johansen cointegration test, the 
next step involved estimating the system as a VECM. As with the VAR 
models, the VECM employs either a three or four variable system with 
five lag lengths and an unconstrained constant in the model (no trend 
in either the first differenced or cointegrating equations). In each 
model, the cointegrating vector is normalized around sales (i.e., the 
sales' coefficient is set to 1), and the model results indicate strong 
evidence of a cointegrating relationship between the variables.
    Aside from general agreement on a cointegrating relationship, the 
VECM performance was weak in nearly every specification attempted, with 
implausible magnitudes for the long-run coefficient estimates and 
insignificant short-run dynamics. Moreover, the adjustment coefficient 
for the sales equation is particularly weak and insignificant.\1577\ 
The limitations of the VECM could be rooted in the system being 
normalized around sales, which lacks significant variation, 
correlation, or possibly true causation with the other variables.
---------------------------------------------------------------------------

    \1577\ The lack of a statistically significant adjustment 
variable could be an indication of weak exogeneity. In this case 
that would not be plausible given the clear endogeneity between 
price and sales, and is more likely an indication of poor data and 
the absence of reliable modelling approaches.
---------------------------------------------------------------------------

    As with the VAR analysis, a similar focus is placed on the IRFs 
presented in Figure VI-62. Here a one percent shock in price on LDV 
sales shows a similar response between the two specifications, with an 
increase during the first several periods before returning to a 
negative and permanent long-run effect. This response is erroneous in 
two ways: First, the sharp positive response during the first 8 to 10 
quarters defies economic logic as an increase in the price of a normal 
good should not induce an increase in sales. Second, the permanent and 
negative effect is equally as confounding because it rules out the 
ability for dealerships or auto manufacturers to adjust prices or 
supply.\1578\
---------------------------------------------------------------------------

    \1578\ Note that error bounds cannot be generated for VECM IRFs 
using most statistical packages, so determining statistical 
significance is difficult. Given the change from positive to 
negative and the low magnitude of the response, it is quite possible 
that this effect is indistinguishable from zero.
---------------------------------------------------------------------------

    The updated econometric models of light duty vehicle sales 
(described above) thus did not provide clear, significant or robust 
insight into the magnitude of the price elasticity of demand. While the 
VAR model specification points to an immediate short-run negative price 
elasticity of demand (i.e., sales fall in the face of an immediate 
price shock), this relationship is relatively small. In addition, the 
fact that this specification excludes the identified cointegration 
between the variables suggests that it is not robust or unbiased. In 
short, the VECM and IV approaches were unable to provide reasonable and 
meaningful results.
    These results strongly suggest that the relationship between sales 
and price is

[[Page 24603]]

not adequately estimated with the macro-level data used in this 
analysis. Recent peer reviewers of the CAFE model had similar concerns. 
In particular, these data are insufficient to explain the individual 
consumer (micro-) level decision making process of purchasing a new 
LDV. Aggregating the sales response to the national level reduces the 
useful variation in the decision making process to levels unsuitable 
for estimation. Commenters generally agreed with this conclusion.
    Even assuming a theoretically and econometrically correct model was 
possible, this relationship is impossible to evaluate at the current 
data aggregation level. Future research may focus on constructing an 
aggregate price elasticity of demand from consumer level data utilizing 
discrete choice modeling or something similar. However, constructing 
such models and integrating them into the simulations of the final rule 
are beyond the scope of this analysis.
[GRAPHIC] [TIFF OMITTED] TR30AP20.307

    Many commenters suggested that the NPRM model was unable to find a 
statistically significant influence of fuel economy on sales because 
the model was too highly aggregated, as the agencies found with the 
econometric experimentation to estimate a price response. EDF, CARB, 
and CA et al. and Oakland et al. expressed concern that using industry 
averages eliminated the variation needed to detect consumer valuation 
of fuel economy in new vehicle purchases. The agencies noted a similar 
concern in the NPRM, citing the level of aggregation as the most likely 
reason that the average fuel economy of a new vehicle was not a 
statistically significant explanatory variable in the ARDL model. The 
approach for the final rule includes an average value of improved fuel 
economy in the sales response, as commenters suggested it should.
(a) How Do Car and Light Truck Buyers Value Improved Fuel Economy?
    Many commenters (CARB, CA et al. and Oakland et al., NRDC, EDF, 
CBD, North Carolina Department of Environmental Quality, IPI, EPA 
Science Advisory Board, Stock et al.) stated that the agencies should 
explicitly consider fuel savings, and the value that consumers ascribe 
to it, in addition to changes in price when estimating the response of 
new vehicle sales to different regulatory alternatives. NRDC stated, 
``The decision between new vehicle purchase alternatives must consider 
both differential costs and differential benefits. The CAFE model sales 
algorithm considers only differential costs and is, therefore, 
flawed.'' \1579\ The agencies agree that the degree to which new 
vehicle buyers value improvements in fuel economy is an important 
consideration when estimating the response of new vehicle sales to 
potential standards. The effect of vehicle prices on sales is difficult 
to detect at the aggregate level because price movements are correlated 
with the current strength of the economy, which can appear as a 
positive price elasticity when modeling sales, and there are various 
technical econometric difficulties in identifying the effect of price 
on sales (simultaneity, cointegration, etc., addressed above). The 
sales response model in the final rule accounts for fuel savings 
realized by buyers of new vehicles.
---------------------------------------------------------------------------

    \1579\ NRDC, Attachment 3, NHTSA-2018-0067-11723, at 4.
---------------------------------------------------------------------------

    Some commenters and EPA's Science Advisory Board noted that the 
sales response equation omitted any value of fuel savings to new 
vehicle buyers, while other elements of the analysis--notably the 
technology application algorithm--assumed that buyers would demand fuel 
economy technologies that ``pay back'' within the first 2.5 years of 
ownership (as a result of avoided fuel costs), and manufacturers would 
supply fuel economy at those levels even in the absence of standards. 
This observation was made in comments by CARB, CBD, and IPI--the last 
of which stated that 2.5 year payback assumption ``clashes directly 
with the contradictory assumption that the agencies rely on in the 
model's sales module, where they implicitly assume that customers 
entirely disregard fuel efficiency in their purchasing decisions.'' 
\1580\ The agencies agree that this represented an internal 
inconsistency. The sales model used to analyze the final rule includes 
the estimated value of fuel savings to vehicle buyers, and is 
consistent with other assumptions throughout the analysis about the 
``pay back'' period.
---------------------------------------------------------------------------

    \1580\ IPI, Appendix, NHTSA-2018-0067-12213, at 16.
---------------------------------------------------------------------------

    How potential buyers value improvements in the fuel economy of new 
cars and light trucks is an important issue in assessing the benefits 
and costs of government regulation. If buyers fully value the savings 
in fuel costs that result from higher fuel

[[Page 24604]]

economy, manufacturers will presumably supply any improvements that 
buyers demand, and vehicle prices will fully reflect future fuel cost 
savings consumers would realize from owning--and potentially re-
selling--more fuel-efficient models. If consumers internalize fuel 
savings this case, more stringent fuel economy standards will impose 
net costs on vehicle owners and can only result in social benefits 
through correcting externalities, because consumers would already fully 
incorporate private savings into their purchase decisions, as discussed 
further below. If instead consumers systematically undervalue some 
market failure such as an information asymmetry leads to an 
underinvestment in fuel-saving technology, the cost savings generated 
by improvements in fuel economy when choosing among competing models, 
more stringent fuel economy standards will also lead manufacturers to 
adopt improvements in fuel economy that buyers might not choose despite 
the cost savings they offer and improve consumer welfare.
    The potential for car buyers voluntarily to forego improvements in 
fuel economy that offer savings exceeding their initial costs is one 
example of what is often termed the ``energy-efficiency gap.'' This 
appearance of such a gap, between the level of energy efficiency that 
would minimize consumers' overall expenses and what they actually 
purchase, is typically based on engineering calculations that compare 
the initial cost for providing higher energy efficiency to the 
discounted present value of the resulting savings in future energy 
costs.
    There has long been an active debate about why such a gap might 
arise and whether it actually exists. Economic theory predicts that 
individuals will purchase more energy-efficient products only if the 
savings in future energy costs they offer promise to offset their 
higher initial costs. However, the additional up-front cost of a more 
energy-efficient product includes more than just the cost of the 
technology necessary to improve its efficiency; because consumers have 
a scarcity of resources, it also includes the opportunity cost of any 
other desirable features that consumers give up when they choose the 
more efficient alternative. In the context of vehicles, whether the 
expected fuel savings outweigh the opportunity cost of purchasing a 
model offering higher fuel economy will depend, among other things, on 
how much its buyer expects to drive, his or her expectations about 
future fuel prices, the discount rate he or she uses to value future 
expenses, the expected effect on resale value, and whether more 
efficient models offer equivalent attributes such as performance, 
carrying capacity, reliability, quality, or other characteristics.
    Published literature has offered little consensus about consumers' 
willingness-to-pay for greater fuel economy, and whether it implies 
over- under- or full-valuation of the expected discounted fuel savings 
from purchasing a model with higher fuel economy. Most studies have 
relied on car buyers' purchasing behavior to estimate their 
willingness-to-pay for future fuel savings; a typical approach has been 
to use ``discrete choice'' models that relate individual buyers' 
choices among competing vehicles to their purchase prices, fuel 
economy, and other attributes (such as performance, carrying capacity, 
and reliability), and to infer buyers' valuation of higher fuel economy 
from the relative importance of purchase prices and fuel economy.\1581\ 
Empirical estimates using this approach span a wide range, extending 
from substantial undervaluation of fuel savings to significant 
overvaluation, thus making it difficult to draw solid conclusions about 
the influence of fuel economy on vehicle buyers' choices.\1582\ Because 
a vehicle's price is often correlated with its other attributes (both 
measured and unobserved), analysts have often used instrumental 
variables or other approaches to address endogeneity and other 
resulting concerns.\1583\
---------------------------------------------------------------------------

    \1581\ In a typical vehicle choice model, the ratio of estimated 
coefficients on fuel economy--or more commonly, fuel cost per mile 
driven--and purchase price is used to infer the dollar value buyers 
attach to slightly higher fuel economy.
    \1582\ See Helfand & Wolverton (2011) and Green (2010) for 
detailed reviews of these cross-sectional studies.
    \1583\ See, e.g., Barry, et al. (1995).
---------------------------------------------------------------------------

    Despite these efforts, more recent research has criticized these 
cross-sectional studies; some have questioned the effectiveness of the 
instruments they use,\1584\ while others have observed that 
coefficients estimated using non-linear statistical methods can be 
sensitive to the optimization algorithm and starting values.\1585\ 
Collinearity (i.e., high correlations) among vehicle attributes--most 
notably among fuel economy, performance or power, and vehicle size--and 
between vehicles' measured and unobserved features also raises 
questions about the reliability and interpretation of coefficients that 
may conflate the value of fuel economy with other attributes (Sallee, 
et al., 2016; Busse, et al., 2013; Allcott & Wozny, 2014; Allcott & 
Greenstone, 2012; Helfand & Wolverton, 2011).
---------------------------------------------------------------------------

    \1584\ See Allcott & Greenstone (2012).
    \1585\ See Knittel & Metaxoglou (2014).
---------------------------------------------------------------------------

    In an effort to overcome shortcomings of past analyses, three 
studies published fairly recently rely on panel data from sales of 
individual vehicle models to improve their reliability in identifying 
the association between vehicles' prices and their fuel economy 
(Sallee, et al. 2016; Allcott & Wozny, 2014; Busse, et al., 2013). 
Although they differ in certain details, each of these analyses relates 
changes over time in individual models' selling prices to fluctuations 
in fuel prices, differences in their fuel economy, and increases in 
their age and accumulated use, which affects their expected remaining 
life, and thus their market value. Because a vehicle's future fuel 
costs are a function of both its fuel economy and expected gasoline 
prices, changes in fuel prices have different effects on the market 
values of vehicles with different fuel economy; comparing these effects 
over time and among vehicle models reveals the fraction of changes in 
fuel costs that is reflected in changes in their selling prices 
(Allcott & Wozny, 2014). Using very large samples of sales enables 
these studies to define vehicle models at an extremely disaggregated 
level, which enables their authors to isolate differences in their fuel 
economy from the many other attributes, including those that are 
difficult to observe or measure, that affect their sale prices.\1586\
---------------------------------------------------------------------------

    \1586\ These studies rely on individual vehicle transaction data 
from dealer sales and wholesale auctions, which includes actual sale 
prices and allows their authors to define vehicle models at a highly 
disaggregated level. For instance, Allcott & Wozny (2014) 
differentiate vehicles by manufacturer, model or nameplate, trim 
level, body type, fuel economy, engine displacement, number of 
cylinders, and ``generation'' (a group of successive model years 
during which a model's design remains largely unchanged). All three 
studies include transactions only through mid-2008 to limit the 
effect of the recession on vehicle prices. To ensure that the 
vehicle choice set consists of true substitutes, Allcott & Wozny 
(2014) define the choice set as all gasoline-fueled light-duty cars, 
trucks, SUVs, and minivans that are less than 25 years old (i.e., 
they exclude vehicles where the substitution elasticity is expected 
to be small). Sallee et al. (2016) exclude diesels, hybrids, and 
used vehicles with less than 10,000 or more than 100,000 miles.
---------------------------------------------------------------------------

    These studies point to a somewhat narrower range of estimates than 
suggested by previous cross-sectional studies; more importantly, they 
consistently suggest that buyers value a large proportion--and perhaps 
even all--of the future savings that models with higher fuel economy 
offer.\1587\

[[Page 24605]]

Because they rely on estimates of fuel costs over vehicles' expected 
remaining lifetimes, these studies' estimates of how buyers value fuel 
economy are sensitive to the strategies they use to isolate differences 
among individual models' fuel economy, as well as to their assumptions 
about buyers' discount rates and gasoline price expectations, among 
others. Since Anderson et al. (2013) found evidence that consumers 
expect future gasoline prices to resemble current prices, the agencies 
use this assumption to compare the findings of the three studies and 
examine how their findings vary with the discount rates buyers apply to 
future fuel savings.\1588\
---------------------------------------------------------------------------

    \1587\ Killian & Sims (2006) and Sawhill (2008) rely on similar 
longitudinal approaches to examine consumer valuation of fuel 
economy except that they use average values or list prices instead 
of actual transaction prices. Since these studies remain 
unpublished, their empirical results are subject to change, and they 
are excluded from this discussion.
    \1588\ Each of the studies makes slightly different assumptions 
about appropriate discount rates. Sallee et al. (2016) use five 
percent in their base specification, while Allcott & Wozny (2014) 
rely on six percent. As some authors note, a five to six percent 
discount rate is consistent with current interest rates on car 
loans, but they also acknowledge that borrowing rates could be 
higher in some cases, which could be used to justify higher discount 
rates. Rather than assuming a specific discount rate, Busse et al. 
(2013) directly estimate implicit discount rates at which future 
fuel costs would be fully internalized; they find discount rates of 
six to 21 percent for used cars and one to 13 percent for new cars 
at assumed demand elasticities ranging from -2 to -3. Their 
estimates can be translated into the percent of fuel costs 
internalized by consumers, assuming a particular discount rate. To 
make these results more directly comparable to the other two 
studies, we assume a range of discount rates and uses the authors' 
spreadsheet tool to translate their results into the percent of fuel 
costs internalized into the purchase price at each rate. Because 
Busse et al. (2013) estimate the effects of future fuel costs on 
vehicle prices separately by fuel economy quartile, these results 
depend on which quartiles of the fuel economy distribution are 
compared; our summary shows results using the full range of quartile 
comparisons.
---------------------------------------------------------------------------

    As Table VI-148 indicates, Allcott & Wozny (2014) found that 
consumers incorporate 55% percent of future fuel costs into vehicle 
purchase decisions at a six percent discount rate, when their 
expectations for future gasoline prices are assumed to reflect 
prevailing prices at the time of their purchases. With the same 
expectation about future fuel prices, the authors report that consumers 
would fully value fuel costs only if they apply discount rates of 24 
percent or higher. However, these authors' estimates are closer to full 
valuation when using gasoline price forecasts that mirror oil futures 
markets, because the petroleum market expected prices to fall during 
this period (this outlook reduces the discounted value of a vehicle's 
expected remaining lifetime fuel costs). With this expectation, Allcott 
& Wozny (2014) find that buyers value 76 percent of future cost savings 
(discounted at six percent) from choosing a model that offers higher 
fuel economy, and that a discount rate of 15 percent would imply that 
they fully value future cost savings. Sallee et al. (2016) begin with 
the perspective that buyers fully internalize future fuel costs into 
vehicles' purchase prices and cannot reliably reject that hypothesis; 
their base specification suggests that changes in vehicle prices 
incorporate slightly more than 100 percent of changes in future fuel 
costs. For discount rates of five to six percent, the Busse et al. 
(2013) results imply that vehicle prices reflect 60 to 100 percent of 
future fuel costs. As Table VI-151 suggests, higher private discount 
rates move all of the estimates closer to full valuation or to over-
valuation, while lower discount rates imply less complete valuation in 
all three studies.
[GRAPHIC] [TIFF OMITTED] TR30AP20.308

    The studies also explore the sensitivity of the results to other 
parameters that could influence their results. Busse et al. (2013) and 
Allcott & Wozny (2014) find that relying on data that suggest lower 
annual vehicle use or survival probabilities, which imply that vehicles 
will not last as long, moves their estimates closer to full valuation, 
an unsurprising result because both reduce the changes in expected 
future fuel costs caused by fuel

[[Page 24606]]

price fluctuations. Allcott & Wozny's (2014) base results rely on an 
instrumental variables estimator that groups miles-per-gallon (MPG) 
into two quantiles to mitigate potential attenuation bias due to 
measurement error in fuel economy, but they find that greater 
disaggregation of the MPG groups implies greater undervaluation (for 
example, it reduces the 55 percent estimated reported in Table VI-148 
to 49 percent). Busse et al. (2013) allow gasoline prices to vary 
across local markets in their main specification; using national 
average gasoline prices, an approach more directly comparable to the 
other studies, results in estimates that are closer to or above full 
valuation. Sallee et al. (2016) find modest undervaluation by vehicle 
fleet operators or manufacturers making large-scale purchases, compared 
to retail dealer sales (i.e., 70 to 86 percent).
    Since they rely predominantly on changes in vehicles' prices 
between repeat sales, most of the valuation estimates reported in these 
studies apply most directly to buyers of used vehicles. Only Busse et 
al. (2013) examine new vehicle sales; they find that consumers value 
between 75 to 133 percent of future fuel costs for new vehicles, a 
higher range than they estimate for used vehicles. Allcott & Wozny 
(2014) examine how their estimates vary by vehicle age and find that 
fluctuations in purchase prices of younger vehicles imply that buyers 
whose fuel price expectations mirror the petroleum futures market value 
a higher fraction of future fuel costs: 93 percent for one- to three-
year-old vehicles, compared to their estimate of 76 percent for all 
used vehicles assuming the same price expectation.\1589\
---------------------------------------------------------------------------

    \1589\ Allcott & Wozny (2014) and Sallee, et al. (2016) also 
find that future fuel costs for older vehicles are substantially 
undervalued (26-30%). The pattern of Allcott and Wozny's results for 
different vehicle ages is similar when they use retail transaction 
prices (adjusted for customer cash rebates and trade-in values) 
instead of wholesale auction prices, although the degree of 
valuation falls substantially in all age cohorts with the smaller, 
retail price based sample.
---------------------------------------------------------------------------

    Accounting for differences in their data and estimation procedures, 
the three studies described here suggest that car buyers who use 
discount rates of five to six percent value at least half--and perhaps 
all--of the savings in future fuel costs they expect from choosing 
models that offer higher fuel economy. Perhaps more important in 
assessing the case for regulating fuel economy, one study (Busse et 
al., 2013) suggests that buyers of new cars and light trucks value 
three-quarters or more of the savings in future fuel costs they 
anticipate from purchasing higher-mpg models, although this result is 
based on more limited information.
    In contrast, previous regulatory analyses of fuel economy standards 
implicitly assumed that buyers undervalue even more of the benefits 
they would experience from purchasing models with higher fuel economy, 
so that, without increases in fuel economy standards, little 
improvement would occur, and the entire value of fuel savings from 
raising CAFE standards represented private benefits to car and light 
truck buyers themselves. For instance, in the EPA analysis of the 2017-
2025 model year CO2 standards, fuel savings alone added up 
to $475 billion (at three percent discount rate) over the lifetime of 
the vehicles, far outweighing the compliance costs: $150 billion). The 
assertion that buyers were unwilling to take voluntary advantage of 
this opportunity implies that collectively, they must have valued less 
than a third ($150 billion/$475 billion = 32 percent) of the fuel 
savings that would have resulted from those standards. In fact, those 
earlier analyses assumed that new car and light truck buyers attach 
relatively little value to higher fuel economy, since their baseline 
scenarios assumed that fuel economy levels would not increase in the 
absence of progressively tighter standards, despite increasing fuel 
prices. The evidence reviewed here makes that perspective extremely 
difficult to justify and would call into question any analysis that 
claims to show large private net benefits for vehicle buyers 
attributable to increases in fuel economy standards.
    What analysts assume about consumers' vehicle purchasing behavior, 
particularly about potential buyers' perspectives on the value of 
increased fuel economy, clearly matters a great deal in the context of 
benefit-cost analysis for fuel economy regulation. In light of this 
recent evidence on this question, warrants a more nuanced approach that 
is more nuanced than merely assuming that buyers drastically undervalue 
benefits from higher fuel economy, (and that, as a consequence, these 
benefits are unlikely to be realized without stringent fuel economy 
standards,) seems warranted. One possible approach would be to use a 
baseline scenario where fuel economy levels of new cars and light 
trucks reflected full (or nearly so) valuation of fuel savings by 
potential buyers in order to reveal whether setting fuel economy 
standards above market-determined levels could produce net social 
benefits. Another might be to assume that, unlike in the agencies' 
previous analyses, where buyers were assumed to greatly to undervalue 
higher fuel economy under the baseline but to value it fully under the 
proposed standards, buyers value improved fuel economy identically 
under both the baseline scenario and with stricter CAFE standards in 
place.
    The agencies requested comment on the consumer valuation of fuel 
economy and its use in the NPRM analysis. CBD and the North Carolina 
Department of Environmental Quality took issue with the agencies' 
characterization of the literature on the value of fuel economy, citing 
EPA's previous determination that the estimates in the literature 
represented too large a range, and the degree of uncertainty made 
including a value of fuel economy challenging. This final rule analysis 
accounts for the value of fuel economy in several places, though it 
uses a more conservative value than is suggested by the literature 
summarized above. Manufacturers have consistently told the agencies 
that new vehicle buyers will pay for about 2 or 3 years' worth of fuel 
savings before the price increase associated with providing those 
improvements begins to impact affect sales. The agencies have assumed 
the same valuation, 2.5 years, in all components of the analysis that 
reflect consumer decisions regarding vehicle purchases and 
retirements.\1590\ This analysis explicitly assumes that: (1) Consumers 
are willing to pay for fuel economy improvements that pay back within 
the first 2.5 years of vehicle ownership (at average usage rates); (2) 
manufacturers know this and will provide these improvements even in the 
absence of regulatory pressure; (3) potential buyers weigh these 
savings against increases in new vehicle prices when deciding to retire 
a vehicle; and (4) the amount of technology for which buyers will pay 
rises (or falls) with rising (or falling) fuel prices.\1591\ Excluding 
the value of fuel economy entirely from these calculations does not 
remove it from the analysis; it merely imposes an implausibly low value 
on the desired payback period of new

[[Page 24607]]

vehicle buyers and manufacturers--regardless of fuel prices or 
technology costs. And while the agencies acknowledge the uncertainty 
around the estimates in the literature, zero is far removed from the 
lower bounds of any study.
---------------------------------------------------------------------------

    \1590\ When accounting for social benefits and costs associated 
with an alternative, the full lifetime value of fuel savings is 
included.
    \1591\ NADA, the Alliance of Automobile Manufacturers, and 
American Fuel and Petrochemical Manufacturers argued that CAFE/
CO2 standards have already reached the point where the 
price increases necessary to recoup manufacturers' increased costs 
for providing further increases in fuel economy outweigh the value 
of fuel savings, and requiring further increases in fuel economy 
will reduce new vehicle sales. The sales response in the final rule 
recognizes and incorporates the effect of fuel prices and fuel 
economy on new vehicle purchases. See NADA, NHTSA-2018-0067-12064, 
at 11; Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073 at 
163-64; AMFP, Comments, NHTSA-2018-0067-12078-29,at 3.
---------------------------------------------------------------------------

    CARB asserted that the various market failures suggested by the 
agencies in past rules (lack of information about fuel savings from 
higher MPG, inability to calculate cost savings from higher MPG, loss 
aversion, first-mover disadvantage), together with advertising that 
only emphasizes fuel economy during periods of high fuel prices, leads 
buyers to undervalue fuel economy.\1592\ In contrast, CARB (and 
others--such as SCAQMD, Alliance to Save Energy, Save EPA, AAA, 
Environmental group coalition, Consumers Union, EDF, and IPI) argues 
elsewhere that new vehicle buyers do value fuel economy highly, and 
nearly fully once fuel prices return to ``normal'' levels.\1593\ The 
agencies' payback period assumption, and the matching adjustment it 
makes to changes in new car prices to account for accompanying changes 
in fuel economy, recognizes that on average potential car buyers value 
a significant share of lifetime cost savings resulting from higher fuel 
economy. The agencies considered longer payback periods along the lines 
suggested by Consumer Federation of America (CFA),\1594\ but chose 2.5 
years as a conservative approach. Our assumption is consistent with 
survey evidence cited by the commenters, but at odds with their 
assertions that this program is necessary to save buyers from their own 
limited ability to make decisions in their best interest.
---------------------------------------------------------------------------

    \1592\ See CARB, Detailed Comments, NHTSA-2018-0067-11873 at 
212-16.
    \1593\ E.g. id. at 190-91. See also, id. at 188-89. See also, 
SCAQMD, Supplemental comments, NHTSA-2018-0067-11813, at 4-5; 
Alliance to Save Energy, Comment, NHTSA-2018-0067-11837, at 2; Save 
EPA, Comments, NHTSA-2018-0067-11930, at 6; AAA, Comments, NHTSA-
2018-0067-11979, at 2-3; Environmental group coalition, Appendix A, 
NHTSA-2018-0067-12000, at 54-56; Consumers Union, Attachment A, 
NHTSA-2018-0067-12068, 27-29; EDF, Appendix B, NHTSA-2018-0067-
12108, at 84-86; and IPI, Appendix, NHTSA-2018-0067-12213, at 40-47.
    \1594\ CFA, Comments, NHTSA-2018-0067-12005, at 12.
---------------------------------------------------------------------------

    More recently, the agencies have justified stricter CAFE and 
CO2 emissions standards by asserting that buyers do not take 
advantage of opportunities to improve their own well-being, by 
purchasing models whose higher fuel economy would more than repay their 
higher initial purchase prices via future savings in fuel costs. This 
newer rationale is fundamentally different from asserting that some 
externality--whereby buyers' choices cause economic harm to others--
exists to justify regulating fuel economy or CO2 emissions, 
or adopting more demanding regulations. EPA and NHTSA have previously 
labeled this behavior an example of the ``energy paradox,'' whereby 
consumers voluntarily forego investments that conserve energy even when 
those initial outlays appear likely to repay themselves--in the form of 
savings in energy costs--over the relatively near term.\1595\
---------------------------------------------------------------------------

    \1595\ See, e.g., EPA Regulatory Impact Analysis: Final 
Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission 
Standards and Corporate Average Fuel Economy Standards, available at 
https://nepis.epa.gov/Exe/ZyPDF.cgi/P100EZI1.PDF?Dockey=P100EZI1.PDF.
---------------------------------------------------------------------------

    However, recent research cast doubt on whether such an energy 
paradox exists in the case of fuel economy--that is, on whether buyers 
of new vehicles inadequately consider the value of future savings in 
fuel costs they would experience from purchasing models that feature 
higher fuel economy--and about how extensive it might be. Several 
recent studies have estimated the fraction of appropriately discounted 
lifetime fuel savings offered by models featuring higher fuel economy 
that car shoppers appear to value or willing to pay for. These 
estimates are typically drawn from one of three sources--(1) buyers' 
choices among competing models with different purchase prices, fuel 
economy levels, and other features; (2) statistically ``decomposing'' 
vehicle prices into the values buyers attach to their individual 
features, one of which is fuel economy; or (3) analyzing how selling 
prices for vehicles with different fuel economy levels respond to 
variation in fuel prices and the changes it causes in their lifetime 
fuel costs.
    The estimates these studies report may partly reflect variation 
among buyers' preferences for different vehicle features (such as fuel 
economy, but also size or utility), the financial constraints they 
face, how much they drive, or their expectations about future fuel 
prices, so they should be interpreted cautiously. However, the most 
careful recent studies suggest that on average buyers appear to 
undervalue the savings from higher fuel economy at most modestly, and 
perhaps not at all, after accounting for the influence of vehicles' 
other attributes on prices and purchasing decisions.\1596\ This 
research suggests that the energy paradox, sometimes described as 
buyers' ``myopia'' in assessing the value of future fuel savings, is a 
much weaker rationale for regulating fuel economy than the agencies had 
previously asserted.
---------------------------------------------------------------------------

    \1596\ For a review of these recent studies, see Table VI-120--
Percent of Future Fuels Costs Internalized in Used Vehicle Purchase 
Price using Current Gasoline Prices to Reflect Expectations (for 
Base Case Assumptions).
---------------------------------------------------------------------------

    IPI commented that the agencies' obligation to consider market 
failures in setting standards derives not just from Executive Order 
12,866 but also from the agencies' respective statutes, and argued that 
the agencies had defined market failures too narrowly in their 
proposal.\1597\ Specifically, IPI stated that NHTSA's task under EPCA 
is ``not so restricted to only protecting consumers from gas price 
spikes,'' and argued that NHTSA must also consider ``externalities 
relating to energy security, national security, positional goods, 
global climate change, and air and water pollution associated with fuel 
production and consumption; asymmetric information, attention costs, 
and other information failures; internalities, including myopia; and 
various supply-side market failures, including first-mover 
disadvantage.'' \1598\
---------------------------------------------------------------------------

    \1597\ IPI, Appendix, NHTSA-2018-0067-12213, at 9-10.
    \1598\ Id.
---------------------------------------------------------------------------

    For EPA's task under the CAA, IPI stated that, although while EPA 
must ``protect the planet from unchecked climate change, [it] must not 
ignore other related market failures that cause harm to public health 
and welfare, including the issues and market failures [as described for 
NHTSA above].'' \1599\ IPI argued that the proposal was arbitrary and 
capricious for not ``consider[ing] important aspects of the problem set 
before the agencies by Congress,'' and also for not considering the 
market failures discussed in the 2012 final rule.\1600\ CBD, et al., 
asserted similarly that the agencies' respective statutes require their 
actions to be more technology-forcing than what markets would otherwise 
achieve, in effect asserting that innovations in technology confer 
external benefits that vehicle manufacturers or buyers do not fully 
consider.\1601\
---------------------------------------------------------------------------

    \1599\ Id.
    \1600\ Id.
    \1601\ CBD, et al., NHTSA-2018-0067-12057, at 2 and 9.
---------------------------------------------------------------------------

    With regard to the specific market failures CAFE and CO2 
standards could potentially address, Global Automakers suggested that 
climate effects are indeed an externality that more stringent standards 
can address,\1602\ while CFA stated that regulating fuel economy and 
CO2 emissions can address an extensive catalog of market 
failures, including externalities, marketing, availability of

[[Page 24608]]

fuel-efficient models, transaction cost friction, information 
asymmetry, behavioral issues, and access to capital, among 
others.\1603\ CFA asserted that advances in economic theory had heavily 
criticized the neoclassical model, and that ``a great deal of empirical 
evidence supports [that the] standards are seen as an important and, in 
many ways, preferred policy approach.'' \1604\ On this basis, CFA 
stated that attribute-based standards that ``are set at a moderately 
aggressive level'' and are ``consistent with the rate of improvement 
that the auto industry achieved in the first decade of the fuel economy 
standard setting program,'' among other things, would address the 
market failure.\1605\
---------------------------------------------------------------------------

    \1602\ Global Automakers, Attachment A, NHTSA-2018-0067-12032, 
at A-22.
    \1603\ CFA, Comments, NHTSA-2018-0067-12005, at 61-64.
    \1604\ Id. at 63.
    \1605\ Id. at 64.
---------------------------------------------------------------------------

    IPI argued that regulation of fuel economy (presumably also 
CO2 emissions) is necessary because ``many vehicle 
attributes, like horsepower and size, are positional goods--that is, 
they confer status on buyers of cars and light truck models that 
feature them prominently, so regulation of fuel economy can help 
correct the positional externality.'' \1606\ IPI also noted the 
externality of health effects associated with refueling. IPI cited 
Alcott and Sunstein (2015) to argue, like CFA, that fuel economy 
standards can correct market failures like informational failure, 
myopia, supply-side failures, positional externalities, etc., and by 
doing so, can provide net private welfare gains--that is, improve the 
utility of vehicle buyers themselves, not just that of other households 
or businesses.\1607\
---------------------------------------------------------------------------

    \1606\ IPI, Appendix, NHTSA-2018-0067-12213, at 33.
    \1607\ Id. at 34. Note, however, that the reference cited does 
not address the question of whether fuel economy standards can be 
effective in correcting those market failures. Instead, it explores 
the circumstances under which fuel economy standards can improve 
welfare when vehicle buyers undervalue savings in fuel costs from 
purchasing more fuel-efficient models. See generally, Allcott, Hunt, 
and Cass R. Sunstein, ``Regulating Internalities,'' Working Paper 
20087, National Bureau of Economic Research, May 2015, available at 
https://www.nber.org/papers/w21187.pdf.
---------------------------------------------------------------------------

    EDF and CARB both asserted that an energy paradox exists in the 
case of fuel economy, with EDF arguing (like CFA) that information 
asymmetry--that is, unequal access of vehicle manufacturers and 
potential buyers to information about the cost savings likely to result 
from owning higher-mpg models--coupled with limited availability of 
fuel-efficient models, leads consumers to purchase vehicles with lower 
fuel economy than they otherwise would.\1608\ CARB simply stated that 
the NPRM analysis did not account for the energy paradox.\1609\
---------------------------------------------------------------------------

    \1608\ EDF, Appendix B, NHTSA-2018-0067-12108, at 88-89.
    \1609\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 188-
89.
---------------------------------------------------------------------------

    The agencies agree with these commenters that the market failures 
CAFE and CO2 standards can help address are likely to exist, 
but note that little of the behavior in the broad catalog identified by 
commenters actually represents market failures, and instead simply 
reflects consumers' preferences for features other than fuel economy. 
Even in the few cases of potential market failures that commenters 
identify related to the hypothetical energy paradox, the agencies 
question whether more stringent CAFE and CO2 standards are 
necessary to address the phenomena, or are even likely to be effective 
in doing so. In the agencies' view, neither the logical arguments nor 
the limited empirical evidence that commenters presented convincingly 
demonstrate the capacity of more stringent CAFE and CO2 
standards to resolve, or even mitigate, most of the various phenomena 
they describe as market failures.
    For example, the idea that regulating fuel economy and 
CO2 emissions can mitigate the consequences of inadequate 
access to information by placing decisions that depend on access to 
complete information in the hands of regulators rather than buyers has 
superficial appeal. Yet commenters do not establish that such a drastic 
step is necessary to overcome any inadequacy of information, or that 
requiring manufacturers to supply higher fuel economy will be more 
effective than less intrusive approaches such as expanding the range of 
information available to buyers. As OMB Circular A-4 notes, ``Because 
information, like other goods, is costly to produce and disseminate, 
your evaluation will need to do more than demonstrate the possible 
existence of incomplete or asymmetric information.'' \1610\
---------------------------------------------------------------------------

    \1610\ Circular A-4, at 5.
---------------------------------------------------------------------------

    In the few cases where commenters cited empirical evidence to 
support their arguments that stricter fuel economy and CO2 
regulations are an appropriate response to market failures, that 
evidence is limited and unpersuasive. As one illustration, the frequent 
assertion that buyers' widespread aversion to the prospect of financial 
losses makes them hesitant to purchase higher-mpg models appears to be 
traceable to findings from classroom experiments on small numbers of 
university students, rather than to large-scale empirical evidence 
drawn from buyers' observed behavior.\1611\ Commenters' repeated 
emphasis on loss aversion as a critical source of buyers' unwillingness 
to choose levels of fuel economy that appear to be in their own 
financial interest also ignores recent research questioning whether 
loss aversion is a plausible motivation for such systematic or 
universal behavior by consumers.\1612\
---------------------------------------------------------------------------

    \1611\ CFA, Comments, NHTSA-2018-0067-12005, at 16 et seq; 
Consumers Union, Attachment 4, NHTSA-2018-0067-12068, at 12; 
Attachment 3, NHTSA-2018-0067-11741, at 5-6, CARB at 214, and States 
at 87 each assert that loss aversion is an important source of car 
buyers' hesitance to purchase higher-mpg models, variously citing 
Greene, David L., John German, and Mark A. Delucchi, ``Fuel Economy: 
The Case for Market Failure,'' Reducing Climate Impacts in the 
Transportation Sector, Springerin James S. Cannon and Daniel 
Sperling, eds., Springer, 2009, at pp. 181-205; (2009); Greene, 
David L. (2010). How consumers value fuel economy: A literature 
review (No. EPA-420-R-10-008); Greene, David L., ``Uncertainty, Loss 
Aversion and Markets for Energy Efficiency,'' Energy Economics, vol. 
33, at pp. 608-616, (2011) and Greene, David L., ``Consumers' 
Willingness to Pay for Fuel Economy: Implications for Sales of New 
Vehicles and Scrappage of Used Vehicles,'' attachment to comments by 
CARB, Oct. 10, 2018. However, none of these sources presents 
empirical evidence on how the frequency of actual common loss 
aversion actually is among real world vehicle buyers, instead simply 
asserting (or implicitly assuming) that loss aversion it is likely 
to be widespread. Further, their (identical) estimates of the degree 
of loss aversion are difficult to trace, and appear to be drawn from 
classroom exercises administered to limited numbers of university 
students, not from empirical research involving real world vehicle 
buyers. One source cited for their repeated assertion that losses of 
a given dollar amount are valued twice as highly as gains of the 
same amount is Gal, David, ``A psychological law of inertia and the 
illusion of loss aversion,'' Judgment and Decision Making, Vol. 1, 
No. 1, at pp. 23-32 (July 2006,), pp. 23-32, but this reference does 
not report such a value. Another source repeatedly cited by Greene 
and co-authors, Benartzi, Shlomo, and Richard H. Thaler, ``Myopic 
Loss Aversion and the Equity Premium Puzzle,'' Quarterly Journal of 
Economics, Vol. 110, No. 1, at pp. 73-92 (February 1995), pp. 73-92, 
does report this value (at p. 74), although only in passing, and 
cites other references as its original source. The original sources 
of the claim that losses are values twice as highly as equivalent 
gains appear to be Kahneman, Daniel, Jack L. Knetsch, and Richard H. 
Thaler, ``Experimental Tests of the Endowment Effect and the Coase 
Theorem,'' Journal of Political Economy, Vol. 98, No. 6, pp. 1325-
48. (Dec., 1990) (pp. 1325-1348, specifically Section II), pp. 1329-
1336; and Tversky, Amos, and Daniel Kahneman, ``Loss Aversion in 
Riskless Choice: A Reference-Dependent Model,'' Quarterly Journal of 
Economics, Vol. 106, No. 4, at pp. 1039-61 (Nov., 1991) (pp. 1039-
1061, specifically pp. 1053-1054). Neither of these references, 
however, makes any claim about the generality of the estimate or its 
applicability to non-experimental settings for consumer behavior.
    \1612\ See Gal, David, ``A psychological law of inertia and the 
illusion of loss aversion,'' Judgment and Decision Making, Vol. 1, 
No. 1, pp. 23-32 (July 2006,) pp. 23-32,; Erev, I., E. Ert, and E. 
Yechiam, ``Loss aversion, diminishing sensitivity, and the effect of 
experience on repeated decisions.'', Journal of Behavioral Decision 
Making, Vol. 21 (2008), pp. 575-97; (2008); Ert, E., and I. Erev, 
``On the descriptive value of loss aversion in decisions under risk: 
Six clarifications,'' Judgment and Decision Making, Vol. 8 (2013), 
at pp. 214-35; (2013); Gal, David and Rucker, Derek, ``The Loss of 
Loss Aversion: Will It Loom Larger Than Its Gain?'' Journal of 
Consumer Psychology, Vol. 28 No. 3, (July 2018), at pp. 497-516 
(July 2018) available at (https://onlinelibrary.wiley.com/doi/abs/10.1002/jcpy.1047); and Gal, David, ``Why the Most Important Idea in 
Behavioral Decision-Making Is a Fallacy,'' Scientific American, 
Observations, (July 31, 2018), available at (https://blogs.scientificamerican.com/observations/why-the-most-important-idea-in-behavioral-decision-making-is-a-fallacy/).

---------------------------------------------------------------------------

[[Page 24609]]

    Another example is commenters' repeated citation of the study of 
households' difficulties in analyzing the financial value of purchasing 
vehicles with higher fuel economy conducted by Turrentine and Kurani, 
which relies on interviews with a limited number of subjects (57 
California households) to conclude that consumers are systematically 
unable to perform the calculations necessary to estimate the value of 
fuel savings.\1613\ These same commenters consistently ignore the 
wealth of detailed, publicly-available information on the fuel economy 
of new vehicle models, and shoppers' ready access to user-friendly 
tools to estimate the savings they are likely to realize from 
purchasing higher-mpg models. These tools include the label that 
prominently displays how much a vehicles' fuel economy will save, or 
conversely, cost a purchaser in fuel costs over 5 years of use in color 
and large type (see Figure VI-63), which is legally required to be 
prominently displayed on all new cars vehicles offered for sale.\1614\ 
Separately, new car dealers are also required to prominently display 
the Federal Fuel Economy Guide for each model year of new vehicles 
offered for sale, which provides fuel economy information for all 
vehicles from that model year.\1615\
---------------------------------------------------------------------------

    \1613\ ICCT at p. 4 and Consumers Union at p. 12 (among others), 
citing Turrentine, T.S., & Kurani, K.S., ``Car buyers and fuel 
economy?,'' Energy policy, Vol. 35 No. 2 (2007), at 1213-1223, 
available at https://www.sciencedirect.com/science/article/pii/S0301421506001200, as evidence that most or all new-car shoppers are 
incapable of calculating the savings they would realize from 
purchasing a higher-mpg model, and further misinterpret the study as 
evidence that buyers invariably underestimate the value of increased 
fuel economy. Yet this widely relied-upon analysis included only 57 
households, all located in California. As an illustration, citing 
Turrentine and Kurani, ICCT asserts ``There is substantial 
circumstantial evidence that most consumers in the U.S. place a low 
value on fuel economy.'' See ICCT at 4 (emphasis added). Similarly, 
Consumers Union simply asserts that ``Households do not track 
gasoline prices over time and cannot accurately estimate future gas 
prices or cost savings.'' See Consumers Union at 12, again citing 
Turrentine and Kurani as authority).
    \1614\ See 15 U.S.C. 1531, et seq., and 49 CFR 575.401.
    \1615\ 40 CFR 600.405-08 and 600.407-08.
    [GRAPHIC] [TIFF OMITTED] TR30AP20.309
    
    Similarly, no commenters offered empirical evidence to support 
their repeated assertions that buyers or the public actually view 
features such as styling, size, or performance as ``positional goods'' 
to which other potential buyers might aspire, or considered the 
possibility that high fuel economy or advanced technology (such as 
hybrid or electric propulsion) might themselves represent such 
positional attributes.\1616\ Nor do commenters

[[Page 24610]]

provide any empirical evidence that the various aspects of behavior 
they allege lead buyers to underinvest in fuel economy--ranging from 
unwillingness to spend time or effort estimating likely fuel savings, 
to inattentiveness to the economic and social importance of improved 
fuel economy, inability to obtain information about the savings it 
offers them, and incorrect ``framing'' of the choice among models with 
different levels of fuel economy--are widespread, empirically 
significant, or systematically likely to lead buyers to under- rather 
than over-invest in fuel economy.
---------------------------------------------------------------------------

    \1616\ For evidence that prestige appears to be a motivation for 
purchasing advanced-technology vehicles, see Hidrue, Michael K., et 
al., ``Willingness to pay for electric vehicles and their 
attributes,'' Resource and Energy Economics, Vol. 33, Issue 3 
(September 2011), at pp. 686-705; Chua, Wan Ying, Lee, Alvin and 
Sadeque, Saalem 2010, ``Why do people buy hybrid cars?,'' 
Proceedings of Social Marketing Forum, University of Western 
Australia, Perth, Western Australia, Edith Cowan University, 
Churchlands, W.A., at pp. 1-13; Liu, Yizao, ``Household demand and 
willingness to pay for hybrid vehicles,'' Energy Economics, Volume 
44, 2014, at pp. 191-197; Hur, Won-Moo, Jeong Woo, and Yeonshim Kim, 
``The Role of Consumer Values and Socio-Demographics in Green 
Product Satisfaction: The Case of Hybrid Cars,'' Psychological 
Reports, Volume 117, issue 2, October 2015, at pp. 406-427. A useful 
summary of many studies appears in Table 1 (p. 196) of Makoto 
Tanaka, Takanori Ida, Kayo Murakami, Lee Friedman, ``Consumers' 
willingness to pay for alternative fuel vehicles: A comparative 
discrete choice analysis between the US and Japan,'' Transportation 
Research Part A: Policy and Practice, Volume 70, 2014, at pp. 194-
209 (Table 1 at p. 196). Some of these studies find that buyers are 
apparently willing to pay significant price premiums for the 
prestige or status value of hybrids or battery-electric vehicles--
which their authors speculate may derive from their ``greenness''--
because their purchases cannot be explained on the basis of economic 
or financial considerations. Others find that average or typical 
shoppers' willingness to pay advanced-technology vehicles is below 
the price premiums they command, suggesting that their purchasers 
must derive some status or prestige value from owning and driving 
them.
---------------------------------------------------------------------------

    The most frequent argument that an energy paradox or energy 
efficiency ``gap'' exists in the case of fuel economy is the 
observation that many U.S. vehicle buyers seem unwilling to pay higher 
prices for models whose increased fuel economy would appear to repay 
their additional investment within a relatively brief ownership period. 
However, this argument is unpersuasive for at least three reasons: Most 
obviously, it does not acknowledge the possibility that engineering 
studies systematically underestimate costs to produce vehicles with 
higher fuel economy, and thus the prices that buyers would be asked to 
pay for models with improved fuel economy. Nor does it account for 
potential sacrifices in other vehicle attributes that manufacturers may 
make in order to achieve higher fuel economy without increasing 
vehicles' purchase prices beyond consumers' willingness to pay. 
Finally, claims that consumers are acting irrationally by refusing to 
purchase higher-mpg models usually reach this conclusion by comparing 
rates at which they implicitly discount future fuel costs--and thus 
evaluate savings from purchasing more fuel-efficient models--to 
interest rates in financial markets that incorporate time horizons or 
risk profiles that may be very different from those of consumers.
    Even putting these concerns aside, comparing future fuel savings to 
the costs of purchasing more expensive models that offer higher fuel 
economy demonstrates only that buyers are not behaving as analysts 
expect them to and believe they should behave. These comparisons do not 
demonstrate that consumers are necessarily acting irrationally, and 
cannot diagnose the nature of information shortcomings buyers face, 
reasons that they might interpret such information incorrectly, or 
identify behavioral inconsistencies they may exhibit. In short, 
conjectures about why buyers might undervalue potential savings from 
investing in higher-efficiency vehicle models do not represent evidence 
that they actually do so, and as discussed above, recent research seems 
to show that such behavior is not widespread, if it exists at all.
    Past joint rulemaking efforts by NHTSA and EPA have repeatedly 
sought to identify a plausible explanation for car buyers' perceived 
undervaluation of improved fuel economy. The agencies have occasionally 
relied on explanations such as consumers' insufficient appreciation of 
the importance of fuel economy, the difficulty of obtaining adequate 
information about the fuel economy of competing models or of converting 
competing models' fuel economy ratings to future fuel costs and 
savings, or consumers' misunderstanding or mistrust of such information 
when it is provided to them. At other times, the agencies have pointed 
to consumers' ``myopia'' about the future--asserting that for some 
reason, they appear to underestimate future fuel costs and savings--or 
argued that shoppers are insufficiently attentive to fuel costs when 
comparing competing models, that the value of improved fuel economy is 
obscured (``shrouded'') by vehicles' other, more visible attributes, or 
that uncertainty about the savings in fuel costs owners will actually 
realize causes them to undervalue those savings when comparing the 
upfront costs of models with different fuel economy.
    Despite the frequency with which the agencies have cited these 
hypotheses, clear support for any of them remains elusive. Consumers 
have long had ready access to detailed information about individual 
models' fuel economy, which appears prominently on the labels displayed 
by new cars,\1617\ and is published online and in printed outlets that 
shoppers use routinely rely widely on to compare models.\1618\ In 
addition, the fuel economy actually experienced by previous buyers of 
individual models is increasingly reported in readily accessible on-
line databases.\1619\
---------------------------------------------------------------------------

    \1617\ Fuel economy labels have been displayed on the window 
sticker of all new light duty cars and trucks since the mid-1970s, 
as required by the Energy Policy and Conservation Act. See https://www.epa.gov/fueleconomy/history-fuel-economy-labeling. Among the 
information currently required to be posted on the fuel economy 
label is both an estimated annual fuel cost for the vehicle, as well 
as an estimate of how that cost compares to the fuel cost over five 
years for an average new vehicle, so it is unclear what information 
consumers lack that prevents them from making an informed decision 
in this regard.
    \1618\ See, e.g., http://www.fueleconomy.gov, where consumers 
can find and compare the fuel economy (and greenhouse gas 
CO2 and smog emissions) of different vehicle models 
across model years, as well as upload information about their own 
real-world fuel economy and compare it to other drivers.
    \1619\ See id.
---------------------------------------------------------------------------

    Similarly, consumers appear to be well aware of the prices they pay 
for gasoline and how those vary among retail outlets, and are reminded 
clearly and frequently of the financial consequences of their fuel 
economy choices each time they purchase fuel. Increasingly, consumers 
also have ready online access to comparisons of fuel prices at 
competing locations near their homes or along routes they travel.\1620\ 
There is also considerable evidence that drivers' forecasts of future 
fuel prices are more accurate than those issued by government agencies 
or private forecasting services.\1621\ Evidence exists

[[Page 24611]]

that car buyers and owners anticipate extreme volatility in fuel 
prices, recognize that there is considerable uncertainty about future 
fuel prices and potential savings from driving a higher-mpg model, and 
respond cautiously to these uncertainties when evaluating competing 
vehicle models,\1622\ none of which suggests a market failure as much 
as it suggests that consumers balance multiple, often competing 
objectives, and make choices based on the outcome of such balancing.
---------------------------------------------------------------------------

    \1620\ See, e.g., Gas Buddy, available at www.gasbuddy.com.
    \1621\ Anderson et al. report evidence that consumers believe 
fuel prices are likely to remain constant in inflation-adjusted 
terms.; see Anderson, Soren T., Ryan Kellogg, and James M. Sallee, 
``What do consumers believe about future gasoline prices?'' Journal 
of Environmental Economics and Management, vol. 66 no. 3 (2013), at 
pp. 383-403. (2013). Other evidence generally supporting this view 
is reported by Allcott, Hunt, ``Consumers' Perceptions and 
Misperceptions of Energy Costs,'' American Economic Review: Papers & 
Proceedings, Vol. 101 No. 3 (2011), at pp. 98-104, (2011), although 
Allcott finds that some fraction of consumers consistently believes 
that gasoline prices will rise in the future. In related research, 
Anderson et al. demonstrate that consumers' expectations that 
gasoline prices will return to their current levels, even after 
sudden and significant variation, is generally accurate; see 
Anderson, Soren T., Ryan Kellogg, James M. Sallee, and Richard T. 
Curtin, ``Forecasting Gasoline Prices Using Consumer Surveys.'' 
American Economic Review: Papers & Proceedings, Vol. 101 No. 3 
(2011), at pp. 110-14. (2011). In contrast to many consumers' 
expectation that fuel prices may vary over the future but will 
generally return to current levels, the U.S. Energy Information 
Administration predicted that gasoline prices would rise 
significantly over the future at the time the two previous rules 
establishing CAF[Eacute]E standards for model years 2012-16 and 
2017-21 were adopted, in 2010 and 2012; see Energy Information 
Administration (EIA), Annual Energy Outlook 2010), Table A12, p. 
131, available at https://www.eia.gov/outlooks/archive/aeo10/pdf/0383(2010).pdf, Table A12, p. 131; and Annual Energy Outlook 2012, 
Appendix A, Table A12, at p. 155, available at https://www.eia.gov/outlooks/archive/aeo12/pdf/appa.pdf, Table A12, p. 155. As of those 
same dates, forecasts of future petroleum prices issued by other 
government agencies and most private forecasting services (with the 
notable exception of HIS-Global Insight, which projected little or 
no increase in future prices) agreed closely with EIA's forecasts 
that prices would increase significantly over both the near- and 
longer-term futures; see EIA, Annual Energy Outlook 2010, Table 10, 
at p. 86; and Annual Energy Outlook 2012, Table 23, available at 
https://www.eia.gov/outlooks/archive/aeo12/table_23.php. Expressed 
in constant-dollar terms, U.S. gasoline prices in 2019 are 
essentially unchanged from those in 2010, although prices have 
varied significantly above and below that level during the 
intervening period. See https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=emm_epm0_pte_nus_dpg&f=m.
    \1622\ For such evidence, see Allcott, Hunt, ``Consumers' 
Perceptions and Misperceptions of Energy Costs,'' American Economic 
Review: Papers & Proceedings, Vol. 101 No. 3 (2011), at pp. 98-104; 
(2011); Greene, David L., (2010). ``How consumers value fuel 
economy: A literature review'' No. EPA-420-R-10-008 (2010) (No. EPA-
420-R-10-008); Brownstone, David, David Bunch, and Kenneth Train, 
``Joint Mixed Logit Models of Stated and Revealed Preferences for 
Alternative-Fuel Vehicles,'' Transportation Research Part B, Vol. 34 
(2000), at pp. 315-338, (2000), among many other sources.
---------------------------------------------------------------------------

    In past rulemakings, the agencies have also hypothesized that 
consumers may ``satisfice''--that is, select some minimum acceptable 
level of fuel economy, and then evaluate models that achieve that 
minimum on the basis of their other attributes. This explanation for 
buyers' reluctance to purchase more fuel-efficient vehicles ignores the 
possibility that they do account fully for the value of higher fuel 
economy in their decision-making, but simply value differences in 
vehicles' other attributes more highly than they do fuel economy, which 
would not reveal irrational or myopic behavior.
    A related argument has been that calculating future savings 
attributable to fuel economy is complicated, so car shoppers resort to 
simplified decision rules to choose among models with different fuel 
economies, and relying on these rules-of-thumb causes them to choose 
models with lower fuel economy.\1623\ However, it is unclear why 
buyers' reliance on simplified procedures or approximations for 
estimating the value of fuel savings would necessarily lead them to 
systematically choose models with lower fuel economies rather than 
leading some to underinvest in fuel economy while others overinvest.
---------------------------------------------------------------------------

    \1623\ See, e.g., 77 FR at 63115 (Oct. 15, 2012).
---------------------------------------------------------------------------

    The agencies have also frequently described consumers as ``loss 
averse,'' making them reluctant to pay the upfront and certain higher 
prices for models offering better fuel economy when the future savings 
they expect to realize are more distant and less certain.\1624\ The 
agencies' past assumption that loss aversion is universal (and equally 
strong) among new-car shoppers appears to be a simplification that is 
largely unsupported by empirical evidence, and in any case has been 
challenged both as a widespread feature of consumer behavior and more 
specifically as an explanation for vehicle shoppers' reluctance to 
purchase more costly models that offer higher fuel economy.\1625\ 
Further, the extremely wide variety of competing models among which car 
buyers can choose enables many of those searching for a model with 
better fuel economy at a comparable price to do so simply by choosing a 
version with fewer other features, which might partly offset the effect 
of their aversion to the prospect of losses from paying a higher 
purchase price. Lastly, the agencies note that both increased fuel 
costs and increased upfront car prices will appear as ``losses,'' so it 
is not obvious why potential buyers would react to the prospects of 
these different forms of losses in different ways.
---------------------------------------------------------------------------

    \1624\ Id. at 63114-15; see also 74 FR at 25511, 25653 (May 7, 
2010).
    \1625\ See supra notes 1611 and 1612.
---------------------------------------------------------------------------

    OMB Circular A-4 does acknowledge that ``[e]ven when adequate 
information is available, people can make mistakes by processing it 
poorly.'' It goes on to say that people may rely on ``mental rules-of-
thumb'' that produce errors, or cognitive ``availability'' may lead to 
consumers overstating the likelihood of an event. However, Circular A-4 
also cautions that ``the mere possibility of poor information 
processing is not enough to justify regulation,'' and that potential 
problems with information processing ``should be carefully 
documented.'' Some of the above examples of potential market failures 
may fall into this category, but lack evidentiary support. As with 
claims of asymmetric information, it is very difficult to distinguish 
between information processing errors and behavior consistent with 
consumer preferences for time and other vehicle attributes that differ 
from what government agency analysts believe they should be.
    Similarly, the agencies have occasionally noted (and seemingly been 
critical of) some consumers' apparent preferences for vehicle 
attributes that convey social status, such as size or styling, and 
suggested that they may give inadequate attention to fuel economy 
because it does not provide similar status. The agencies have also 
suggested that consumers may be reluctant to purchase more fuel-
efficient models because they associate higher fuel economy with 
inexpensive, less well-designed vehicles. These might be plausible 
explanations, were they not contradicted by concurrent arguments that 
potential buyers are inattentive to or uninformed about fuel economy, 
or have difficulty isolating it from vehicles' other attributes. 
Moreover, the market currently offers a wide range of highly fuel 
efficient (and advanced technology) vehicles at many different price 
points, including in the luxury and performance segments, which belies 
the assumption that fuel economy is inconsistent with positional 
attributes. In any case, consumers' hesitance to choose models offering 
higher fuel economy because they are reluctant to sacrifice 
improvements in other vehicle attributes on which they place higher 
values cannot reasonably be characterized as a market failure.
    Although past rulemakings have raised the possibility that car 
buyers' apparent tendency to underinvest in fuel economy could 
plausibly be explained by their use of discount rates exceeding those 
the agencies employ to assess the present value of fuel savings, the 
agencies have generally dismissed that possibility. In combination with 
factors such as their valuation of vehicles' attributes other than fuel 
economy, differences in driving habits that affect fuel economy and in 
how much they expect to drive newly- purchased cars, and variation in 
their expectations about future fuel prices, differing attitudes about 
the importance of future costs relative to more immediate ones could 
readily explain buyers' apparent reluctance to purchase models offering 
fuel economy levels that the agencies interpret as privately 
``optimal.''
    As with consumption of any good or service, the agencies believe 
consumers'

[[Page 24612]]

choice in vehicles represents what economists call ``constrained 
optimization.'' That is, consumers select a bundle of vehicle 
features--within their budget constraint--that optimizes the value to 
them. The agencies also believe, as is the case in every constrained 
consumer choice, that each of these attributes provide what economists 
call diminishing marginal returns (or value) to consumers. For 
instance, the agencies believe that consumers value vehicle size, 
comfort, performance, trim-level, appearance, etc. As such, fuel-saving 
technologies that increase the cost of the car are just one of many 
vehicle attributes that consumers balance against each other. And 
instead of using their entire budget on a single vehicle attribute, 
consumers tend to sacrifice some degree of many or all attributes in a 
degree that varies according to their preferences so that they can 
consume some degree of most or all attributes they value. This means 
that many consumers may not maximize fuel-saving technologies in their 
vehicle selection, but instead may choose some other bundle of 
attributes. The agencies' use of a 30 month pay-back period in this 
analysis--as opposed to fuel-savings over the life of the vehicle--is 
consistent with the constrained optimization consumers perform when 
selecting a vehicle. It is a reasonable representation of consumers' 
valuation of fuel-saving technologies, given the diminishing marginal 
returns of additional fuel economy. If the agencies had used the entire 
undiscounted fuel-savings over the entire life of the vehicle, the 
agencies would be effectively modeling a scenario where consumers 
maximize fuel economy to the detriment of all other vehicle 
attributes--an assumption that is evidently wrong. As such, it is not 
necessary that purchasers do not value lifetime fuel savings--and, in 
all likelihood, purchasers would prefer vehicles with better fuel 
efficiency and all of their preferred attributes--but rather consumers 
are forced to choose between fuel economy and other vehicle attributes 
while weighing how much each attribute contributes to the total cost of 
the vehicle.
    Finally, the agencies have also previously speculated that vehicle 
producers may be reluctant to offer models featuring the higher levels 
of fuel economy that buyers are willing to pay for, and that buyers' 
apparent underinvestment in fuel economy reflects this lack of choice. 
The agencies have speculated that such behavior by manufacturers could 
arise from their collective underestimation of the value that buyers 
attach to fuel economy, or failing this, from limitations on 
competition among them to supply improved fuel economy, whether 
voluntarily or as a consequence of the industry's structure.\1626\ The 
agencies have also raised the seemingly contradictory argument that 
producers have more complete knowledge about fuel economy than 
potential buyers (``asymmetric information'') causing them to provide 
lower levels than buyers demand, and speculated that deliberate 
decisions by manufacturers may limit the range of fuel economy they 
offer in particular market segments.\1627\
---------------------------------------------------------------------------

    \1626\ See 75 FR at 25653-64 (May 7, 2010); and 77 FR at 63115 
(Oct. 15, 2012).
    \1627\ See, e.g. 75 FR 25510-13; 76 FR 57315-19; 77 FR 62914.
---------------------------------------------------------------------------

    The overarching theme of these arguments seems to be that vehicle 
manufacturers cannot identify--or can, but voluntarily forego--
opportunities to increase sales and profits at the expense of their 
rivals by offering models that feature higher fuel economy. The 
agencies have sometimes ascribed this behavior to the risk that 
producers might incur large investments to produce the more fuel-
efficient models that would enable them to seize these opportunities, 
but subsequently lose sales and profits to competitors who simply 
followed suit after their rivals were successful. This explanation is 
at odds with the customary view that innovative producers can be 
rewarded--substantially, even if only temporarily--with commensurate 
profits that justify taking such risks, when they correctly assess 
consumer demand for innovative features or products.
    In any case, behavior on the part of individual businesses that 
leaves obvious opportunities to increase profits unexploited by an 
entire industry seems extremely implausible, particularly in light of 
the fact that auto manufacturers are profit-seeking businesses whose 
ownership shares are publicly traded and subject to regular market 
valuation. This notion also seems to ignore the range of choices 
already available in the current automobile market, where 
extraordinarily efficient models are available in nearly every vehicle 
class or market segment, including plug-in hybrid and fully electric 
versions of a rapidly increasing number of models. Automobile 
manufacturers can, and in fact are, competing on the basis of fuel 
economy.
    The central analysis presented in this final regulatory impact 
analysis does not account for the possibility that imposing stricter 
standards may require manufacturers to make sacrifices in other vehicle 
features that compete with fuel economy, and that some buyers may value 
more highly. If this proved to be the case, more stringent alternatives 
could impose offsetting losses on buyers well beyond the increases in 
vehicle prices that are necessary for manufacturers to recover their 
outlays for adding new technology (or changing design features) to 
improve fuel economy. By doing so, it could significantly reduce the 
estimates of total and net benefits the agencies report. To further 
illustrate this issue, the agencies have conducted a sensitivity 
analysis that incorporates a conservative estimate of consumers' 
valuation of other vehicle attributes, as further discussed in Chapter 
VII of the FRIA accompanying today's notice.\1628\ The agencies also 
recognize that buyers may have time preferences that cause them to 
discount the future at higher rates than the agencies are directed to 
consider in their regulatory evaluations.
---------------------------------------------------------------------------

    \1628\ This sensitivity analysis assumes that consumer's value 
of other vehicle attributes is at least as great as a portion of the 
fuel savings that consumers supposedly ``leave on the table.'' In 
this analysis, the private net benefits of the final rule are a 
positive $15 billion using a 7% discount rate--which is consistent 
with the theory that providing consumers with greater choices will 
enhance their private welfare. The net external benefits are 
identical to the primary analysis, or $34 billion, so the 
sensitivity results show the final rule improves net social benefits 
by $49 billion.
---------------------------------------------------------------------------

    If either case is true--that the analysis is incomplete regarding 
consumer valuation of other vehicle attributes or discount rates used 
in regulatory analysis inaccurately represent consumers' time 
preferences--no market failure would exist to support the hypothesis of 
a fuel efficiency gap. In either case, the agencies' central analysis 
would overstate both the net private and social benefits from adopting 
more stringent fuel economy and CO2 emissions standards. For 
instance, Table VII-93 (Combined LDV Societal Net Benefits for MYs 
1975-2029, CAFE Program, 7% Discount Rate) shows that the CAFE final 
rule would generate $16.1 billion in total social net benefits using a 
7% discount rate, but without the large net private loss of $26.1 
billion, the net social benefits would equal the external net benefits, 
or $42.2 billion. Because government action cannot improve net social 
benefits in the absence of a market failure, if no market failure 
exists to motivate the $26.1 billion in private losses to consumers, 
the net benefits of these final standards would be $42.2 billion.
    In sum, the agencies do not take a position in this rule on whether 
a fuel

[[Page 24613]]

efficiency gap exists or constitutes a failure of private markets. 
Accordingly, the final regulatory impact analysis is not constrained in 
any manner that ensures the private net benefits of more stringent 
standards will necessarily be either positive or negative. In fact, 
however, the analysis supporting this final rule does present a 
situation where adopting more stringent CAFE and CO2 
emission standards aligns consumers' decisions with a simplified 
representation of their own economic interests, and by doing so 
improves their well-being from what they would experience under less 
stringent standards. In other words, our final modelling results 
reflect the case where some fuel efficiency gap persists (albeit of 
smaller magnitude than the agencies found in previous analyses), 
despite our expressed reservations about its likelihood.
(b) Representing Sales Responses in CAFE/CO2 Analysis
    The approach used in the NPRM relied on a single model to produce 
the total number of new vehicle sales in each calendar year for a given 
regulatory scenario. Many commenters expressed reservations about the 
predictive capabilities of the model (CARB, North Carolina Department 
of Environmental Quality, EDF, Aluminum Association). As the Aluminum 
Association commented, ``[D]eveloping a model to predict consumer 
reaction to changes in prices is complicated and highly sensitive to 
macroeconomic conditions, consumer confidence and employment levels.'' 
\1629\ As discussed above, the agencies agree that development of such 
a model is complicated, and the agencies have elected to simplify the 
approach for the final rule. For the purposes of regulatory evaluation, 
the relevant sales metric is the difference between alternatives rather 
than the absolute number of sales in any of the alternatives. As such 
and in response to these comments and others previously addressed, the 
agencies divided the sales response model for the final rule into two 
parts: A nominal forecast that provides the level of sales in the 
baseline (based primarily upon macroeconomic inputs), and a price 
elasticity that creates sales differences relative to that baseline in 
each year. The nominal forecast does not include price, and is merely a 
(continuous) function of several macroeconomic variables that are 
provided to the model as inputs. While the statistical model used in 
the NPRM attempted to account for the influence of these other factors 
in estimating the price elasticity, the forecast in this analysis 
separates the two completely (as described further below). The price 
elasticity is also specified as an input, but this analysis assumes a 
unit elastic response of 1.0--meaning that a one percent increase in 
the average price of a new vehicle produces a one percent decrease in 
total sales.\1630\
---------------------------------------------------------------------------

    \1629\ NHTSA-2018-0067-11952-4.
    \1630\ The ``price increase'' in this case represents the new 
vehicle price net of a portion of fuel savings, described further in 
this section.
---------------------------------------------------------------------------

    The revised sales model features three broad changes: (1) It uses 
the change in average vehicle price net of fuel costs instead of 
vehicle prices on their own, (2) it uses macroeconomic factors to 
project baseline sales without considering vehicle prices, and (3) it 
assesses the change in sales across the various regulatory alternatives 
considered using an own-price elasticity from the literature. These 
changes were made in response to comments that consumers are willing to 
pay for some level of fuel economy and vehicle prices and sales are 
simultaneously and jointly determined (e.g. endogenous). This section 
discusses these three broad changes, as well as other more technical 
and minor changes.
    The first component of the new sales response model is the nominal 
forecast, which is a function (with a small set of inputs) that 
determines the size of the new vehicle market in each calendar year in 
the analysis for the baseline. It leverages some of the same structure 
of the statistical model used in the NPRM, though the dependent 
variable and some of the explanatory variables have changed. It is of 
some relevance that this statistical model is intended only as a means 
to project a baseline sales series. Some commenters raised econometric 
objections about the NPRM specification's ability to isolate the causal 
effect of new vehicle prices on new vehicle sales. The agencies note 
that the nominal forecast model does not include prices and is not 
intended for statistical inference.
    The forecast is derived from a statistical model that accounts for 
a similar set of exogenous factors related to new light-duty vehicle 
sales. In particular, the model accounts for the number of households 
in the U.S., recent number of new vehicles sold, GDP, and consumer 
confidence. The structure of the forecast model is similar to the NPRM 
model, which also used a ARDL specification, but even the variables 
that are common between the two models have different structural forms 
in the final rule version. In particular, the dependent variable has 
been transformed to reflect the fact that, as some commenters 
suggested, households are an important component of demand for new 
vehicles. As such, the dependent variable is defined as new vehicles 
sold per household.\1631\ While this variable still exhibits the cyclic 
behavior that new vehicle sales exhibit over time, the trend shows the 
number of new vehicles sold per household declining since the 1970's, 
as shown in Figure VI-64, where the dotted line is the trend over time. 
As this time series is non-stationary,\1632\ a lagged variable (the 
value in the previous year) is included on the right-hand side of the 
regression equation. In addition, the model includes a lagged variable 
that represents the three-year running sum of new vehicle sales, 
divided by the number of households in the previous year. This variable 
represents the saturation effect, where the existing number of 
households can only buy so many new vehicles before a significant 
number of households already have one (and do not need to buy another). 
As vehicle durability and cost has increased over time, and average 
length of initial ownership has increased similarly, this variable acts 
to put downward pressure on sales after successive years of high sales 
(particularly during extrapolation).
---------------------------------------------------------------------------

    \1631\ Number of U.S. households is taken from Federal Reserve 
Economic data, https://fred.stlouisfed.org/series/TTLHH.
    \1632\ Stationary refers to whether a time series statistical 
properties are constant over time. Since car sales are increasing 
over time, the time series non-stationary.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24614]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.310

BILLING CODE 4910-59-C
    Similar to the NPRM model, the forecast model includes real U.S. 
GDP,\1633\ but in natural logarithm form (as some commenters suggested 
was more appropriate).\1634\ The final variable is consumer sentiment, 
as measured by the University of Michigan survey of consumers.\1635\ As 
both of these series are non-stationary (determined by applying 
augmented Dickey-Fuller unit root tests to the time series), lagged 
versions of the variables are included to ensure stationarity in the 
residuals. The functional form appears below in Equation 2.
---------------------------------------------------------------------------

    \1633\ Federal Reserve Economic Data, available at https://fred.stlouisfed.org/series/GDPC1#0.
    \1634\ EPA-HQ-OAR-2018-0283-6220-1.
    \1635\ http://www.sca.isr.umich.edu/tables.html.
---------------------------------------------------------------------------

    Equation 2--Statistical Model Used to Generate Nominal Forecast
    The model fit is described in Table VI-152. The included lag term 
of the dependent variable and both GDP variables are statistically 
significant at nearly zero, while both the lagged three year sum term 
and consumer sentiment are both marginally significant. Being a time 
series model, the agencies also computed the Durbin-Watson test 
statistic for autocorrelation (1.77) and the Breusch-Godfrey test for 
serial correlation (0.65) at order 1. The signs of the coefficients are 
all correct, in the sense that they are consistent with our 
expectations.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.311


[[Page 24615]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.312

BILLING CODE 4910-59-C
    Because the dependent variable is the number of new vehicles sold 
per household, it is necessary to multiply by the number of households 
to produce an estimate of new vehicle sales. This model is used to 
produce a forecast of new vehicle sales out to 2050, so it is necessary 
to have projections of each variable used in Equation 2 through 
calendar year 2050. In an effort to be consistent with other inputs to 
the analysis, the projection of U.S. GDP is taken from the 2019 AEO. 
The forecast of households in this analysis comes from the Harvard 
Joint Center for Housing Studies 2018 Household projections.\1636\ The 
consumer confidence forecast is taken directly from the University of 
Michigan index for 2017 and 2018, and from the Global Insight forecast 
of consumer confidence for all subsequent years.
---------------------------------------------------------------------------

    \1636\ https://www.jchs.harvard.edu/research-areas/working-papers/updated-household-growth-projections-2018-2028-and-2028-2038.
---------------------------------------------------------------------------

    While the analysis could have relied on a forecast of new vehicle 
sales taken from a published source (the 2019 AEO, for example), using 
a function is an attractive option because it allows the CAFE Model 
dynamically to adjust the forecast in response to input changes. If a 
sensitivity case requires a forecast that is consistent with a set of 
specific, possibly unlikely, assumptions, a forecast of new vehicle 
sales that is consistent with those assumptions may not exist in the 
public domain, for example low GDP growth sensitivity cases. As 
implemented in this rulemaking, using a functional form allows the user 
to vary some of the assumptions to the analysis without creating 
inconsistencies with other elements of the analysis. However, it is 
incumbent upon the analyst to ensure that any set of assumptions that 
deviate from the central analysis are logically consistent.
    This function, and the set of assumptions contained in the central 
analysis, produces a projection that is comparable in magnitude to the 
forecast in the 2019 AEO reference case, though there are differences. 
The two forecasts, and the percentage difference relative to the AEO 
2019, appear in Table VI-153, as does a recent forecast published by 
the Center for Automotive Research.\1637\ The reader will notice that 
even 2017 shows a discrepancy of nearly 7 percent between the final 
rule forecast and the Annual Energy Outlook, one of the larger 
differences between annual forecasts. However, the final rule analysis 
is based upon the certified production volumes of MY2017, which exceed 
17 million units. So, while the difference may seem significant, the 
final rule volumes in 2017 represent the ground truth for model year 
production.\1638\ The CAR forecast, while shorter in length, is 
consistently higher than both the AEO and final rule forecasts--though 
likely also includes class 2b (and possibly class 3) pickup trucks in 
its light vehicle forecast. Finding a public forecast that explicitly 
excludes light-duty vehicles exempt from these regulations is 
challenging. However, all three forecasts exhibit similar trends--
decreases in sales starting in 2019 that last for a few years before 
ticking up again slowly. As commenters observed, all forecasts are 
almost guaranteed to have some errors, and projections out to 2050 
should be taken as potential future projections limited by our 
knowledge at the time, rather than an ironclad prediction of the 
future.
---------------------------------------------------------------------------

    \1637\ https://www.cargroup.org/u-s-light-vehicle-sales-expected-to-take-a-dip-in-2019/, last accessed 11.21.2019.
    \1638\ See CAFE Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24616]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.313

BILLING CODE 4910-59-C
    Although the forecast produces the total number of new vehicle 
sales in the baseline, an elasticity is imposed on price differences to 
produce sales changes between alternatives. The NPRM version of the 
model considered only differences in average new vehicle prices between 
alternatives, and the agencies received a number of comments (from CBD, 
IPI, EDF, CARB, CA et al., and Oakland et al., as well as recent peer 
reviewers) encouraging the agencies to account for some component of 
fuel savings associated with those price changes. In their comment, 
California et al. and Oakland et al. stated the model failed ``to 
consider

[[Page 24617]]

how consumers will respond to the reduced cost of operating the vehicle 
from better gas mileage and therefore inaccurately predicts a decline 
in vehicle sales under the existing standards.'' \1639\ The agencies 
agree that price is not the only consideration, and that the value of 
fuel savings to new vehicle buyers is also relevant to the purchase 
decision.
---------------------------------------------------------------------------

    \1639\ States and Cities, Attachment 1, NHTSA-2018-0067-11735, 
at 86.
---------------------------------------------------------------------------

    In previous rules, while the agencies produced analyses that 
qualitatively considered sales and employment impacts, the agencies 
acknowledged that fuel economy and CO2 standards were likely 
to increase vehicle prices, while simultaneously reducing operating 
costs, and that estimating how consumers would choose to balance those 
two factors in the new vehicle market was challenging.\1640\ 
Furthermore, the agencies recognized that there is a broad consensus in 
the economic literature that the price elasticity of demand for 
automobiles is approximately -1.0.\1641\ The agencies feel that a unit 
elasticity of -1.0 is still a reasonable estimate.\1642\
---------------------------------------------------------------------------

    \1640\ Final Regulatory Impact Analysis, Corporate Average Fuel 
Economy for MY 2017-MY 2025 Passenger Cars and Light Trucks, August 
2012, at 821.
    \1641\ See, e.g., Kleit, A.N., ``The Effect of Annual Changes in 
Automobile Fuel Economy Standards,'' Journal of Regulatory 
Economics, Vol. 2 (1990), at pp 151-72; Bordley, R., ``An 
Overlapping Choice Set Model of Automotive Price Elasticities,'' 
Transportation Research B, Vol. 28B no. 6 (1994), at pp 401-408; and 
McCarthy, P.S. ``Market Price and Income Elasticities of New Vehicle 
Demands,'' The Review of Economics and Statistics, Vol. LXXVII no. 3 
(1996), at pp. 543-547.
    \1642\ For example, a recent review of 12 studies examining 
vehicle price elasticities conducted by the Center of Automotive 
Research (``CAR'') found an ``average short-run elasticity of -
1.09'' and focusing ``only those models which also employ time 
series methods, the average short-run own-price elasticity is higher 
yet, at -1.25.'' CAR's own analysis found a -.79 short-run 
elasticity. Appendix II of the CAR report shows that the long-run 
elasticities ranged from -.46 and -1.2 with an average of -.72. In 
sum, a -1.0 elasticity is well-aligned with the totality of 
research. McAlinden Ph.D., Sean P., Chen, Yen, Schultz, Michael, 
Andrea, David J., The Potential Effects of the 2017-2025 EPA/NHTSA 
GHG/Fuel Economy Mandates of the US Economy, Center for Automotive 
Research, Ann Arbor, MI (Sept. 2016), available at https://www.cargroup.org/wp-content/uploads/2017/02/The-Potential-Effects-of-the-2017_2025-EPANHTSA-GHGFuel-Economy-Mandates-on-the-US-Economy.pdf.
---------------------------------------------------------------------------

    Because the elasticity assumes no perceived change in the quality 
of the product, and the vehicles produced under different regulatory 
scenarios have inherently different operating costs, the price metric 
must account for this difference. As commenters suggested is 
appropriate, the price to which the unit elasticity is applied in this 
analysis represents the residual price change between scenarios after 
accounting for 2.5 years' worth of fuel savings to the new vehicle 
buyer. This approach is consistent with the 2012 FRIA analysis of sales 
impacts, that which considered several payback periods over which the 
value of fuel savings was subtracted from the change in average new 
vehicle price.
    Similar to the NPRM, the price elasticity is applied to the 
percentage change in average price (in each year). However, the average 
price to which the elasticity is applied is calculated differently in 
the final rule in response to comments. As discussed below the price 
change does not represent an increase/decrease over the last observed 
year, but rather the percentage change relative to the baseline. In the 
baseline, the average price is defined as the observed new vehicle 
price in 2017 plus the average regulatory cost associated with the 
alternative. In the case of CO2 standards, the regulatory 
cost is equivalent to the retail equivalent price of technology 
improvements. In the case of CAFE standards, the regulatory cost 
includes both technology costs and civil penalties paid for non-
compliance in a model year. So the change in sales for alternative a in 
year y is:
[GRAPHIC] [TIFF OMITTED] TR30AP20.314

    [Delta]RegCost is the difference in average regulatory cost between 
alternative a and the baseline scenario in year y to make a vehicle 
compliant with the standards, $34,449 is the average transaction price 
of a new vehicle in 2016, NominalSales is the forecasted sales (in the 
baseline) in year y, [Delta]FuelCosts is the change in average fuel 
costs over 2.5 years relative to the baseline in year y and 
PriceElasticity is -1.0:
[GRAPHIC] [TIFF OMITTED] TR30AP20.315


    Where 35,000 miles is assumed to be equivalent to 2.5 years of 
vehicle usage.\1643\ The agencies assume that consumers behave as if 
the fuel price faced at the time of purchase is the fuel price that 
they will face over the first 2.5 years of ownership and usage. 
Essentially, they behave as if fuel prices follow a random walk, 
where the best prediction of (near) future prices is the price 
today. Scrappage rates in the first few years of ownership are close 
to zero, so buyers can reasonably expect to travel the full annual 
mileage in each of the first three years of ownership. Total sales 
in each alternative (that is not the baseline) will equal 
NominalSalesy + [Delta]Salesa,y for 
alternative a in year y.
---------------------------------------------------------------------------

    \1643\ Based on odometer data, 35,000 miles is a good 
representation of typical new vehicle usage in the first 2.5 years 
of ownership and use--though the distribution of usage is large.

    This implementation produces a range of differences in total sales, 
both between alternatives and over time. Table VI-154 shows the range 
of differences in the final rule at the industry level for 
CO2, and Table VI-155 shows the sales changes under CAFE. 
While cost decreases between the baseline and alternatives differ by 
program, one can see that removing the value of fuel savings from the 
price limits the sales increases in the alternatives to under 300,000 
units in a single year under the preferred alternative, and about one 
percent of total sales between 2017 and 2050.
BILLING CODE 4910-59-P

[[Page 24618]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.316

[GRAPHIC] [TIFF OMITTED] TR30AP20.317

BILLING CODE 4910-59-C
    Table VI-154 and Table VI-155 show sales under the baseline 
(augural standards), and differences under the proposal (0 percent 
increase in stringency) and final rule (1.5 percent increase in 
stringency) of MYs 2017-2050.
c) Dynamic Fleet Share (DFS)
    The first module described above (the forecast function and applied 
elasticity)

[[Page 24619]]

determine the total industry sales in each model year from 2018 (in 
this analysis, 2017 is based on certified compliance data) to 2050. A 
second module, the dynamic fleet share, acts to distribute the total 
industry sales across two different body-types: ``cars'' and ``light 
trucks.'' While there are specific definitions of ``passenger cars'' 
and ``light trucks'' that determine a vehicle's regulatory class, the 
distinction used in this phase of the analysis is more simplistic. All 
body-styles that are obviously cars--sedans, coupes, convertibles, 
hatchbacks, and station wagons--are defined as ``cars'' for the purpose 
of determining fleet share. Everything else--SUVs, smaller SUVs 
(crossovers), vans, and pickup trucks--are defined as ``light 
trucks''--even though they may not be treated as such for compliance 
purposes. In the case of SUVs, in particular, many models may have 
sales volumes that reside in both the passenger car and light fleets 
for regulatory purposes, but the dynamic fleet share does not make this 
distinction. The fleet share model was applied at the same level in the 
NPRM--namely, at the level of body-style rather than regulatory class. 
EDF expressed concern that any simulated increase in the light truck 
share represented consumers shifting from sedans to either 4WD drive 
crossovers, SUVs or pickup trucks.\1644\ However, this was not the 
case. All crossovers are considered light trucks for the purposes of 
fleet share, even though they may be 2WD crossovers treated as 
passenger cars for compliance purposes. So, while the number may 
increase overall for a given scenario, the proportion of crossovers 
sold as 4WD, rather than 2WD, does not.
---------------------------------------------------------------------------

    \1644\ EDF, Appendix B, NHTSA-2018-0067-12108, at 40-41.
---------------------------------------------------------------------------

    EDF was also concerned that the sales implementation in the NPRM, 
which relied on the absolute average price to determine differences 
between alternatives, was unduly influenced by fleet share--as 
differences in the share of light-trucks had the potential to skew 
differences in average price because light-trucks are generally more 
expensive than sedans and hatchbacks. The final rule implementation, 
which starts from an observed average transaction price and evolves the 
average price in the alternatives based on average regulatory cost, is 
less vulnerable to this potential distortion. Even if the fleet share 
model (described in greater detail below) increases the share of light 
trucks (for example), the inherent price difference between passenger 
cars and light trucks does not pass through to the average price--only 
the relative difference in compliance costs associated with the vehicle 
types. Despite the fact that light trucks have generally higher 
transaction prices than passenger cars, there is no guarantee that 
regulatory costs will be higher for light-trucks than for cars (which 
depend upon the mix of footprints, their distance from the relevant 
curve, and the technology cost needed to bring each fleet into 
compliance). Thus, the average price differences used in the sales 
calculations are relatively unaffected by the fleet share model.
    As in the NPRM, the dynamic fleet share represents two difference 
equations that independently estimate the share of passenger cars and 
light trucks, respectively, given average new market attributes (fuel 
economy, horsepower, and curb weight) for each group and current fuel 
prices, as well as the prior year's market share and prior year's 
attributes. The two independently estimated shares are then normalized 
to ensure that they sum to one. As with the Sales Response model, the 
DFS utilizes values from one and two years preceding the analysis year 
when estimating the share of the fleet during the model year being 
evaluated. For the horsepower, curb weight, and fuel economy values 
occurring in the model years before the start of analysis, the DFS 
model uses the observed values from prior model years. After the first 
model year is evaluated, the DFS model relies on values calculated 
during analysis by the CAFE model. The DFS model begins by calculating 
the natural log of the new shares during each model year, independently 
for each vehicle class, as specified by the following equation:

[[Page 24620]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.318

HPVC,MY	1: The average horsepower of all vehicle models belonging to 
vehicle class VC, in the year immediately preceding model year MY,
---------------------------------------------------------------------------

    \1645\ As discussed elsewhere in this final rule, model year and 
calendar year are assumed to be equivalent in the simulation--as 
they always have been in all prior rulemaking analyses.
---------------------------------------------------------------------------

HPVC,MY	2: The average horsepower of all vehicle models belonging to 
vehicle class VC, in the year preceding model year MY by two years,
CWVC,MY	1: The average curb weight of all vehicle models belonging 
to vehicle class VC, in the year immediately preceding model year 
MY,
CWVC,MY	2: The average curb weight of all vehicle models belonging 
to vehicle class VC, in the year preceding model year MY by two 
years,
FEVC,MY	1: The average on-road fuel economy rating of all vehicle 
models (excluding credits, adjustments, and petroleum equivalency 
factors) belonging to vehicle class VC, in the year immediately 
preceding model year MY,
FEVC,MY	2: The average on-road fuel economy rating of all vehicle 
models (excluding credits, adjustments, and petroleum equivalency 
factors) belonging to vehicle class VC, in the year preceding model 
year MY by two years,
0.423453: a dummy coefficient, and
1n(ShareVC,MY): The natural log of the calculated share of the total 
industry fleet classified as vehicle class VC, in model year MY.

    In the equation above, the beta coefficients, [beta]C through 
[beta]Dummy, are provided in the following table. The beta coefficients 
differ depending on the vehicle class for which the fleet share is 
being calculated.

[[Page 24621]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.319

    Once the initial car and light truck fleet shares are calculated 
(as a natural log), obtaining the final shares for a specific vehicle 
class is simply a matter of taking the exponent of the initial value, 
and normalizing the result at one (or 100%). This calculation is 
demonstrated by the following:
[GRAPHIC] [TIFF OMITTED] TR30AP20.320

    These shares are applied to the total industry sales derived in the 
first stage of the sales response. This produces total industry volumes 
of car and light truck body styles. Individual model sales are then 
determined from there based on the following sequence: (1) individual 
manufacturer shares of each body style (either car or light truck) 
times the total industry sales of that body style, then (2) each 
vehicle within a manufacturer's volume of that body-style is given the 
same percentage of sales as appear in the 2017 fleet. This implicitly 
assumes that consumer preferences for particular styles of vehicles are 
determined in the aggregate (at the industry level), but that 
manufacturers' sales shares of those body styles are consistent with 
MY2017 sales. Within a given body style, a manufacturer's sales shares 
of individual models are also assumed to be constant over time. The 
agencies assume that manufacturers are currently pricing individual 
vehicle models within market segments in a way that maximizes their 
profit. Without more information about each OEM's true cost of 
production and operation, fixed and variables costs, and both desired 
and achievable profit margins on individual vehicle models, the 
agencies have no reason to assume that strategic shifts within a 
manufacturer's portfolio will occur in response to standards.
    The Global Automakers noted in their comments that the market share 
of SUVs continues to grow, while conventional passenger car body-styles 
continue to lose market share.\1646\ The agencies are aware of this, 
and include the DFS model in an attempt to address these market 
realities. In the 2012 final rule, the agencies projected fleet shares 
based on the continuation of the baseline standards (MY2012-2016) and a 
fuel price forecast that was much higher than the realized prices since 
that time. As a result, that analysis showed passenger car body-styles 
comprising

[[Page 24622]]

about 70 percent of the new vehicle market by 2025. The reality, as 
Global Automakers note, has been quite different.
---------------------------------------------------------------------------

    \1646\ Global Automakers, Attachment A, NHTSA-2018-0067-12032, 
at 13.
---------------------------------------------------------------------------

    The coefficients of the DFS model show passenger car styles gaining 
share with higher fuel prices and losing them when prices are lower. 
Similarly, as fuel economy increases in light truck models, which offer 
consumers other desirable attributes beyond fuel economy (ride height 
or interior volume, for example) their relative share increases. NRDC, 
in particular, found this counterintuitive.\1647\ However, this 
approach does not suggest that consumers dislike fuel economy in 
passenger cars, but merely recognizes the fact that fuel economy has 
diminishing returns. As the fuel economy of light trucks increases, the 
tradeoff between passenger car and light truck purchases increasingly 
involves a consideration of other attributes. Similarly, the 
coefficients show a relatively stronger preference for power 
improvements in cars than light trucks because that is an attribute 
where trucks have outperformed cars, like cars have outperformed trucks 
for fuel economy.
---------------------------------------------------------------------------

    \1647\ NRDC, Attachment 3, NHTSA-2018-0067-11723, at 5.
---------------------------------------------------------------------------

    Rather than estimate new functions to determine relative market 
shares of cars and light trucks, the agencies applied existing 
functions from the transportation module of the National Energy 
Modeling System (NEMS) that was used to produce the 2017 Annual Energy 
Outlook. The functions above appear in the ``tran.f'' input file to 
that version of NEMS, and were embedded (in their entirety) in the CAFE 
model in the NPRM (and this final rule). NEMS uses the functions to 
estimate the percent of total light vehicles less 8,500 GVW that are 
cars/trucks. While NRDC asserted that the agencies must demonstrate the 
propriety of the fleet share model before relying on its 
estimates,\1648\ they ignore the fact that, by using the AEO to develop 
a static fleet in prior rulemakings, the agencies have always relied on 
NEMS estimates. The primary difference between those analyses and the 
NPRM (and this final rule), is that prior analyses applied the fleet 
share that was simulated for the baseline to all regulatory scenarios 
considered. Based on the fleet share functions in NEMS, NPRM corrected 
this internal inconsistency found in previous analyses. This approach 
also enables consistent sensitivity cases--where higher fuel prices 
produce fleets with more transitional passenger car body styles, for 
example--and ensures that the starting point (MY 2017) evolves in 
response to both fuel economy improvements and fuel prices in a way 
that is internally consistent.
---------------------------------------------------------------------------

    \1648\ Id.
---------------------------------------------------------------------------

    The agencies are making one change to the DFS function, which is 
the level of application. While NEMS intended the fleet shares to be 
defined by regulatory classes, vehicles are defined much more coarsely 
in NEMS than in the CAFE model, and manufacturers are not 
differentiated at all. In order to produce well-behaved fleet share 
projections with this model, the agencies applied the share functions 
to body-styles rather than regulatory classes. For many years, there 
was little overlap between nameplates in a manufacturer's passenger car 
regulatory class and its light truck regulatory class. However, with 
the recent emergence of smaller FWD SUVs and crossovers, it is 
increasingly common to have nameplates with model variants in both the 
passenger car and light truck regulatory classes, and it is also common 
for there to be only minor differences (like the presence of 4WD or 
AWD) between versions regulated as cars and versions regulated as light 
trucks. The agencies have modified the application of the fleet share 
equations to focus on body-style, rather than regulatory class, in 
recognition of the increased ambiguity between the regulatory class 
distinction for popular models like the Honda CR-V and Toyota RAV4, 
that sell more than 100K units in each regulatory class (typically 
using the same powertrain configuration). The Nissan Rogue sold more 
than 400K units in MY2017, and almost exactly half of them were in the 
light truck (LT) regulatory class. Applying the fleet share at the 
body-style level preserves the existing regulatory class splits for 
nameplates that straddle the class definitions. It also serves to 
minimize the deviation from the observed MY2017 regulatory class shares 
over time. Had the agencies applied the share equations at the 
regulatory class level, as some commenters incorrectly claimed the 
agencies were doing in the proposal, the passenger car regulatory class 
would have eroded much faster than we've seen in the real world and 
ceased to resemble the composition of the MY2017 fleet. Our 
implementation allows the passenger car (PC) regulatory class to 
continue evolving toward crossover-type cars, if that is what economic 
and policy conditions favor.\1649\
---------------------------------------------------------------------------

    \1649\ The ``passenger car'' fleet for CAFE represents the 
combination of both imported passenger cars (IC) and domestic cars 
(DC). While Table VI-157 illustrates shares for the CAFE program, 
resulting shares under the tailpipe CO2 emissions 
standards are comparable.

---------------------------------------------------------------------------

[[Page 24623]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.321

    Table VI-157 shows the regulatory class shares under the baseline 
(augural standards), proposal (0 percent increase in stringency), and 
final rule (1.5 percent increase in stringency) between 2017 and 2030. 
The shares move relatively little between the classes in the baseline, 
with larger (but still small) deviations occurring in the least 
stringent alternative (0 percent increase) and the final rule. As the 
sensitivity cases show, the changes in shares (both over time and 
between regulatory classes) respond to the fuel price case, but remain 
internally consistent due to the inclusion of the DFS.
    Some commenters encouraged the agencies to consider vehicle 
attributes beyond price and fuel economy when estimating a sales 
response to fuel economy/CO2 standards, and suggested that a 
more detailed representation of the new vehicle market would allow the 
agencies to simulate strategic mix shifting responses from 
manufacturers and diverse attribute preferences among consumers. Doing 
so would have required a discrete choice model (at some level), and 
below the reasons why the agencies have not chosen to employ that 
approach in this final rule.
d) Using Vehicle Choice Models in Rulemaking Analysis
    Some commenters argued that the NPRM's statistical model used to 
estimate changes in sales between alternatives was too highly 
aggregated and missed consumers' valuation of other vehicle attributes. 
CARB, Cities and States, and EDF all made some version of the argument 
that the sales model in the NPRM operated at too high a level of 
aggregation to estimate the real sales response, which primarily occurs 
at the model level where consumers are making decisions based on the 
comprehensive set of attributes and body styles available in the 
market. They also argued that a model must operate at the same level, 
such as a discrete choice model, in order to capture consumer response 
accurately. EPA's Science Advisory Board, Bento, Toyota, Automobile 
Alliance, RFF, and Bunch (writing on behalf of CARB) insisted that the 
best approach to estimating the change in sales across alternatives is 
to use a discrete choice model and embed it in the simulation.
    Other commenters expressed different views on the importance of a 
consumer choice model. For example, while the Aluminum Association 
supported a consumer choice model, they suggested that total new 
vehicle sales may not change due to increases in price, but rather the 
attributes of new vehicles would shift, as consumers would likely shift 
their purchases toward lower content vehicles (in terms of safety, 
luxury, or other option content) when faced with generally higher 
prices. Other commenters, including UCS and CBD, strongly encouraged 
the agencies to avoid using consumer choice models; commenters asserted 
that consumer choice models have historically lacked reliability and 
predictive power.\1650\
---------------------------------------------------------------------------

    \1650\ UCS, Technical Appendix, NHTSA-2018-0067-12039 at 50.
---------------------------------------------------------------------------

    In general, these various comments present the agencies with 
considerably different suggestions on how to address these issues, and 
certain suggestions are in direct opposition to each other. That is, 
while some commenters argue that only micro-level consumer responses 
are relevant to the analysis, and that a consumer choice model is 
required to estimate these responses, others argue that it is 
inappropriate to use a discrete choice model--the method by which those 
responses are econometrically estimated--in a regulatory analysis. 
Adding to the confusion, some of the same commenters who argued against 
a consumer choice model,\1651\ also argued that it was necessary for 
the analysis to account for the influence of other vehicle attributes 
in purchasing decisions, which would require incorporating a discrete 
choice model.
---------------------------------------------------------------------------

    \1651\ For example, see EDF, NRDC, RFF, NCAT, and CBD comments.
---------------------------------------------------------------------------

    CARB argued that ``accurately capturing the relative impact of 
sales shifts versus no-buy decisions would require a more detailed 
consumer choice model, as recommended by the CAFE Model peer reviewers. 
The current new vehicle sales model has no

[[Page 24624]]

way of capturing these types of effects.'' \1652\
---------------------------------------------------------------------------

    \1652\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 192.
---------------------------------------------------------------------------

    David Bunch, writing for CARB, said, ``In fact, in previous 
versions of the CAFE model there were no attempts to directly simulate 
consumer response from within the CAFE model at all. Instead, NHTSA 
relied on fixed projections of future vehicle market behavior from 
multiple sources for the purpose of performing the required economic 
cost and benefit calculations. While this might possibly be less than 
ideal, this approach is only a problem if, in the real world, there 
[are] notable differences in future market behavior [that] occur under 
different regulation scenarios, and, moreover, that these differences 
would be large enough to compromise the validity of the net benefit 
comparisons.'' Bunch essentially argues that the old approach, 
asserting that standards can have no impact on sales, even at the 
individual model level, is more appropriate than trying to capture the 
general idea that when all new vehicles get more expensive, consumers 
are likely to buy fewer of them, all else being equal. The agencies 
disagree with that perspective.
    There are a number of practical challenges to using estimates of 
consumer attribute preferences to simulate market responses. Discrete 
choice models typically rely on fixed effects (or alternative-specific 
constant terms) to account for the unobserved characteristics of a 
given model that influence purchasing decisions, such as styling,\1653\ 
but are not captured by independent variables that represent specific 
vehicle attributes (horsepower, interior volume, or safety rating, for 
example). Ideally, these constant terms would contribute relatively 
little to the fit and performance of the model, assuming that the most 
salient characteristics are accounted for explicitly. In practice, this 
is seldom the case. While the fixed effects at the model level are 
statistically sound estimates of consumer preferences for the 
unobserved vehicle characteristics of the individual models, the 
estimates are inherently historical--based on observed versions of the 
specific vehicle models to which they belong. However, once the 
simulation starts, and new technologies are added to each 
manufacturer's product portfolio over successive generations, it is no 
longer obvious that those constant terms would still be valid in the 
context of those changes.
---------------------------------------------------------------------------

    \1653\ Aesthetics such as styling are difficult, if it not 
impossible, to define in a manner that allows meaningful comparison 
between choices.
---------------------------------------------------------------------------

    Another complication is that discrete choice models are highly 
dependent on their inputs and are unable to account for future market 
changes. For example, the Draft TAR relied on a MY 2014 market (for 
EPA's analysis) and a MY 2015 market (for NHTSA's analysis), while the 
NPRM used a MY 2016 fleet, and this final rule has updated the market 
characterization to a MY 2017 fleet. A discrete choice model estimated 
on any of those model years would probably produce different fixed 
effects estimates for each model variant in the fleet. Even assuming 
that no new variants of a given model are offered over time, new 
nameplates emerge as others are retired--and for those new nameplates 
and all of their model variants, no constant terms would exist. They 
would have to be imputed (either from comparable vehicles in the 
market, some combination of their attributes, or both). Some studies 
have attempted to estimate fixed effects for a single new entrant to 
the market,\1654\ but none have attempted to do so at the scale 
required to migrate a discrete choice model fit on an earlier model 
year to a newer model year for simulation.
---------------------------------------------------------------------------

    \1654\ Berry, Steven, James Levinsohn, and Ariel Pakes (2004). 
Differentiated products demand systems from a combination of micro 
and macro data: The new car market. Journal of Political Economy 
112(1): 68-105.
---------------------------------------------------------------------------

    Figure VI-65 shows the cumulative percentage of nameplates in the 
2017 new vehicle market by year of introduction. About ten percent of 
nameplates in 2017 have been around since the 1970s, but another ten 
percent have only existed since about 2010. This fact illustrates the 
likely necessity of constructing vehicle model fixed effects for the 
inevitable new entrants between the estimating fleet and the rulemaking 
fleet. But it also suggests another challenge. New model entrants are 
driven by the dynamics of the market, where some vehicle models succeed 
and others fail, but a simulated market with a discrete choice model 
can only simulate failure--where consumer demand for specific 
nameplates erode to the point that the nameplate volumes trend toward 
zero. It has no mechanism to generate new nameplates to replace those 
nameplates whose sales it estimates will erode beyond some minimal 
practical level of production.
    Consumer choice models are typically fit on a single year of data 
(a cross-section of vehicles and buyers), but this approach misses 
relevant trends that build over time, such as rising GDP or shifting 
consumer sentiment toward emerging technologies. If such a model is 
used to estimate total sales, but lacks trends in GDP growth or 
employment, etc., it will have the wrong set (likely a smaller set) of 
new vehicle buyers and exaggerate price responses and attribute 
preferences. Consumer preferences change over time in response to any 
number of factors--given manufacturers' recent investments in electric 
powertrains, they are counting on this fact. But a choice model 
estimated on observed consumer preferences for EVs--or other vehicle 
attributes with comparatively little experience in the market--would 
necessarily disadvantage a technology that is currently (or only 
recently) unpopular, but gaining popularity. While these are problems 
that may not matter in the estimation process, where a researcher is 
attempting to measure revealed consumer preference for given attributes 
at a single point in time, they become material once that model is 
integrated into the simulation and dynamically carried forward for 
three decades. The agencies note that models that examine aggregate 
trends, such as the one utilized in this analysis, are able to side-
step this issue by not placing a value on unique vehicle attributes.

[[Page 24625]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.322

    The agencies' compliance simulation model estimates the additional 
cost of technology required to achieve compliance, or to satisfy market 
demand for additional fuel economy. While it necessarily calculates 
these costs on a per-vehicle basis, estimating the cost of additional 
technologies as they are applied to each specific model in order to 
bring an entire fleet into compliance, it is agnostic about how these 
costs are distributed to buyers. Manufacturers have strategic, complex 
pricing models that rely on extensive market research and reflect each 
company's strategic interests in each segment. Automobile companies 
attempt to maximize profit from the sale of their vehicles, rather than 
solely focusing on minimizing the cost of compliance, as this 
rulemaking simulates. Lacking reliable data for each manufacturer on 
production costs and profit margins for each vehicle model in their 
portfolios, the most reasonable course of action is to simulate 
compliance as if OEMs are attempting to minimize costs, and, worth 
noting, this approach is also the one NHTSA takes in its rulemakings 
related to the FMVSS. However, it is obvious that some market segments 
and individual models are much less elastic than others.\1655\ As 
reflected in the prices of those models, consumers are able to bear a 
greater share of the total cost of compliance before negatively 
affecting sales and manufacturer profits.
---------------------------------------------------------------------------

    \1655\ See, for example, Kleit, A.N. (2004), Impacts of Long-
Range Increases in the Fuel Economy (CAFE) Standard. Economic 
Inquiry, 42: 279-294. doi:10.1093/ei/cbh060.
---------------------------------------------------------------------------

    Several commenters (CARB, CBD, IPI, and Bento et al.) suggested 
that the agencies should employ a pricing model that allows 
manufacturers to vary prices in response to heterogeneous consumer 
preferences and different levels of willingness to pay for fuel 
economy, and other attributes, in the new vehicle market. 
Fundamentally, this would require the agencies to model strategic 
pricing for each manufacturer individually--no single pricing model 
would be appropriate for every manufacturer. Bento et al. stated that 
the agencies should simulate the market by allowing manufacturers to 
dynamically adjust vehicle prices to ensure compliance with the 
standards.\1656\ There is no reasonable expectation that the agencies 
could embed and utilize each manufacturer's pricing strategy, as this 
is an essential feature of competitive corporate behavior and that 
automakers closely hold pricing strategy information and the agencies 
have insufficient information to model manufacturer pricing strategies. 
Furthermore, models in the academic literature that commenters have 
suggested are superior because they allow prices to adjust, merely 
demonstrate that the mechanics of those adjustments work; they do not 
imply that the resulting prices are reasonable or realistic. Given the 
burden to estimate each manufacturer's standard under the attribute-
based system, where the mix of vehicles sold defines not only the 
achieved fuel economy of each fleet but also the standard to which it 
is compared, the agencies are understandably reluctant to implement 
models that might shift a manufacturer's mix of vehicles sold within a 
market segment.
---------------------------------------------------------------------------

    \1656\ NHTSA-2018-0067-12326 at 10.
---------------------------------------------------------------------------

    Bunch suggested the agencies use a joint model of household vehicle 
holdings and sales that encompasses decisions to purchase new vehicles, 
retain existing ones, or reduce or augment current holdings of vehicles 
of all types and vintages in each period. Manufacturers would modify 
either new vehicle content, prices, or both to produce a supply of new 
vehicles that allowed them each to comply with standards. And, 
subsequently, households and manufacturers would iteratively interact 
until the market reached equilibrium. The model described by Bunch 
would face many of the same issues outlined above. There are 
significant econometric challenges associated with estimating a 
household's decision to buy a new vehicle instead of a used vehicle (of 
some vintage), or to maintain its current set. And integrating such a 
model would

[[Page 24626]]

require the agencies to simulate the dynamics of the used vehicle 
market--hundreds of unique nameplates for each of dozens of vintages--
in order to provide the correct choice set in each simulated year. Such 
a model is beyond the scope of the current analysis.
    While the agencies believe that these challenges provide a 
reasonable basis for not employing a discrete choice model in today's 
final rule analysis, the agencies also believe they are not 
insurmountable, and that some suitable variant of such models may yet 
be developed for use in future fuel economy and CO2 
emissions rulemakings. The agencies have not abandoned the idea and 
plan to continue experimenting with econometric specifications that 
address heterogeneous consumer preferences in the new vehicle market as 
they further refine the analytical tools used for regulatory analysis.
    Operating at the level of individual auto and light truck model 
variants--the same level at which compliance is, necessarily, 
simulated--may not be tractable for rulemaking analyses. However, 
market shares for brands and manufacturers within market segments are 
more stable over time--even if the volumes of segments across the 
industry fluctuate. In the 2012 final rule, the agencies' analysis 
showed a new vehicle market where the share of passenger car body 
styles--sedans, coupes, hatchbacks--reached almost 70 percent of the 
new vehicle market by 2025, while light trucks, including many 
crossovers, accounted for the remaining 30 percent. Those results were 
consistent with the assumptions made in 2012, but the combination of 
low fuel prices and decreasing differences in fuel consumption between 
body styles has instead reduced the market share of those body styles 
significantly (only 40% in the MY 2017 fleet), and, thus eroded the 
value of the 2012 analysis to inform current decisions. Including a 
choice model that operated on existing market shares, albeit at a 
higher level of aggregation than specific nameplates, such as brand/
segment/powertrain, may be able to improve internal consistency with 
the interaction of assumptions about fuel prices and regulatory 
alternatives. The agencies will continue to engage with the academic 
community and other stakeholders to ensure that future work on this 
question improves our analysis of regulatory alternatives.
3) Scrappage
a) The Impacts of New Vehicle Fuel Economy Standards on Fleet Turnover
    Economic literature and theory indicate that the retirement (or 
scrappage) rates of existing vehicles slows when new vehicle fuel 
economy standards increase and cause new vehicle price increases. 
Slower retirement rates result in an older distribution of the on-road 
fleet. Today's on-road fleet is the oldest it has ever been, 
approaching an average of 12 years old.\1657\ Since older vehicles are, 
on average, less safe and less fuel efficient, modeling this reduction 
in the scrappage rates of existing vehicles has important implications. 
As mentioned in the sales section above, past quantitative analyses of 
CO2 and CAFE standards excluded the scrappage effect (though 
the agencies discussed the scrappage effect qualitatively), which could 
have resulted in an overestimate of the benefits of increasing 
standards.
---------------------------------------------------------------------------

    \1657\ Bureau of Transportation Statistics (BTS). ``Average Age 
of Automobiles and Trucks in Operation in the United States.'' 
Available at https://www.bts.gov/content/average-age-automobiles-and-trucks-operation-united-states.
---------------------------------------------------------------------------

    For the NPRM, the agencies chose for the first time to model the 
change in existing vehicle retirement rates across regulatory 
alternatives. The agencies used a logistic function to estimate the 
instantaneous scrappage rate for vehicles of different body styles and 
model year vintages using registration data from Polk, the estimated 
durability of specific model year vintages, the prices of new vehicles, 
a measure of the cost of travel for the model year cohort versus new 
vehicles in any given calendar year, and other cyclical macroeconomic 
indicators.\1658\
---------------------------------------------------------------------------

    \1658\ For a more detailed explanation of the NPRM model, see 
PRIA Chapter 8.10.
---------------------------------------------------------------------------

    The agencies received many comments about the NPRM's scrappage 
model. While some commenters objected to the inclusion of a scrappage 
model, most commenters supported the inclusion of a dynamic scrappage 
model as an improvement in the agencies' analysis; these comments are 
discussed in Section VI.C.1.b)(3)(a)(ii). Other commenters raised 
concerns about the specific scrappage models used in the NPRM analysis; 
these are discussed in Section VI.C.1.b)(3)(b). Specifically, 
commenters raised concerns about overfitting in the models, the 
identification strategy, the modeling of new and used vehicle fuel 
economy in general, the exclusion of certain variables, about how the 
agencies captured macroeconomic effects, and about the lack of 
integration with the sales model.
    The agencies contemplated all of the comments and suggestions made 
by commenters and, in response, have made several changes to final 
rule's model. First, the agencies changed the time-series strategy used 
in the model, as discussed in Section VI.C.1.b)(3)(c)(iii)(a). This 
change allows the agencies to simplify the models significantly, 
addressing commenters' concerns about potential overfitting of the 
model and difficulty of interpreting individual coefficient values 
(discussed in Section CI.C.1.b)(3)(b)(i)). Second, the agencies changed 
the modeling of the durability effect as discussed in Section 
VI.C.1.b)(3)(c)(iii)(c); this change reduces the reliance on the decay 
function and has the added benefit of addressing concerns about 
overfitting and out-of-sample projections discussed in Section 
VI.C.1.b)(3)(b)(i). Third, a portion of anticipated fuel savings from 
increased fuel economy are netted from new vehicle prices--meaning 
consumers are now assumed to value fuel economy at the time of purchase 
to a certain extent--as discussed in Section VI.C.1.b). This change is 
in response to comments discussed in Section VI.C.1.b)(3)(c)(iii)(d) 
and addresses inconsistent treatment of consumer valuation within the 
NPRM's analysis. Finally, the agencies consider the inclusion of 
additional or alternative variables in the scrappage model in response 
to comments discussed in Section VI.C.1.b)(3)(b)(ii). After extensive 
testing, the agencies concluded that these additional variables do not 
improve the model fits or would introduce autocorrelation in the error 
structures (see Sections VI.C.1.b)(3)(c)(iii)(e) and 
VI.C.1.b)(3)(c)(iii)(f) for further discussion). As such, the agencies 
rejected the additional terms suggested by commenters. Input from 
commenters was used to simplify the scrappage model, make it more 
consistent with modeling of new vehicle prices elsewhere in the 
analysis, and improve its predictions for the instantaneous scrappage 
rates of vehicles beyond age 20.
i) Basis for `The Gruenspecht Effect'
    Gruenspecht (1981) and (1982) recognized that since fuel economy 
standards affect only new vehicles, any increase in price (net of the 
portion of reduced fuel savings valued by consumers) will increase the 
expected life of used vehicles and reduce the number of new vehicles 
entering the fleet. The effects of differentiated regulation in the 
context of fuel

[[Page 24627]]

economy is often deemed the Gruenspecht Effect.\1659\ Jacobsen and van 
Bentham (2015) first quantified the Gruenspecht Effect, or the share of 
new vehicle fuel savings lost to the used vehicle fleet due to delayed 
scrappage, to be between 13 and 16 percent.\1660\
---------------------------------------------------------------------------

    \1659\ Gruenspecht, H. ``Differentiated Regulation: The Case of 
Auto Emissions Standards.'' American Economic Review, Vol. 72(2), 
pp. 328-331 (1982).
    \1660\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and 
Gasoline Policy,'' American Economic Review, Vol. 105, pp. 1312-38 
(2015).
---------------------------------------------------------------------------

    As discussed in the write up of the sales model, fuel economy 
standards increase the cost of acquiring new vehicles, but also improve 
the quality of those vehicles by increasing their fuel economy. The 
CAFE analysis assumes that consumers value 30 months of fuel savings, 
so that the quality-adjusted change in new vehicle prices is the 
increase in regulatory costs less 30 months of fuel savings. As long as 
the quality-adjusted price is positive,\1661\ it becomes more expensive 
for manufacturers to produce vehicles and, as a result, prices of new 
vehicles increase. From a supply and demand perspective, this equates 
to the supply curve for new vehicles moving inwards or to the left and 
a corresponding increase in the equilibrium price and decrease in the 
equilibrium quantity of new vehicles purchased.
---------------------------------------------------------------------------

    \1661\ The quality adjusted price is positive when regulatory 
compliance costs exceed 30 months of fuel savings.
---------------------------------------------------------------------------

    New and used vehicles are substitutes. When the price of a good's 
substitute increases, the demand curve for that good shifts upwards and 
the equilibrium price and quantity supplied also increases. Thus, 
increasing the quality-adjusted price of new vehicles will result in an 
increase in equilibrium price and quantity of used vehicles. Since, by 
definition, used vehicles are not being ``produced'' but rather 
``supplied'' from the existing fleet, the increase in quantity must 
come via a reduction in their scrappage rates. Practically, when new 
vehicles become more expensive, demand for used vehicles increases (and 
they become more expensive). Because used vehicles are more valuable in 
such circumstances, they are scrapped at a lower rate, and just as 
rising new vehicle prices push marginal prospective buyers into the 
used vehicle market, rising used vehicle prices force marginal 
prospective buyers of used vehicles to acquire older vehicles or 
vehicles with fewer desired attributes.
ii) Commenter Response to the Inclusion of the Gruenspecht Effect
(a) Many Commenters Support the Inclusion of the Effect
    Academic researchers and automakers widely agree with the existence 
and direction of the Gruenspecht Effect. For example, RFF commented, 
``There's good evidence supporting the scrappage effect.'' \1662\ The 
Auto Alliance stated that the agencies ``made significant strides 
toward improving their modeling of consumer behavior by adding new 
modules to estimate new vehicle sales and in-use vehicle scrappage in 
response to changes to new vehicle prices.'' \1663\ FCA agreed ``that 
an outcome of the current augural stringency of the CAFE/
[CO2] emission regulations may be a decreasing trend in 
vehicle scrappage rates as consumers delay purchases [. . .] forc[ing] 
consumers to hold their current vehicles for additional time.'' \1664\
---------------------------------------------------------------------------

    \1662\ RFF, Comments EPA NHTSA, NHTSA-2018-0067-11789, at 4.
    \1663\ Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073, 
at 47.
    \1664\ FCA, Comments for CAFE-GHG NPRM Final Public Version, 
NHTSA-2018-0067-11943, at 22.
---------------------------------------------------------------------------

    Other commenters agreed with the existence of the effect, but took 
issue with the implications of the combination of the sales and 
scrappage models. Mark Jacobsen stated ``while we agree that the 
scrappage effects we study will mitigate changes in the used fleet, we 
do not believe they could be strong enough to reverse completely the 
direction of change in the used fleet.'' \1665\ Jacobsen's contention 
was echoed by many commenters; the main point was that they believed 
that the prices of both new and used vehicles should be less expensive 
in the NPRM's preferred alternative than the augural standards, and 
that this should, if anything, result in a larger fleet in the NPRM's 
preferred alternative. This issue is further discussed in Section 
(b)(iv) with other comments about integrating the sales and scrappage 
models and the incremental fleet size across alternatives. Here it is 
important to note that this concern does not suggest that a scrappage 
model should not exist, but takes issue with the specific modeling of 
scrappage and/or sales implemented in the NPRM analysis.
---------------------------------------------------------------------------

    \1665\ Mark Jacobsen and Arthur van Benthem, Letter Describing 
Scrappage Effects, NHTSA-2018-0067-7788, at 2.
---------------------------------------------------------------------------

b) Some Commenters Worry About the Shift in Agency Perspective
    Some commenters argued that the agencies modeling of sales and 
scrappage in the NPRM analysis contradicted previous positions that 
these effects were too uncertain to model. For example, the Center for 
Biological Diversity (CBD) commented:

    In the 2012 rulemaking for fuel economy and [CO2] 
standards, both NHTSA and EPA stated that analysis of the standards' 
impact on new vehicles sales and on the ``scrappage'' of used 
vehicles was too uncertain to be used in the rulemaking. The 
agencies reiterated this position in their 2016 technical assessment 
of the standards.\1666\
---------------------------------------------------------------------------

    \1666\ CBD, Appendix A, NHTSA-2018-0067-12000, at 171.

---------------------------------------------------------------------------
    They further stated:

    The agencies have not provided a meaningful rationale or 
justification for the change in position regarding their ability to 
present quantified estimates of the impact of the standards on new 
vehicle sales and the scrappage of used vehicles.\1667\
---------------------------------------------------------------------------

    \1667\ CBD, Appendix A, NHTSA-2018-0067-12000, at 178.

    To respond to these comments, it is useful to look at the reasons 
the agencies gave for not considering fleet turnover effects on pages 
---------------------------------------------------------------------------
845-46 of the 2012 rulemaking:

    If the value of fuel savings resulting from improved fuel 
efficiency to the typical potential buyer of a new vehicle outweighs 
the average increase in new models' prices, sales of new vehicles 
will rise, while scrappage rates of used vehicles will increase 
slightly. This will cause the ``turnover'' of the vehicle fleet--
that is, the retirement of used vehicles and their replacement by 
new models--to accelerate slightly, thus accentuating the 
anticipated effect of the rule on fleet-wide fuel consumption and 
CO2 emissions. However, if potential buyers value future 
fuel savings resulting from the increased fuel efficiency of new 
models at less than the increase in their average selling price, 
sales of new vehicles will decline, as will the rate at which used 
vehicles are retired from service. This effect will slow the 
replacement of used vehicles by new models, and thus partly offset 
the anticipated effects of the final rules on fuel use and 
emissions.
    Because the agencies are uncertain about how the value of 
projected fuel savings from the final rules to potential buyers will 
compare to their estimates of increases in new vehicle prices, we 
have not attempted to estimate explicitly the effects of the rule on 
scrappage of older vehicles and the turnover of the vehicle 
fleet.\1668\
---------------------------------------------------------------------------

    \1668\ 77 FR 62,623, 63,112-13 (emphasis added).

The agencies' reason for not modeling the fleet turnover effects in 
prior rulemakings was not uncertainty about the direction or impact of 
vehicle prices on sales or scrappage rates, but rather uncertainty 
about how consumers value fuel savings. The agencies now have 
sufficient knowledge regarding the amount of fuel savings consumers are 
assumed to value at the time they purchase new vehicles and make these

[[Page 24628]]

assumptions in the technology application simulation. With this 
assumption, it becomes possible to model the fleet turnover effects, 
including the scrappage effect.
c) Some Commenters Think the Effects Are Uncertain
    Other commenters argue that the sales and scrappage effects are too 
uncertain to include in a rulemaking analysis. For example, CBD argued 
that ``the models are attempting to evaluate the small and uncertain 
effects of changes in vehicle standards on certain dynamics--vehicle 
sales, scrappage rates, and vehicle usage--which are largely determined 
by much stronger forces, such as the state of the economy.'' \1669\
---------------------------------------------------------------------------

    \1669\ CBD, Appendix A, NHTSA-2018-0067-12000, at 177.
---------------------------------------------------------------------------

    The agencies agree that there is uncertainty around the magnitude 
of the sales and scrappage response, but do not agree that sign of 
either effect is uncertain. Importantly, excluding modeling of the 
sales and scrappage effects would only make sense if there was a 
legitimate existential concern--the sales and scrappage effects are 
founded in very basic economic theory, as noted above, in Section 
VI.C.1.b)(3)(a)(i). Furthermore, the agencies believe that assessing 
the magnitudes of the sales and scrappage effects is a tractable task 
for researchers and sufficient data exists to quantify these effects. 
Thus, excluding these effects would be a serious omission that limits 
accurate accounting of the costs and benefits of fuel economy 
standards. Other stakeholders commented that the NPRM analysis did not 
thoroughly consider the uncertainty around the magnitudes of the sales 
and scrappage responses. These comments and the agencies response is 
discussed in Section VI.C.1.b)(3)(b)(i), below. The agencies believe it 
is better to consider a range of the scrappage and sales response to 
address concerns about uncertainty, and that excluding them would be 
inappropriate.\1670\ The agencies did just that with the proposal 
through sensitivity analyses--including seeking comment and having the 
scrappage modeling peer reviewed--and continue to do so for the final 
rule.
---------------------------------------------------------------------------

    \1670\ See, e.g. Ctr. for Biological Diversity v. Nat'l Highway 
Traffic Safety Admin., 538 F.3d 1172, 1203 (9th Cir. 2008), (finding 
that NHTSA inappropriately assigned no value to reducing carbon 
emissions when the value for doing so was ``certainly not zero.'').
---------------------------------------------------------------------------

b) Summary of Notice, Request for Comments, and the Agencies' Response
    The comments related to the scrappage model are summarized here 
into five major categories: Overfitting and identification strategies, 
modeling fuel economy and new vehicle prices, consideration of other 
additional variables, integration with sales or VMT, and evaluations of 
associated costs and benefits due to changes in scrappage rates within 
the CAFE model. Specific modeling decisions the agencies have made or 
considered in response to the public comments summarized in this 
section are discussed in Sections VI.C.1.b)(3)(c)(ii)(d) and 
VI.C.1.b)(3)(c)(iii).
i) Overfitting and Identification Strategy
    Several commenters argued that the NPRM scrappage model did not 
have a clear identification strategy, or that the model over-fit the 
data. These commenters suggest that the NPRM model may not capture a 
causal relationship, but picks up other correlation or noise within the 
data. This section outlines the specific claims made by commenters.
a) Overfitting and the Use of Lagged and Interactions Terms
    Several commenters argued that the results presented in the NPRM 
could be driven by the specific structure of the price effect used in 
the scrappage models that were implemented into the CAFE Model. IPI, 
California States et. al., CARB, and other commenters suggested that 
the NPRM model is over-fit. CARB outlined its argument that the 
agencies overfit the data in the following passage:

    [T]he model appears to be significantly overfit and to suffer 
from multicollinearity. An overfit model means that the model is 
able to precisely replicate past trends, but only through the use of 
too many variables. An overfit model fits the data too well, fitting 
the noise or errors in the data in addition to the underlying 
relationships between the variables of interest. Because an overfit 
model also fits the noise and errors of the data, the out-of-sample 
predictions are unreliable. Comments from Jeremy Michalek and Katie 
Whitefoot suggest that choice of specification of the scrappage 
model could result in substantially different predictions, and that 
the agencies should make only those claims that are robust to 
reasonable variations in the model specifications.\1671\
---------------------------------------------------------------------------

    \1671\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 245.

    The agencies agree that it is important that the scrappage model 
results are robust across those specifications that meet a set of 
econometric criteria (these criteria are discussed further in Section 
VI.C.1.b)(3)(c)(iii)). However, the agencies acknowledge that the NPRM 
could have provided further evidence that the specification did not 
drive the results. In the analysis for the final rule the agencies have 
presented more than one specification of the price effect as evidence 
that the specification chosen here does not drive the results of the 
analysis. Further, claims that the specification of the scrappage 
response in the NPRM is inconsistent with economic theory are false.
    Theoretically, changes in average new prices may have longer-term 
trends that can be picked up by including lagged terms, and/or be non-
linear with age, so that vehicles of different ages have different 
elasticities of scrappage (relative to changes in average new vehicle 
prices). Further, sometimes the effect of one independent variable on 
the dependent variable depends on the magnitude of another independent 
variable--this is called an interaction effect. Regression analysis can 
capture these interaction effects by defining a new variable using some 
combination of independent variables.\1672\ It is necessary to retain 
such interaction terms when doing so.\1673\ For example, it is not 
obvious that the elasticities of scrappage rates to changes in new 
vehicle prices should be constant for all vehicle ages, or put another 
way, the older a vehicle is, the higher likelihood the vehicle will be 
scrapped instead of being retained or resold.
---------------------------------------------------------------------------

    \1672\ Davis, J. B., Statistics using SAS enterprise guide. 
Cary, NC: SAS Institute, pp. 411-415 (2012).
    \1673\ As explained in more detail in Section 
I.A.1.a)(1)(a)(ii)(a), below, the agencies perform several 
sensitivity analyses to ensure the model captures the correct impact 
of interactive effects.
---------------------------------------------------------------------------

    Michalek and Whitefoot, Honda, and other commenters, argued that 
the fact that some of the interaction terms were not statistically 
significant was evidence that the response measured is uncertain. CBD 
in particular claimed that the ``scrappage model is poorly constructed, 
and its results are not statistically significant.''
    In response to such comments, it is important to note that when 
interaction terms are included, the significance of the overall effect 
of a variable should be tested by performing a restricted F-test, which 
simultaneously tests that all coefficients of the variable of interest 
are jointly indistinguishable from zero. The insignificance of one term 
of the interaction does not imply that the effect is indistinguishable 
from zero.\1674\
---------------------------------------------------------------------------

    \1674\ Davis, J. B., Statistics using SAS enterprise guide. 
Cary, NC: SAS Institute, pp. 411-415 (2012).
---------------------------------------------------------------------------

    Commenters also noted the lagged terms and age interactions make 
the new vehicle price effect difficult to interpret. IPI argued that 
``[t]he inclusion of interaction variables make it very difficult to 
evaluate the results of the regression for an individual variable

[[Page 24629]]

of interest.'' Michalek and Whitefoot suggested ``using a Monte Carlo 
analysis to understand the distribution of scrappage outcomes implied 
by uncertainty of the value of the coefficients in the model regression 
and reporting 95% confidence intervals.''
    We agree that the inclusion of lags and age interactions of new 
vehicle prices can make interpreting the sign and magnitude of the 
price effect difficult. It also makes it difficult to use the 
confidence intervals on the coefficients as a way to capture 
uncertainty, since the interaction variables are jointly estimated. 
Thus, for the NPRM analysis, the agencies could not independently 
sample each coefficient from the confidence intervals and perform a 
Monte Carlo analysis.
    While the agencies think that the inclusion of lags and interaction 
terms is theoretically plausible, in response to commenter and peer 
reviewer concerns about overfitting and the difficulty of interpreting 
coefficients, the agencies reconsidered the time series approach. The 
agencies found that new vehicle prices are integrated to order one and 
that the dependent variable is stationary (as discussed in Section 
VI.C.1.b)(3)(c)(iii)(a)). It is therefore sufficient to fit the first 
difference of new vehicle prices within the models. Thus, the agencies 
have simplified the central model of the response of scrappage rates to 
changes in new vehicle prices to exclude lags of the effect. The 
agencies further simplified the central scrappage models to exclude 
interaction of new vehicle prices and vehicle age; this allows the 
agencies to take the 95 percent confidence intervals as a low and high 
range for the magnitude of the price effect for the sensitivity 
analysis. The agencies also include a sensitivity analysis which 
includes interaction terms between new vehicle price and vehicle age to 
allow the elasticity of scrappage to changes in new vehicle price to 
vary by vehicle age.
    Commenters also noted that the model did not perform well for 
vehicles beyond age 20. The agencies noted in the PRIA that the Polk 
dataset for older vehicles was limited and likely led to the inability 
to estimate the scrappage rates for older ages.\1675\
---------------------------------------------------------------------------

    \1675\ FR, Vol 83, No. 165, August 24, 2018, p.43097.
---------------------------------------------------------------------------

    The final rule dataset includes almost 30 percent more data for 
vehicles fifteen years or older than the NPRM, which improves estimates 
of the scrappage rate of vehicles aged 20 to 30 (see Table VI-158). The 
agencies are still unable to capture the scrappage trends for vehicles 
over 30, as the dataset is still limited for the oldest ages of 
vehicles, and still rely on the decay function used in the NPRM for 
vehicles over the age of 30. The limited data explains the inability to 
predict the scrappage rates for older vehicles. However, including 
model year fixed effects and including the share of the initial cohort 
remaining does improve predictions of the final share remaining in the 
final rule models. These changes are discussed in Section 
VI.D.1.b)(c)(i)(c).
b) Reduced Form and Endogenous Prices
    California States et. al., CARB, EDF, IPI and academic commenters 
expressed concerns that the NPRM analysis fit a reduced form of the 
scrappage model, rather than a structural model. In other words, 
instead of explicitly modeling new and used vehicle prices in 
equilibrium under different regulatory alternatives and applying a 
measurement of the elasticity of scrappage to the resulting used 
vehicle prices, the agencies modeled the elasticity of scrappage from 
changes to new vehicle prices. For example, California States et. al., 
argued that the model ``does not link the new and used vehicle markets 
as required by economic theory, nor does it attempt to measure used 
vehicle prices, which form the basis of scrappage theory.''
    While the agencies recognize that there are certain advantages to a 
structural model, they disagree that the sales of new and used vehicles 
must be modeled simultaneously. The agencies do link the new and used 
car markets by including new vehicle prices as an independent variable 
in scrappage regression equation. However, it would be inappropriate to 
include used vehicle prices in this equation due to endogeneity 
concerns. A change in used vehicle prices may change scrappage rates, 
but also an exogenous shock to scrappage rates may cause used car 
prices to vary.
    Furthermore, the agencies are unaware of a viable structural model 
for the scrappage effect. The agencies performed an extensive review of 
economic of literature, both before creating the scrappage model for 
the proposal and revising it for the final rule, but were unable to 
find such a model or any insights on how to construct one. The agencies 
note that commenters did not suggest a structural model that the 
agencies should use or give any indication of whether such a model 
exists.
    In order to understand why such a model is difficult to construct, 
it is important to understand what a structural model of the sales and 
scrappage responses would entail. A hypothetical structural model for 
the new vehicle market can be represented by the following simultaneous 
demand and supply equations:

DNew = [beta]0 + [beta]1 * PNew + 
[beta]2 * PUsed + [beta]3 * PTransit + 
[beta]4 * Income + [beta]5 * Households
SNew = [beta]6 + [beta]7 * PNew + 
[beta]8 * Production CostNew

The demand equation for new vehicles in a given year is determined by 
the annual price of owning and operating new vehicles, the annual price 
of owning and operating used vehicles, the annual price of other 
substitutes, average household income, and the number of households. 
The supply equation is made up of the average price of new vehicles and 
the average cost to produce them.
    As noted in the sales model write up, reducing required fuel 
economy stringency reduces the cost of producing new vehicles, and 
shifts the supply curve to the right. This results in an increase in 
the quantity supplied of new vehicles.
    The structural model for the used vehicle market can be represented 
by the following simultaneous demand and supply equations:

DUsed = [gamma]0 + [gamma]1 * PUsed + 
[gamma]2 * PNew + [gamma]3 * PTransit + 
[gamma]4 * Income + [gamma]5 * Households
SUsed = [gamma]6 + [gamma]7 * PUsed + 
[gamma]8 * Maint RepairUsed + [gamma]9 * Scrap 
ValueUsed

    The aggregate demand equation for used vehicles is determined by 
the price of owning and operating used vehicles, the price of owning 
and operating new vehicles, the price of other transit substitutes, 
average income, and the number of households. The supply curve equation 
for used vehicles is determined by the price of used vehicles, the cost 
to repair and maintain them in service, and the opportunity cost of the 
scrappage value of doing so. Relaxing new vehicle standards reduces new 
vehicle prices and shifts the demand curve for used vehicles downward, 
which reduces demand for used vehicles and the equilibrium price and 
quantity of used vehicles, and increases the annual scrappage rate.
    Modeling the structural equations would require that the agencies 
predict new and used vehicle prices in equilibrium, allowing prices of 
new and used vehicles be determined simultaneously from estimates of 
the supply and demand curves for each market. As CARB stated in the 
following comment, new and used vehicle prices are endogenous--the 
equilibrium prices of each good are simultaneous:


[[Page 24630]]


    Because both scrappage rates and new vehicle prices may 
influence one another, the Agencies would need to utilize different 
statistical techniques to credibly identify the impact of new 
vehicle prices on scrappage rates. For example, the Agencies would 
need to identify an instrumental variable that impacts new vehicle 
price but that does not impact the scrappage rate. Models that 
suffer from endogeneity problems will have biased estimates. In 
other words, the estimates from these models cannot be used to 
inform policy, because they do not actually tell us how new vehicle 
prices impact scrappage.

    CARB suggested a way to correct for endogeneity: Using an 
instrumental variable in a two-stage least squares methodology where 
the instrumental variable is correlated with new vehicle prices, but 
not scrappage rates.\1676\ The agencies could also address the 
potential for endogeneity in two steps: First, they could model the 
impacts of exogenous changes in new vehicle prices on used vehicle 
prices, and second, they could model the impacts of exogenous changes 
in used prices on scrappage rates. Implementing the first step would 
require using an instrumental variable to isolate exogenous shifts to 
the new vehicle supply curve, and then using the predicted values of 
new vehicle prices to model changes in prices for used vehicles of all 
ages. Because prices and scrappage rates are jointly determined in the 
market for used vehicles, predicting the elasticity of scrappage with 
respect to price variation also requires isolating exogenous changes in 
used vehicle price via the use of an instrumental variable.
---------------------------------------------------------------------------

    \1676\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
---------------------------------------------------------------------------

    There is one literature example that approaches the structural 
model that some commenters would like the agencies to implement. 
Jacobsen and van Bentham \1677\ developed a structural model that 
simultaneously solves for prices that clear new and used vehicle 
supplies, and then applies an elasticity of scrappage measure that 
corrects for potential endogeneity of used vehicle values and scrappage 
rates using an instrumental variable methodology. Specifically, they 
use changes in fuel prices as an instrumental variable; changes in fuel 
prices shift the demand for different vehicle models, but not the cost 
of supplying them. This should capture exogenous changes in value, so 
that an exogenous measure of the scrappage elasticity can be isolated 
in the second stage of the two-staged least squares method.
---------------------------------------------------------------------------

    \1677\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and 
Gasoline Policy,'' American Economic Review, Vol. 105, pp. pp. 1312-
38 (2015).
---------------------------------------------------------------------------

    While Jacobsen and van Bentham are able to correct for potential 
endogeneity between used vehicle values and their scrappage rates, 
their structural model to set new and used vehicle values 
simultaneously makes some presumptions that the agencies are not 
comfortable making. First, they calibrate their constant elasticity of 
substitution (CES) utility function using 1999 data from GM's internal 
model. This type of model would estimate elasticities of specific 
vehicle models and require a pricing strategy other than allotting all 
additional technology costs to the vehicle models to which they are 
applied. The agencies have avoided a pricing strategy for the reasons 
cited in the sales model write up. Second, by relying on GM's internal 
model, Jacobsen and van Bentham used elasticities calculated using only 
1999 data of the GM fleet. The agencies do not expect that elasticities 
estimated from 20-year old data from a single OEM's portfolio of 
vehicles would translate to the entirety of the current vehicle 
fleet.\1678\ Finally, Jacobsen and van Bentham represent total vehicle 
demand of a representative consumer from a composite vehicle. This 
approach precludes the realistic consideration that a household may 
prefer two used vehicles over one new vehicle, which is accounted for 
in the agencies' functional equations.
---------------------------------------------------------------------------

    \1678\ Kleit, Andrew N., 2004. ``Impacts of Long-Range Increases 
in the Corporate Average Fuel Economy (CAFE) Standard.'' Economic 
Inquiry 42:279-94.
---------------------------------------------------------------------------

    Jacobsen's and A. van Benthem's model is not a household level 
choice model, and is not meant to determine fleet size, as noted in 
their comment:

    In summary, while the Jacobsen and van Benthem (2015) paper 
cannot inform by how much the total vehicle fleet would expand under 
a CAFE rollback (since we do not estimate by how much it shrinks 
under CAFE), all the evidence and economic logic points to a larger 
total vehicle fleet under a rollback, at odds with NHTSA's fleet 
turnover model.\1679\
---------------------------------------------------------------------------

    \1679\ Mark Jacobsen and Arthur van Benthem, Letter Describing 
Scrappage Effects, NHTSA-2018-0067-7788, at 2.

    The agencies agree that the long-term fleet should be smaller in 
the augural case, as fewer new vehicles flow into the used car market 
(because of lower sales), but do think it is plausible that in the 
short term the fleet size could increase under augural standards if in 
some cases consumers substitute two used vehicles for one new one or 
choose to retain an additional vehicle on the margin because the higher 
value makes doing so a more reasonable investment (at the annual 
level). This sort of outcome is not possible with the Jacobsen and van 
Bentham 2015 model, because the overall demand for vehicles is set by 
the annual rent prices of a composite vehicle. The updates to the 
scrappage model for the final rule are consistent with this view, but 
do show a smaller fleet size under the augural standards relative to 
the proposal. This is discussed further in Section 
VI.C.1.b)(3)(b)(iv)(b).
    Fitting the reduced form equation requires that endogenous 
variables are excluded from the model to avoid biased coefficients. As 
a result, used vehicle prices were omitted by design, because used 
vehicle prices and scrappage rates are endogenous.\1680\ Some 
commenters argue that new vehicle prices and scrappage rates are also 
endogenous; CARB argued that ``the model tries to rely solely on new 
vehicle prices to predict scrappage rates without realizing or 
controlling for the fact that scrappage rates may also affect new 
vehicle prices.'' \1681\
---------------------------------------------------------------------------

    \1680\ Hill, R. C., Griffiths, W. E., & Lim, G. C. Chapter 11: 
Simultaneous Equation Models. In Principles of Econometrics (3rd 
ed., pp. 303-24). Hoboken, NJ: John Wiley & Sons, Inc. (2008).
    \1681\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
---------------------------------------------------------------------------

    Commenters provided neither evidence nor an explanation as to why 
there may be some degree of ``reverse causality'' or endogeneity 
between new vehicle prices and scrappage rates. Two potential 
econometric explanations for such endogeneity could be that: (1) These 
variables are jointly or simultaneously determined, so each one 
influences the other; or (2) the model omitted a variable that causes 
covariance between new vehicle prices and scrappage rates. The agencies 
believe the first source of potential endogeneity can be dismissed, as 
any causal relationship between scrappage rates and new vehicle prices 
would necessarily flow through the used car market, which are 
substitute products for new vehicles, and specifically through the 
mechanism of used car prices. For example, an exogenous shock to 
scrappage rates might cause the supply curve in the market for the 
lowest-price used vehicles to shift, and the resulting change in their 
price might cause price responses in higher-price segments of the used 
vehicle market, which in turn might eventually filter up to the new 
vehicle market and affect the prices for new vehicles. This chain of 
events suggests omitted variable bias might be a concern, rather than 
simultaneity.
    The agencies believe that supply and demand for used vehicles (or 
some measure of their interaction, such as

[[Page 24631]]

used vehicle prices) are the most likely sources of any potential 
omitted variable bias. If an omitted variable is causing bias in the 
estimates, then the bias is observable. Whether endogeneity--through an 
omitted variable--is causing bias is an empirical question, which can 
be answered by conducting common empirical test--the Durbin-Wu-Hausman 
test. The Durbin-Wu-Hausman test requires identifying a suitable 
instrument(s)--a variable--that is correlated with new vehicle prices 
but not with scrappage rates, so any effect exerted on scrappage rates 
by the instrument will occur through their association with prices for 
new vehicles.\1682\ The agencies tested a few alternative approaches, 
which included using the change in new vehicle prices during the 
preceding time period and the level of prices during the current period 
as instrumental variables for the change in prices during the current 
period, and another test using the current-period growth rate in GDP as 
an instrument for the change in new vehicle prices during the current 
period. Each of these tests fails to reject the null hypothesis that no 
endogeneity is present at the 0.05 level of significance.
---------------------------------------------------------------------------

    \1682\ For a conceptual overview of this test, see https://www.statisticshowto.datasciencecentral.com/hausman-test/. For a more 
detailed description of the logic underlying the test and how to 
interpret its results, see http://personal.rhul.ac.uk/uhte/006/ec2203/Lecture%2015_IVestimation.pdf.
---------------------------------------------------------------------------

    For both theoretical and empirical reasons, the agencies are 
therefore skeptical about both the likelihood that scrappage rates will 
affect prices for new vehicles, and the extent to which they might do 
so. The agencies find the theoretical underpinnings for endogeneity to 
be tenuous, and believe the empirical evidence suggests such 
endogeneity is not an issue for today's analysis.
    The agencies chose not to fit a model predicting used vehicle 
prices directly from new vehicle prices for the proposal because 
currently-available time-series data on the prices of used vehicles of 
a given vintage going back to 1975 is limited. EDF cited the lack of 
available data as the reason not to fit the structural model:

    In the absence of any data or analysis, NHTSA did not describe 
the extent to which changes in new vehicle prices affect used 
vehicle prices of varying age, condition, etc. \1683\
---------------------------------------------------------------------------

    \1683\ EDF, Appendix B, NHTSA-2018-0067-12108, at 56.

The agencies note that acquisition, assembly, and cleaning of a 
nationally representative database for calendar years 1974 to 2017 on 
used vehicle prices by vintage from Kelly Blue Book (or a similar 
source) would take months to years, and would push the final rule 
beyond the necessary April 2020 lead time requirement to set MY 2022 
standards. Kelly Blue Book data is readily searchable for current 
prices, but without a time series of used vehicle prices the data 
cannot be used to answer the causal relationship of changes in used 
vehicle prices over time on vehicle retirement rates. Even assembling a 
nationally representative sample of used vehicle prices by vintage 
would be a major undertaking. This is not to suggest that doing so is 
out of scope for future analyses; the agencies plan to consider further 
the possibility of conducting additional analysis on the relationship 
between new and used vehicle prices.
    The agencies considered use of the Consumer Expenditure Survey 
(CEX), which has reported vehicle transaction data annually since 
1984.\1684\ However, the sample of used vehicle purchase prices aged 
twenty and older is severely limited. For vehicles purchased between 
1996 and 2017, the average number of transaction prices reported for 
vehicles aged 20 is 58, and for vehicles aged 25 is 18. Any computation 
of average used vehicle prices from such a small sample would not be 
reliable, and in fact, would be quite noisy. The agencies do not think 
that estimates of a structural model based on such limited sampling 
would improve the prediction of the scrappage effects over use of the 
reduced form equation.
---------------------------------------------------------------------------

    \1684\ U.S. Bureau of Labor Statistics. (2016). Consumer 
Expenditures and Income: Collections & Data Sources. Retrieved from 
https://www.bls.gov/opub/hom/cex/data.htm.
---------------------------------------------------------------------------

    EDF argued that modeling the impact of changes in new vehicle 
prices directly on used vehicle scrappage may not capture the fact that 
changes in used vehicle prices impact vintages differently. Further, 
they argue that if new and used vehicle prices change by the same 
proportion, the effect will have a very small impact on the prices of 
the oldest used vehicles. They argue that these small changes are not 
enough to change the scrappage decisions:

    Given that vehicles can sell for as little as a couple of 
hundred dollars and new vehicle prices average over $30,000, used 
vehicle prices can be as little as 1% of that of a new vehicle. 
Given that the largest increase in new vehicle prices projected by 
NHTSA in the NPRM is less than $3000, and assuming that its effect 
on used vehicle prices is likely to be roughly proportional to 
current relative prices, this might mean that the value of a very 
old vehicle or one in poor condition might only increase by $30 
(decline by $30 under the proposal). It is difficult to see how such 
a change in value would have a measurable impact on scrappage. Of 
course, the impact of an increase in new vehicle prices on used 
vehicle prices might be more or less than proportional to their 
current relative values. However, NHTSA has done nothing to show 
which might be the case. The probability of any realistic change in 
used vehicle prices to induce the scrappage of used vehicles is 
still a complete mystery.\1685\
---------------------------------------------------------------------------

    \1685\ EDF, Appendix B, NHTSA-2018-0067-12108, at 52.

    However, the age interaction on the new vehicle price effect allows 
that the elasticity of scrappage to changes in new vehicle prices may 
not be constant for all ages. Allowing the scrappage elasticity to new 
vehicle prices to vary by age incorporates the fact that the elasticity 
of scrappage of used vehicles and the cross-price elasticity of used 
vehicle demand to new vehicle prices may not be constant with age. At 
some point, the thirty-dollar increase EDF cited could be the 
difference in keeping a marginally used vehicle on the road; it would 
be a 10 percent increase in the price of a used vehicle, and may cover 
State registration fees on a marginally scrapped vehicle.
(c) Time Series
    The scrappage model utilizes panel data. Panel data observes 
multiple individuals or cohorts over time. The data employed by the 
scrappage model observes the scrappage rates of individual model year 
cohorts between successive calendar years. The model allows for the 
isolation of trends over time and across individuals.\1686\ Since the 
scrappage model uses aggregate model year cohorts to estimate scrappage 
rates by age and time-dependent variables (new vehicle prices, fuel 
prices, GDP growth rate, etc.) panel data is necessary to estimate the 
model. A major challenge to using panel data is that the data structure 
requires consideration of potential violations of econometric 
assumptions necessary for consistent and unbiased estimates of 
coefficients both across the cross-section and along the time 
dimension. The cross-section of the scrappage data introduces potential 
heterogeneity bias--where model year cohorts may have cohort-specific 
scrappage patterns. \1687\ Another way to put this is that each model 
year may have its own inherent durability. The NPRM captured this 
potential bias by including model year as a continuous variable, but 
the model amended for the final rule includes the more traditional

[[Page 24632]]

individual fixed effects. This is discussed in Section 
VI.C.1.b)(3)(c)(iii)(a). The time dimension of a panel introduces a set 
of potential econometric concerns present in time series analysis. The 
agencies considered potential autocorrelation in the error structures 
and included lags of the dependent and specific independent variables 
to correct for it; this is not an uncommon practice in dynamic panel 
models.\1688\ Some commenters argued that time series approaches were 
not appropriate in the scrappage model at all. CARB stated the 
following:
---------------------------------------------------------------------------

    \1686\ Cambridge University Press. (1989). Analysis of Panel 
Data. New York, NY.
    \1687\ Cambridge University Press. (1989). Analysis of Panel 
Data. New York, NY.
    \1688\ Bun, M. J. G., & Sarafidis, V. (2015). Dynamic Panel Data 
Models. In The Oxford Handbook of Panel Data (pp. 76-110). New York, 
NY: Oxford University Press.

    Time-series analysis for modeling scrappage is also 
inappropriate for the same reasons as it was for the new vehicle 
sales model--particularly because time-series analysis does not 
capture structural changes, which the scrappage model seeks to 
illustrate.\1689\
---------------------------------------------------------------------------

    \1689\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 243.

    The agencies disagree with CARB's assessment. The potential 
scrappage effect can only be measured with a time series dimension; the 
agencies are interested in how changes in new vehicle prices over time 
impact the retirement rate of the on-road fleet over time. In order to 
isolate this effect, the agencies need multi-period data on the 
scrappage rates of used vehicles and prices of new vehicles.
    The literature on vehicle scrappage rates utilizes panel data, but 
most research has ignored potential autocorrelation issues caused by 
the structural properties of independent variables that vary along the 
time dimension. With the NPRM analysis, the agencies found evidence of 
auto-correlated errors, which were corrected by including three lagged 
terms of the dependent variable.\1690\ While in a pure time series 
analysis, this can be an appropriate methodology to account for 
autocorrelation in the error structure; estimates of the coefficients 
of the lagged dependent variable are biased downwards when applied in 
fixed or random effects panel models. The reason for this is that the 
constant individual specific terms are correlated with the lagged 
dependent variable (by definition, since the individual specific terms 
are constant for all time periods, including the previous period), 
creating a bias in the estimate of the coefficient on the lagged 
dependent variable, and potentially other measures.\1691\ The eponymous 
bias was first discussed in a paper written by Nickell in 1982.\1692\ 
There is an increasing body of work developing estimators built 
specifically for dynamic panel data (DPD), or panel data where there is 
an autoregressive component to the data-generating process. In other 
words, the previous value of the dependent variable impacts the current 
value.
---------------------------------------------------------------------------

    \1690\ FR, Vol 83, No. 165, August 24, 2018, p.43097.
    \1691\ Allison, P., Don't Put Lagged Dependent Variables in 
Mixed Models, (2015, June 2). Retrieved June 1, 2019, from https://statisticalhorizons.com/lagged-dependent-variables.
    \1692\ Nickell, Stephen. ``Biases in Dynamic Models with Fixed 
Effects.'' Econometrica, vol. 49, no. 6, 1981, pp. 1417-26. JSTOR, 
www.jstor.org/stable/1911408.
---------------------------------------------------------------------------

    Further research into this literature (discussed above), comments 
on the NPRM, and peer review comments prompted the agencies to 
reconsider the approach developed for the NPRM. The NPRM analysis did 
not use fixed effects for specific model years, but instead imposed a 
parametric logarithmic relationship of successive model years. This 
parametric model year term will still result in biased estimates of the 
lagged dependent variable because it also does not vary over time for 
the same model year, and is therefore correlated with the 
autoregressive term. Since the autoregressive term carries through 
effects from the previous period (the new vehicle price effect), this 
will also bias the predicted Gruenspecht effect in the NPRM model. 
Updates to the model used for the final rule correct this issue by more 
deliberately considering the time series properties of both the 
dependent and independent variables.
    In reconsidering the appropriate way to address the time series 
properties of the scrappage model, the agencies first consider the 
stationarity of dependent and independent variables. This was suggested 
in James Sallee's peer review:

    In contrast to the new vehicle sales regression reported in the 
PRIA's section 8.6, the discussion of the scrappage regressions does 
not include any discussion of the time series properties of the 
estimators. It is important to test for non-stationarity, for 
example.\1693\
---------------------------------------------------------------------------

    \1693\ CAFE Model Peer Review (Report No. DOT HS 812 590). 
Washington, DC--National Highway Traffic Safety Administration, B-
64.

Importantly, the agencies find that the instantaneous scrappage rate is 
stationary, so that there is no longer term information in the 
scrappage rates to recover with an autoregressive term. This means that 
a DPD model is not necessary to correct for potential autocorrelation 
in the model. This also implies that the autocorrelation in the errors 
is a result of non-stationarity in some or all of the regressors, and 
not the independent variable. The solution to this problem is to 
identify the order of integration of each regressor and difference 
until each is non-stationary. Table VI-160 in Section 
VI.C.1.b)(3)(c)(iii)(a) shows the order of integration of variables 
considered in the scrappage modelling.
(ii) Modeling Fuel Economy
(a) Counterintuitive Signs
    In the NPRM analysis, the agencies controlled for the changes in 
the relative fuel economy of new and used vehicles by including the 
cost per mile of travel in the current period and the previous period 
for both new vehicles and the model year cohort whose scrappage is 
being predicted. This allowed fuel prices to alter the scrappage rates 
of existing vehicles, meaning model year cohorts with lower-than-
average fuel economies were impacted by increases to fuel prices to a 
greater extent than cohorts with higher-than-average average fuel 
economies. It also allowed increases in the fuel economy of new 
vehicles to impact the scrappage rates of existing vehicles; the idea 
is that when new vehicles have a higher average fuel economy, holding 
price constant, the demand for new vehicles should increase relative to 
used vehicles, and scrappage rates should increase. While this was a 
plausible way of controlling for changes in the relative fuel cost per 
mile of usage of new and used vehicles, the agencies noted in the NPRM 
that some of the signs on new vehicle cost per mile were 
counterintuitive, so that increases in the average new vehicle fuel 
economy of certain body styles actually increased the scrappage rates 
of existing vehicles.
    IPI, CARB, CBD, Natural Resources Defense Council (NRDC), and other 
commenters argued that these results were driven more by modeling 
decisions than by actual relationships within the data. NRDC suggested 
that the conclusions from the NPRM model should be treated with 
suspicion until validated by further research:

    [A]n increase in fuel price for a given level of fuel economy 
results in longer vehicle retention even though operational costs 
per mile increase. While it is not possible to rationalize this 
response without significant additional research, it is indicative 
of the fact that the algorithm response functions may not be 
properly defined.\1694\
---------------------------------------------------------------------------

    \1694\ NRDC, Attachment 3: CAFE Model Activity Review, NHTSA-
2018-0067-11723, at 20.

    The agencies agree that the results were counter-intuitive--having 
identified this issue in the NPRM and

[[Page 24633]]

specifically seeking comment on the matter--and considered multiple 
alternative methods of capturing the fuel economy improvements of new 
vehicles within the scrappage model in response to comments. Among the 
changes considered were alternate forms of modeling the form of new 
---------------------------------------------------------------------------
vehicle fuel economy, as suggested by IPI:

    A paper by Shanjun Li et al., provides a useful example of how 
the agencies could include fuel efficiency in their regression 
without raising the econometric concerns that may be leading to 
their nonsensical results. Li et al. include fuel price and vehicle 
fuel efficiency (gallons per mile) of used vehicles as well as a 
variable that captures the interaction of fuel efficiency of used 
vehicles and fuel price in their regression as explanatory 
variables. Unlike the agencies' model, the regression analysis used 
in the Li et al. paper found results that are consistent with 
economic theory: A decrease in overall demand for vehicles and an 
increase in demand for more fuel-efficient cars.\1695\
---------------------------------------------------------------------------

    \1695\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 72.

    The NPRM included changes in new vehicle cost-per-mile, but did not 
include separate variables for fuel prices or fuel economy. This could 
potentially have conflated changes in the cost-per-mile of new vehicles 
from changes in fuel prices and changes in new vehicle fuel economy. 
The agencies considered including changes in fuel prices and new 
vehicle fuel economy as separate measures, as suggested in IPI's 
comment above, but opted for a different method of addressing the 
concern of how to include changes to new vehicle fuel economy in the 
scrappage model. However, specifications considering this approach are 
shown in Section VI.C.1.b)(3)(c)(iii)(d).
(b) New Vehicle Prices Net of Fuel Savings
    UCS, CBD, NRDF, EDF, and other commenters expressed concern that 
quality adjustments were not included in the price series used to fit 
the NPRM model. In particular, commenters suggested that the valuation 
of fuel savings at the time of purchase should be deducted from the new 
vehicle price increases. For example, CBD argued:

. . . [T]he agencies rely heavily on work by Howard Gruenspecht 
regarding the scrappage effect, and the NPRM acknowledges that 
Gruenspecht considered the effect of an increase in price ``net of 
the portion of reduced fuel savings valued by consumers.'' Yet 
consumer valuation of fuel savings is excluded from the scrappage 
model, as well.\1696\
---------------------------------------------------------------------------

    \1696\ CBD, Appendix A, NHTSA-2018-0067-12000, at 177.

    The scrappage model cannot include both independent variables on 
the fuel economy and cost-per-mile of new vehicles, and adjust the new 
vehicle prices by the value of fuel savings considered at the time of 
purchase, which would account for the improvement of the fuel economy 
of new vehicles twice. Thus, the agencies must choose between these 
methods to capture the value improvement of new vehicles when their 
fuel economy increases. The agencies show both methods in Section 
VI.C.1.b)(3)(c)(iii)(d). However, additional comments give reason to 
prefer a methodology that does not model the fuel economy or cost per 
mile of new model year cohorts directly, but instead adjusts the new 
vehicle price series by the amount of fuel savings valued at the time 
of purchase.
    IPI expressed concern that the cost-per-mile measure was included 
in the scrappage model, but not in the sales model:

    [T]he CPM results in the scrappage model are inconsistent with 
the agencies' sale model. In the sales module, the agencies have 
chosen to ignore consumer demand for fuel economy and significantly 
boosted the price impact of the baseline standards as a result. But 
in the scrappage model, the agencies have incongruously allowed 
consumer valuation of fuel economy to drive a significant portion of 
the estimated fatalities.\1697\
---------------------------------------------------------------------------

    \1697\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 79.

The agencies note that the fuel economy of new vehicles was not 
included in the sales model because the signs were statistically 
insignificant when it was included, and the fit of the overall model 
was not improved. It was not excluded because the agencies do not think 
that new vehicle fuel economy does not affect their sales. One way to 
consider the value of increased fuel economy in both the sales and the 
scrappage model (in the same way) is to adjust the price of new 
vehicles by the amount of fuel savings consumers value at the time of 
purchase in both models. This is also consistent with how the CAFE 
model applies technology in the absence of CAFE standards, or when a 
manufacturer is already in compliance with existing standards. In 
response to comments about the counterintuitive signs of the change in 
new vehicle cost per mile for some body styles, and about the 
disconnect in how the fuel economy of new vehicles is modelled in the 
sales and scrappage models, the agencies have adjusted the new vehicle 
price series in both models by the amount of fuel savings consumers are 
assumed to value at the time of purchase (30 months of fuel savings). 
As noted in Section VI.C.1.b)(3)(b)(ii)(a), alternatives to this 
solution are presented in Section VI.C.1.b)(3)(c)(iii)(d). The agencies 
also discuss consideration of other quality improvements over 
successive model years in Section VI.C.1.b)(3)(b)(iii)(d).
(iii) Consideration of Other Additional Variables
    Some commenters expressed concern that the scrappage model 
implemented in the NPRM analysis omitted several theoretically 
important variables in predicting the scrappage rates of the existing 
vehicle fleet. To understand these comments more fully it is useful to 
recall that existing vehicle owners can be private households/
individuals, businesses, or dealerships. They supply the used vehicle 
(in the sense of making it available for use) to the market either by 
reselling them, or continuing to own the vehicle for their own use. 
Theoretically an existing owner will supply a used vehicle for 
additional use if the value of the vehicle (net of the opportunity cost 
of its value as scrap metal and used parts) exceeds the cost of 
maintenance, repair, insurance, and registration fees for the vehicle. 
If a seller does not perform necessary repair or maintenance services 
on the vehicle prior to sale, the value of the vehicle should be offset 
by the cost of those services. Accordingly, the scrappage threshold for 
a vehicle should remain the same regardless of whether the seller or 
buyer pays for any necessary maintenance or repair services on the 
vehicle.
    Under this framework, commenters have argued that the agencies 
should include maintenance and repair costs, the value of the used 
vehicle when scrapped, and other costs to purchase the vehicle, all of 
which were excluded in the NPRM version of the scrappage models. IPI 
stated the following:

    The agencies should include the variables that Gruenspecht and 
others have traditionally included in their scrappage analysis, 
including price of vehicles indexed by maintenance and repair costs, 
the price of scrap metal, and interest rates.\1698\
---------------------------------------------------------------------------

    \1698\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 91.

The agencies agree that these variables are relevant to determining the 
scrappage rates of existing vehicles, but have concerns that the level 
of aggregation of available series related to each of these factors may 
obscure the ability of a statistical model to capture their impact on 
vehicle scrappage rates.

[[Page 24634]]

Below, the agencies discuss commenter concerns about the omission of 
maintenance and repair costs, scrap steel prices, and interest rates, 
in turn. This rulemaking then outline the agencies' further 
consideration of each factor in this final rule analysis, and why each 
chose whether to consider each factor in the analysis for the final 
rule. Empirical results of models considering these factors are shown 
in Sections VI.C.1.b)(3)(c)(iii)(e) and VI.C.1.b)(3)(c)(iii)(f); the 
decision to exclude them from the primary analysis is further explained 
in these sections.
(a) Maintenance and Repair Costs
    EDF, IPI, California States et. Al., CARB, CBD, and other 
commenters suggest that the omission of maintenance and repair costs by 
the agencies was not justified, and that the measure should be included 
in future models. CARB claimed that:

parameters for repair costs and used vehicle prices towards the end 
of life should likely be included in a scrappage model. However, 
neither of these variables appear in the Agencies' model.\1699\
---------------------------------------------------------------------------

    \1699\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.

The agencies agree that the theoretically ideal model of scrappage 
would include maintenance and repair costs. For this reason, the 
agencies explored several methods for explicitly incorporating 
maintenance and repair costs. Section VI.C.1.b)(3)(c)(iii)(f) reports 
model results both with and without a maintenance and repair variable. 
Since the variable is integrated of order one, (see Table VI-158), the 
models including it take the first difference; in this form, increases 
in maintenance and repair costs result in an increase in the scrappage 
rate of existing vehicles, as expected. The sign is also statistically 
significant. While the agencies would prefer a maintenance and repair 
price series that varies by calendar year and vintage, such a series is 
not currently available. The agencies hope to continue to improve this 
variable in future work on the scrappage model, but respond to comments 
by including the first difference of the maintenance and repair series 
in some of the models considered for the model used for the final rule.
    Commenters were apparently confused about the agencies' discussion 
of the impact of fuel economy standards on durability. The agencies 
discussed a finding from the Greenspan and Cohen (1996) paper that 
suggested that higher EPA emission standards actually decreased the 
durability of certain model years. The discussion from the PRIA 
follows:

    In addition to allowing new vehicle prices to affect cyclical 
vehicle scrappage [agrave] la the Gruenspecht effect, Greenspan & 
Cohen also note that engineering scrappage seems to increase where 
EPA emission standards also increase; as more costs goes towards 
compliance technologies, it becomes more expensive to maintain and 
repair more complicated parts, and scrappage increases. In this way, 
Greenspan and Cohen identify two ways that fuel economy standards 
could affect vehicle scrappage--(1) through increasing new vehicle 
prices, thereby increasing used vehicle prices, and finally, 
reducing on-road vehicle scrappage, and (2) by shifting resources 
towards fuel-saving technologies--potentially reducing the 
durability of new vehicles by making them more complex.\1700\
---------------------------------------------------------------------------

    \1700\ PRIA at 1000.

EDF and IPI misinterpret the agencies' discussion of findings from 
Greenspan and Cohen's work to imply that the fuel efficiency variable 
is meant to control for changes in maintenance and repair costs. The 
---------------------------------------------------------------------------
following quote from IPI exemplifies their confusion:

    In addition, the agencies have explicitly excluded several 
theoretically important explanatory variables (e.g., the cost of 
maintenance and repair), which are potentially correlated with fuel 
efficiency. [Footnote 405: Id. at 1000 (indirectly making this point 
with respect to fuel efficiency and maintenance and repair costs 
when emphasizing that `Greenspan & Cohen also note that engineering 
scrappage seems to increase where EPA emission standards also 
increase; as more costs goes towards compliance technologies, it 
becomes more expensive to maintain and repair more complicated 
parts, and scrappage increases'). In other words, maintenance and 
repair costs are correlated with respect to fuel efficiency and 
scrappage rates.]\1701\
---------------------------------------------------------------------------

    \1701\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 78.

The agencies did not mean to imply that including some measure of the 
fuel economy of a model year cohort (cost per mile, in the NPRM model) 
would control for variation in maintenance and repair costs over time. 
The discussion of Greenspan and Cohen's results was intended only to 
demonstrate that durability and standards that increase technological 
complexity may be correlated, so that durability increases may not be 
independent of CAFE/CO2 standards.
    Maintenance and repair costs for a given model year cohort likely 
are correlated with the fuel saving technologies applied to that 
cohort, but there is also a dimension of maintenance and repair costs 
that are correlated with other macroeconomic factors (i.e., wages, 
materials, etc.). Controlling for fuel economy would not capture 
calendar-year-specific changes to maintenance and repair costs that are 
caused by factors other than fuel economy. It also does not seem likely 
that variation in maintenance and repair costs from different fuel 
savings technology would be linearly related to fuel consumption, so 
that even model year variation in maintenance and repair costs could 
not be captured by including some measure of fuel economy or fuel 
consumption. As noted above, the agencies agree that maintenance and 
repair prices exist in the theoretically ideal scrappage model, and 
consider the variable in some of the models presented in Section 
VI.C.1.b)(3)(c)(iii)(f).
(b) Scrap Values
    In the NPRM model, the agencies considered inclusion of the BLS 
scrap steel CPI series. The agencies gave the following reasons for 
excluding the measure in the final NPRM models in the PRIA:

    As noted by Parks (1977), the value of a scrapped vehicle can be 
derived either from the value of recoverable scrap metal or from the 
value of sellable used parts. There are several issues with using 
the BLS scrap steel CPI. First, as in Park's work, the coefficient 
on scrap steel is statistically insignificant--model results 
including the CPI of scrap steel are not shown, as there were other 
theoretical problems with the measure. The material composition and 
mass of vehicles has changed over time so that the absolute amount 
of recoverable scrap steel is not constant over the series. The 
average weight of recoverable steel by vintage would have to be 
known, and this measure would still be missing any other recoverable 
metals and other materials. Further, projecting the future value of 
the recoverable scrap metal would involve computing the amount of 
recoverable steel under all scenarios of fuel economy standards, 
where mass and material composition are assumed to vary across all 
alternatives. This value is not calculated explicitly in the current 
model, which is another reason some estimate of the value of 
recoverable metal is not included in the preferred model 
specification.\1702\
---------------------------------------------------------------------------

    \1702\ PRIA at 1012.

The concerns the agencies raised in the NPRM continue to be present for 
the model used for the final rule. The BLS scrap steel CPI will not 
have the same effect on the opportunity cost (the scrap value) of 
keeping an existing vehicle on the road as opposed to scrapping it for 
successive model year cohorts. The average weight of vehicles has 
changed over successive model years, as has the average steel 
composition.
    Even considering the limitation of using the BLS scrap steel price 
series, commenters expressed concern about the exclusion of a variable 
to capture changes in the value of a vehicle as

[[Page 24635]]

scrapped metal and/or used vehicle parts. As noted in Section 
VI.C.1.b)(3)(b)(iii)(a), IPI suggested that ``the price of scrap 
metal'' should be included, while CARB suggested the model include 
``used vehicle prices towards the end of life.'' The agencies made 
several further attempts to capture this component of vehicle 
scrappage, and address commenters' concerns, in the scrappage models 
used in the final rule. The agencies continue to consider models which 
include the BLS iron and scrap steel CPI series; results of these 
considerations are shown in Section VI.C.1.b)(3)(c)(iii)(f).
(c) Interest Rates
    IPI and EDF expressed concerns that changes in the real interest 
rates of vehicle loans had not been included in the final NPRM 
scrappage model. EDF commented the following:

    NHTSA's model also does not include interest rates or the cost 
of financing a vehicle, another variable which NHTSA acknowledges 
affects scrappage. NHTSA itself states that ``[a]s the real interest 
rate increases so does the cost of borrowing and the opportunity 
cost of not investing. For this reason, it is expected that as real 
interest rates increase that vehicle scrappage should decline. 
Consumers delay purchasing new vehicles because the cost of 
financing increases. Conversely, as real interest rates decrease, 
vehicle scrappage should increase . . . . Yet, NHTSA chooses not to 
include interest rates in its model since inclusion of interest 
rates yields results that are opposite to what is expected--``as 
real interest rates increase, so does the scrappage rate'' in 
NHTSA's model. As discussed above, this is yet another indication 
that the model is flawed and cannot be relied upon.\1703\
---------------------------------------------------------------------------

    \1703\ EDF, Appendix A, NHTSA-2018-0067-12108, at 41.

    The agencies considered real interest rates in the NPRM analysis. 
Increasing the cost of purchasing a vehicle should increase the 
incentive for households to hold onto existing vehicles (as opposed to 
purchasing one) and scrappage rates should decline. The agencies 
excluded real interest rates from the final NPRM model for the reasons 
---------------------------------------------------------------------------
stated in the PRIA:

    Table 8-14, Table 8-15, and Table 8-16 include interest rates 
and maintenance and repair CPI for cars, vans/SUVs, and pickups, 
respectively. For cars, as shown in Table 8-8, real interest rate is 
of the opposite sign than expected; as real interest rates increase, 
so does the scrappage rate--this model is also a worse fit by 
measures of AIC and BIC relative to the preferred model.\1704\
---------------------------------------------------------------------------

    \1704\ PRIA at 1028.

    In response to commenters' concerns, the agencies continue to 
consider interest rates in the model used for the final rule, as shown 
in Section VI.C.1.b)(3)(c)(iii)(e). However, interest rates only affect 
scrappage rates where a household might be unable to finance the 
purchase of a new or used vehicle and instead decides to maintain an 
existing vehicle that would have otherwise been scrapped. The most 
likely substitute for a marginal scrapped vehicle would not be a 
vehicle that could be financed. Accordingly, the relationship between 
interest rates and scrappage rates may be weaker than that between new 
vehicle prices and scrappage rates. The most likely substitutes for new 
vehicles are vehicles just off lease, and the resulting increase in 
residual values will affect slightly older vehicles. Eventually, the 
price of the most likely substitutes for marginally scrapped vehicles 
will also increase, so that scrappage rates will also be affected.
(d) Other Vehicle Quality Adjustments
    CARB and other commenters expressed concerns that the NADA series 
used by the agencies in development of the NPRM scrappage model did not 
make quality adjustments. CARB made the following specific comment:

    By only including new vehicle prices and no other controls for 
vehicle quality, the Agencies' scrappage model omits variables that 
are important predictors of scrappage rates and of vehicle prices. 
Prior work that has relied on new vehicle prices to estimate 
scrappage rates have also included some aspects of quality 
improvements, meaning considering that the vehicle is improving in 
some way. For example, Greenspan and Cohen (1996) include both the 
Bureau of Labor Statistics (BLS) new vehicle price index and the BLS 
cost of repair index.\1705\
---------------------------------------------------------------------------

    \1705\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.

    The NADA average new vehicle transaction price does not control for 
other average characteristics that may change over successive model 
years. The agencies considered controlling for average body style and 
model year characteristics in the scrappage model as an alternative to 
including fixed effects in the model. The considered characteristics 
included: Horsepower to weight, zero to sixty acceleration time, and 
average curb weight. However, performing the pFtest implementation of 
an F-test of goodness-of-fit, from the ``plm'' R package, suggested 
that fixed effects are necessary to control for heterogeneity across 
model years.\1706\ For this reason, average characteristics that are 
constant over calendar years for a given model year cohort cannot be 
included in the model. The agencies do present specifications that 
include the ratio of new to used vehicle performance (since this has 
calendar year level variation and can be included with model year fixed 
effects) in Section VI.C.1.b)(3)(c)(iii)(f).
---------------------------------------------------------------------------

    \1706\ Croissant, Y., Millo, G., & Tappe, K. (2019, September 
7). Package `plm.' Retrieved from https://cran.r-project.org/web/packages/plm/plm.pdf.
---------------------------------------------------------------------------

(iv) Integration of Sales and/or VMT, Total Fleet Size, and Total VMT
    Some commenters believe the ideal model of how CAFE/CO2 
standards affect sales, scrappage, and usage would be a joint household 
choice model. RFF makes the following comment:

    The agencies can fix those problems by making two changes. 
First, they can jointly model VMT and vehicle holdings (i.e., 
scrappage and new-vehicle purchases). The literature provides many 
examples of such modeling for guidance (see citations above). 
Jointly modeling these choices will make the analysis internally 
consistent and will account for the fact that households do not make 
scrappage and vehicle use decisions in isolation. If the model 
predicts that weaker standards cause more scrappage, it will 
simultaneously estimate any increase in VMT for the remaining 
vehicles.\1707\
---------------------------------------------------------------------------

    \1707\ RFF, Comments EPA NHTSA, NHTSA-2018-0067-11789, at 14.

    The advantage of such a model is that sales, scrappage, and usage 
would be jointly determined so that the impacts on scrappage is 
conditional on how increased new vehicle prices affect sales and 
vehicle prices, and usage is dependent on both effects. The agencies 
agree that this type of model would better capture the joint nature of 
the choices of which vehicles to buy, which to sell or scrap, and how 
much to use each than modelling each effect separately. However, the 
agencies are not aware of any national dataset that would allow sales, 
scrappage and usage to be jointly predicted, nor are they confident of 
such a model's ability to predict better than carrying current market 
shares forward.
    The papers cited in the RFF comment, Linn and X. Dou, 2018; \1708\ 
Berry, Levinsohn, and Pakes, 1995; \1709\ and Jacobsen and van Bentham, 
2015,\1710\ either use the CEX or the NADA transaction price series 
merged with the Polk registration counts. The CEX is a relatively small 
sample of households (about 160,000), their vehicle holdings,

[[Page 24636]]

vehicle purchases, and usage. However, it does not report retirement 
rates, but only when a vehicle exits a household's fleet (most often it 
is sold or traded in). Thus, at best, the CEX could be used to build a 
household consumer vehicle holdings and usage model, but the vehicles 
that are scrapped would be implied; scrappage would not be modeled 
directly, nor would it be attached to the number of miles on a vehicle. 
The NADA and Polk datasets used by Jacobsen and van Bentham links 
vehicles prices and scrappage rates, but does not track individual 
household decisions. The Jacobsen and van Bentham paper relies instead 
on a model of the new and used vehicle market which takes cross-price 
elasticities as an assumption derived from the outputs of a 1997 GM 
consumer choice model.1728 1711 The agencies will continue 
investigating whether a consumer/household choice model can serve as an 
alternative to aggregate estimates of sales and scrappage, but are 
skeptical about the ability of such models to predict future model 
shares accurately.
---------------------------------------------------------------------------

    \1708\ J. Linn and X. Dou, ``How Do US Passenger Vehicle Fuel 
Economy Standards Affect Purchases of New and Used Vehicles?'' 
(Washington, DC: Resources for the Future, 2018).
    \1709\ Berry, S., J. Levinsohn, and A. Pakes, ``Differentiated 
Product Demand Systems from a Combination of Micro and Macro Data: 
The New Car Market,'' Journal of Political Economy 112(1) (2004): 
68-105.
    \1710\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and 
Gasoline Policy,'' American Economic Review 105 (2015): 1312-38.
    \1711\ Kleit, Andrew N., 2004. ``Impacts of Long-Range Increases 
in the Corporate Average Fuel Economy (CAFE) Standard.'' Economic 
Inquiry 42:279-94.
---------------------------------------------------------------------------

    As was the case with the 2012 final rule and the 2016 TAR, the 
agencies again note there is no credible consumer choice model which 
can be implemented in the CAFE model. Literature comparing the 
performance of consumer choice models to holding manufacturers constant 
suggest that the latter predicts future market shares better than the 
former. NCAT raises this point in their comment below:

    Academic and other researchers have developed a number of 
vehicle demand (consumer choice) models for the new and/or used 
vehicle markets to look at effects on sales and fleet mix. Rarely 
has there been any effort to validate these models, either for 
consistency across models, or for ability to predict out of sample. 
Recent academic research, as well as work by EPA, has found that 
these models commonly perform worse, especially in the short run, 
than simply holding market shares constant.\1712\
---------------------------------------------------------------------------

    \1712\ NCAT, NCAT Comments, NHTSA-2018-0067-11969, at 11.

For these reasons, the agencies have not used a consumer choice model 
to capture the sales and/or scrappage impacts, but have built reduced 
form equations from aggregate data instead.
    NCAT and CBD also refer to EPA attempts to develop a consumer 
choice model in conjunction with Oak Ridge National Labs, and note that 
the agencies did not use this model for the NPRM analysis. This 
specific choice model, as referenced in the excerpted NCAT comment 
above, has not predicted future market shares as well as projecting 
current shares forward. For this reason the model was not deemed fit to 
include in the policy analysis. NHTSA also worked to develop a consumer 
choice model, but when implemented, the model predicted that some OEM's 
would have unrealistic declines in total sales. The limitations of the 
consumer choice models the agencies have considered is overlooked in 
the following comments from CBD:

    The sales model the agencies use is not the consumer-choice 
model that EPA has been developing and refining for almost a decade. 
Rather, both it and the scrappage model appear to have been 
developed by NHTSA in just the last two years. Neither model has 
been peer-reviewed, nor even released publicly until the publication 
of this NPRM.\1713\
---------------------------------------------------------------------------

    \1713\ CBD, Appendix A, NHTSA-2018-0067-12000, at 175.

The agencies did not use the consumer choice models either agency 
developed because the predictions are not reliable--which has 
disappointed not only the commenters mentioned above, but the agencies 
and researchers who have spent significant resources attempting to 
develop models for these purposes. Instead, the agencies have modelled 
the effects from reduced form equations from aggregate data.
(a) Integration With Sales Model
    The NPRM models did not include any direct linkage between the 
sales, scrappage, and usage functions, as noted by the agencies. Here, 
the agencies consider comments from stakeholders about the lack of 
integration of the scrappage model with sales (and the effect on total 
fleet size), and the lack of integration with the vehicle usage 
schedules (and the effects on total VMT).
    NCAT, EDF, CBD, CARB, and other commenters argued that the sales 
and scrappage models should be directly linked, and that their 
independence predicts the higher fleet size and total VMT under the 
augural standards. CBD makes the following statement:

    The agencies now, irrationally, decouple those two effects, such 
that the number of new vehicles sold (or left unsold) has no effect 
on the number of vehicles scrapped. Relying on the deeply flawed 
scrappage model, the agencies have predicted a massive ballooning of 
fleet size under the existing standards that leads, automatically 
under their model, to a massive increase in VMT. \1714\
---------------------------------------------------------------------------

    \1714\ CBD, Appendix A, NHTSA-2018-0067-12000, at 185.

    The agencies note that the structural model presented in Section 
VI.C.1.b)(3)(b)(i)(b) demonstrates that both the equilibrium quantity 
and the price of new vehicles sold are changed when the production cost 
of new vehicles changes under different regulatory alternatives. 
Specifically, under relaxed standards, the equilibrium price is lower 
and equilibrium sales are higher than the counterfactual augural 
standards. Controlling for other variables that might shift the new 
vehicle supply or demand curves, either new vehicle prices or sales 
could enter the used vehicle demand equation (as in the structural 
model, there is a functional relationship between the two, again, 
controlling for factors that shift the supply and demand curves for new 
vehicles). Thus, the agencies could use either new vehicle sales or 
prices to control for changes in the new vehicle equilibrium solution 
in the scrappage equation. It is important to control for factors that 
affect the demand for vehicles overall (business cycle conditions, 
etc.). The agencies present the preferred models using either new 
vehicle prices or new vehicles sales in Section 
VI.C.1.b)(3)(c)(iii)(d). Since there should be a collinearity between 
the two, it would be inappropriate to include both variables 
simultaneously.
(b) Total Fleet Size
    NCAT, EDF, CBD, CARB, UCS, IPI, California et. al., academic 
commenters, and other stakeholders argue that the fleet size should not 
change much with new vehicle prices. Some commenters go further to 
argue that higher vehicle prices under the augural standards should 
result in a smaller fleet size in the augural case relative to the 
proposal. The agencies agree that the long-term impact of higher new 
vehicle prices should be a slight reduction in fleet size, but do not 
agree that the short-term impacts of the standards on fleet size are 
obvious.
    Many examples from the literature make assumptions that ensure that 
the fleet size under different regulatory alternatives remain constant. 
UCS cites this assumption in the original Gruenspecht works (their 
emphasis):

    Though the agencies cite the Gruenspecht effect for its basis 
for the scrappage model, they ignore a central constraint of 
Gruenspecht's work--namely, his assumption that FLEET SIZE AND TOTAL 
VMT ARE INSENSITIVE TO PRICE.\1715\
---------------------------------------------------------------------------

    \1715\ UCS, UCS MY2021-2026 NPRM: Technical Appendix, NHTSA-
2018-0067-12039, at 60.

Other works ensure the same conclusion with different assumptions. 
Within the

[[Page 24637]]

Jacobsen and van Bentham, 2015 and Goulder et. al., 2012 framework, a 
household first chooses the number of vehicles to own based on the 
average price of all vehicles subject to a budget constraint. After 
choosing the number of vehicles to hold, the household chooses the 
specific type and age of vehicles to hold. However, for some households 
the choice of how many and which vehicles to hold is not disjoint, so 
that a household may choose to hold two used vehicles as a second 
choice to one new vehicle. When new vehicle prices increase, under the 
same budget constraint, they may choose to hold two vehicles instead of 
one. If enough households make this choice, the fleet size could 
slightly increase.
    IPI gives a literature example of a model that does not ensure this 
outcome with initial assumptions. This model directly predicted fleet 
size, and not sales and scrappage. The fleet size in the CAFE model is 
the result of the sales and scrappage models, and not the result of a 
single of the models. Small and Van Dender, 2007 finds that higher new 
vehicle prices are associated with lower total vehicle stock, as IPI 
states in the quote below: \1716\
---------------------------------------------------------------------------

    \1716\ Auto Alliance, Attachment 1: NERA Evaluation, NHTSA-2018-
0067-1207, at D-3.

    In their 2007 study estimating the rebound effect caused by 
changes in fuel efficiency, Kenneth Small and Kurt Van Dender 
derived estimates of the relationship between vehicle price and 
fleet size. By simultaneously estimating a system of equations for 
VMT per capita, fleet size, and fuel efficiency for the United 
States from 1966 to 2001, Small and Van Dender also found that an 
increase in new vehicle price has a negative, statistically 
significant effect on total vehicle stock.\1717\
---------------------------------------------------------------------------

    \1717\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 70.

However, it is worth noting that Hymel, Small, and Van Dender in 2010 
published a study finding a statistically insignificant result of the 
opposite sign.\1718\ The general framework of the two papers are very 
similar, so that the updated results show that the fleet size impact is 
ambiguous.
---------------------------------------------------------------------------

    \1718\ Hymel, Kent M. & Small, Kenneth A. & Dender, Kurt Van, 
2010. ``Induced demand and rebound effects in road transport,'' 
Transportation Research Part B: Methodological, Elsevier, vol. 
44(10), pages 1220-1241.
---------------------------------------------------------------------------

    Toyota and the Automobile Alliance mentioned that NERA built sales 
and scrappage models, and requested that the agencies ``review the NERA 
econometric study's methodologies for adoption or to refine their own 
models.'' The agencies considered the NERA scrappage model, but note 
that the model merges the data for all vehicle types, so that the 
scrappage relationship by age for pickups is adjusted by the same 
constant for all ages. However, the agencies note that each body style 
has a unique functional form with age--as evidenced in Section 
VI.C.1.b)(3)(c)(iii)(c))--so that it does not seem appropriate to merge 
them. Further, it does not seem likely that the elasticity of scrappage 
is the same for all vehicle types.
    While the agencies think there are reasons not to adopt the NERA 
scrappage model as is, this suggested general approach does support 
simplifying the model as further suggested in Section 
VI.C.1.b)(3)(b)(i). Also, this research supports the notion that the 
relative fleet size of the proposed and augural standards is not a 
given. NERA's comments about their model provided:

    The separate changes in new vehicle sales and changes in 
scrappage rates would lead to differences in the overall fleet size 
for the CAFE standard alternatives. The net effects of these two 
changes did not have a substantial effect on the overall fleet 
population under any of the three CAFE alternatives (never more than 
0.25% change in fleet size compared to the augural standards).\1719\
---------------------------------------------------------------------------

    \1719\ Auto Alliance, Attachment 1: NERA Evaluation, NHTSA-2018-
0067-1207, at D-3.HONDA.

The NERA model shows the same directional fleet impacts as the NPRM 
sales and scrappage model. This lends some further support to the 
notion that the fleet impacts are not as certain as some commenters 
suggest.
    Another empirical model predicts a larger total fleet size under 
the augural standards than under the proposed standards. Comments by 
David Bunch offer an extended comparison of the sales, fleet size, and 
retirement rate results of the Department of Energy's National Energy 
Modeling System (NEMS) model under the proposed and augural standards. 
NEMS predicts fleet size from input assumptions about the size of the 
on-road fleet, endogenous new vehicle sales estimates, and exogenous 
assumptions about scrappage.\1720\ However, in his comments Bunch said:
---------------------------------------------------------------------------

    \1720\ From page 109 of 2016 NEMS documentation ``exogenously 
estimated vehicle scrappage and fleet transfer rates.'' https://www.eia.gov/outlooks/aeo/nems/documentation/archive/pdf/m070(2016).pdf.

    Scrappage is an implied behavior determined by projecting total 
fleet size and new vehicle sales. Through this mechanism, all else 
equal, an increase in new vehicle sales would yield an increase in 
scrappage.\1721\
---------------------------------------------------------------------------

    \1721\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling, 
at 77.

    NEMS does not project total fleet size endogenously in their model 
as Bunch assumes. Nor is scrappage an implied behavior determined by 
fleet size and new sales projections. Instead, total fleet size is 
implied from an endogenous sales model, and constant age- and body-
style-specific scrappage rates. The difference between the CAFE Model 
and NEMS is that the CAFE model has both endogenous new vehicles sales 
and scrappage rates--scrappage rates are not assumed to be constant for 
all regulatory alternatives. Fleet size is the implied variable in both 
models.
    Bunch finds that the NEMS model also predicts a larger fleet size 
under the augural standards than the proposed standards. Specifically, 
he finds the following:

    The differences are initially about 100K, increasing linearly 
from 2031 from 200K to 1.8M in 2050. Because even the Existing 
standards remain at the same level after 2025, this would seem to 
represent a very different effect from what might be going on in the 
CAFE model results.\1722\
---------------------------------------------------------------------------

    \1722\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling, 
at 69.

    Bunch goes on to discuss the relationship between sales, scrappage 
---------------------------------------------------------------------------
and fleet size in NEMS in the following passage:

    New vehicle sales generally are growing in both scenarios, so 
economic theory suggests that fleet sizes should also be growing 
(they are). Specifically, although the Gruenspecht effect logic 
suggests that increasing new vehicle sales should lead to increased 
used vehicle scrap rates, the total ``value'' of the fleet is 
increasing, so this would suggest an increase in the fleet size. 
Moreover, new vehicle sales are higher under Existing, so the fleet 
size should be also.\1723\
---------------------------------------------------------------------------

    \1723\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling, 
at 71.

    Bunch makes several claims that are not consistent with available 
data and the agencies' understanding of how the NEMS model. First, he 
states that because sales are growing fleet size should also be 
growing. However, change in fleet size is the result of new vehicle 
sales less the number of existing vehicles scrapped; if new vehicle 
sales and used vehicle scrappage rates both increase, the fleet size is 
not necessarily increasing. Second, he states that the `Gruenspecht 
effect logic' suggests that increasing new vehicle sales results in 
increasing scrappage rates. However the NEMS model does not change 
vintage-specific scrappage rates endogenously, but takes them as an 
exogenous input. Thus, the NEMS model does not capture the Gruenspecht 
effect, and its fleet size projections can only vary from changes in 
new vehicle sales. Any differences in the projected total fleet 
scrappage rates Bunch considers later are due to

[[Page 24638]]

different initial sales of each body style, and therefore a different 
weighting of the constant body-style- and vintage-specific scrappage 
rates. This makes the comparison of the fleet size and scrappage rates 
of the two models not particularly meaningful. However, the difference 
in the projected sales impacts are worth a second glance. NEMS predicts 
prices that are at most about $1,000 higher in the Augural than the 
proposed standards, while the CAFE model predicts prices that are up to 
approximately $2,500 higher. The difference in the projected costs to 
meet the CAFE standards is likely the main reason for the difference in 
the sales outcomes--if the average fuel savings exceed the average 
incremental cost of the augural standards (relative to the proposal) in 
the NEMS model, the expected outcome is that sales should be higher in 
the augural case, as shown.
    It is also worth noting Bunch's discussion of the empirical results 
of the CAFE scrappage model. Bunch purports to calculate the scrappage 
elasticity relative to new vehicle price increases, but his point of 
comparison does not hold constant other factors that might impact used 
vehicle scrappage rates. Instead, Bunch calculates the inter-annual 
percentage change in the scrappage rates for each regulatory 
alternative, then calculates the inter-annual change in new vehicle 
prices for each regulatory alternative, and finally takes the quotient. 
However, for inter-annual changes in scrappage rates, different 
projected GDP growth rates and fuel prices will have also played a 
critical role in the scrappage rates. The better point of comparison 
would be the incremental percentage decrease in scrappage rates for the 
augural standard relative to the proposal, over the incremental 
percentage increase in new vehicle price in the augural standard 
relative to the proposal for each calendar year. This ensures that the 
point of comparison holds constant all other factors that determine 
scrappage, as the regulatory alternatives use the same GDP growth rate 
and fuel price projections. When computing the implied scrappage 
elasticity in this way, the implied elasticities vary between 
approximates -0.1 and -1.1, with the average being approximately -0.5--
which is more in line with what Bunch determines reasonable for his 
incorrect calculations of the NEMS model scrappage elasticities, as 
cited below:

    Finally, the average values are -0.90 and -0.88 for the Existing 
and Rollback scenarios, respectively. On one hand, these are 
reasonably close to the Jacobsen and van Benthem (2015) estimate for 
scrap elasticity with respect to used vehicle prices. On the other 
hand, the Bento et al. (2018) estimate was -0.4, and one might 
expect the elasticity with respect to new vehicle price to be 
smaller. In any case, these results are not unreasonable.\1724\
---------------------------------------------------------------------------

    \1724\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling, 
at 79.

The implied elasticities from the NEMS model are approximately zero, 
which is not a surprise since these are merely the result of different 
new vehicle sales affecting the relative weighting of NEMS' constant 
age-specific scrappage rates. Figure VI-66, below, shows a comparison 
of fleet sizes under the baseline, preferred alternative, and AEO 2019. 
The agencies see that, as commenters believed likely, the fleet size 
under the preferred alternative (where sales are larger in many years 
and scrappage rates higher) is eventually larger than in the baseline. 
However, those differences are minimal in the early years of the 
simulation where policy differences produce only small differences in 
sales and scrappage. Furthermore, the agencies see that the magnitudes 
of the fleet sizes in today's rule are generally similar to those 
produced by the AEO 2019 model. NEMS tends to produce growth that is 
more linear, leading to slightly smaller fleet sizes than those 
simulated by the CAFE Model through the 2030's and slightly larger 
fleet sizes through the 2040's. However, these differences are at most 
three percent of fleet size, and typically closer to one or two 
percent.

[[Page 24639]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.323

    As discussed above, commenters offered NERA's model and NEMS as 
points of comparison for NHTSA's sales and scrappage models and their 
combined implied fleet size. However, since NEMS does not model the 
scrappage effect, but takes static scrappage rates, it is not a fair 
point of comparison. NERA's model shows a larger fleet under the 
Augural standards, providing evidence that the impacts of the sales and 
scrappage models are ambiguous.
(c) Integration With VMT
    In the NPRM the agencies noted that the average VMT by age is 
constant regardless of instantaneous or cumulative scrappage rates. The 
agencies noted that this was a limitation of the model, and sought 
comment on ways to integrate the two effects:

    [O]ur scrappage model assumes that the average VMT for a vehicle 
of a particular vintage is fixed--that is, aside from rebound 
effects, vehicles of a particular vintage drive the same amount 
annually, regardless of changes to the average expected lifetimes. 
The agencies seek comment on ways to further integrate the survival 
and mileage accumulation schedules.\1725\
---------------------------------------------------------------------------

    \1725\ EDF, Appendix B, NHTSA-2018-0067-12108, at 51.

Several commenters suggest that the lack of integration between VMT and 
scrappage rates is not justified. Some commenters suggested that the 
VMT should be determined from a household holdings model, while others 
suggested merely that delayed scrappage under higher standards should 
increase average mileage accumulation, which will have some feedback 
for the next year's scrappage rates.
    Joshua Linn and other commenters suggest that VMT is determined at 
the household level and should thus be modelled as such. EDF makes the 
following comment, which seems to reflect a fundamental 
misunderstanding of the type of model used to predict the scrappage 
effect:

    When describing the process whereby a potential new vehicle 
purchaser chooses to forego buying a new vehicle and continues to 
drive their existing vehicle, NHTSA's scrappage model ignores the 
fact that this action shifts VMT from a new vehicle with a higher 
average mileage per year to a used vehicle with a lower average 
mileage. Either the driver of this vehicle will drive their older 
vehicle less, causing overall VMT to decline, or the average mileage 
of the used vehicle will increase without any need to affect 
scrappage. By focusing solely on scrappage, and focusing the change 
in scrappage on those vehicles with the worst fuel economy (i.e., 
the oldest vehicles), NHTSA essentially shifts new vehicle VMT to 
the oldest vehicles. According to NHTSA's own rationale, much of the 
lost VMT from new vehicles will be replaced by vehicles only a few 
years old. The VMT of these relatively new used vehicles which is 
then replaced by VMT from older used vehicles, and so on.\1726\
---------------------------------------------------------------------------

    \1726\ EDF, Appendix B, NHTSA-2018-0067-12108, at 51.

    The agencies' scrappage model does not capture household choices, 
but uses aggregate data to predict new vehicle sales and age-specific 
scrappage rates in response to changes in new vehicle prices. In 
addition, the scrappage rates of all ages change in response to 
increases in new vehicle prices, not just the oldest vehicles. Further, 
the household that does not buy a new vehicle but holds onto an 
existing vehicle instead, in EDF's example, results in one fewer used 
vehicle supplied to the used market--this will result in an increased 
price for used vehicles and potentially lead to some used vehicles not 
being scrapped. Because the VMT schedules the agencies use in modelling 
show usage declining with age, the agencies' model does assume that 
younger vehicles that are not scrapped are driven more than older 
vehicles that are not scrapped.

[[Page 24640]]

    EDF, IPI, and Honda further argue that mileage accumulation should 
not be constant under all scrappage rates. Specifically, they suggest 
that the assumption that average VMT accumulation by age is constant 
even when scrappage rates decline, results in an overestimate of VMT. 
IPI suggests that the marginally unscrapped vehicles should drag down 
the average VMT accumulation under higher standards in the following 
comment:

    Because those schedules assume each vehicle of a certain age and 
type in the fleet drives a set amount of miles without any 
adjustment for the increase in total fleet size or vehicle quality 
(i.e., wear and tear and durability), the finding that the standards 
cause the fleet size to increase results in a significant increase 
in total VMT.\1727\
---------------------------------------------------------------------------

    \1727\ IPI, Policy Integrity Comments: NHTSA Final--Appendix, 
NHTSA-2018-0067-12213, at 61.

The agencies note that mileage accumulation and scrappage are not 
disjoint. A vehicle that is driven more miles is more likely to be 
scrapped. However, since the National Vehicle Population Profile (NVPP) 
data does not track individual vehicles, there is no obvious way to 
merge individual vehicle odometer readings with those that are 
scrapped. The agencies explored different data sources that could be 
used to capture the joint relationship of the two effects, but 
unfortunately were unable to identify a workable dataset. Furthermore, 
the agencies note that while commenters could be correct about the 
relationship between mileage accumulation and scrappage, they did not 
provide the agencies with any empirical evidence supporting their 
assertions.\1728\ In the meantime, the agencies have adjusted the final 
rule analysis to conservatively assume that total demand for VMT, not 
including the rebound effect, should be constant for all regulatory 
alternatives, as discussed in Section VI.C.1.b)(3)(b)(iv)(d), below. 
This requires that the VMT schedules are no longer constant for all 
fleet sizes.
---------------------------------------------------------------------------

    \1728\ EDF, Appendix B, NHTSA-2018-0067-12108, at 54.
---------------------------------------------------------------------------

(d) Total VMT
    Many commenters think that total VMT, not considering rebound 
miles, should be constant, regardless of the number of new vehicles 
sold and used vehicles scrapped. NCAT, Global, Auto Alliance, CBD, EDF, 
IPI, CARB, and Honda all make this argument. CARB makes the following 
statement suggesting that even a larger fleet size should not increase 
aggregate demand for VMT (again, not including rebound miles):

    A change in the overall fleet size due to the Augural standards 
might not in and of itself be problematic, as long as the VMT 
schedules are adjusted to account for overall travel activity that 
is distributed over a larger number of vehicles. However, the As-
Received version of the [scrappage] model does not adjust VMT 
schedules, with the result that the additional unscrapped vehicles 
inflate total VMT proportionally.\1729\
---------------------------------------------------------------------------

    \1729\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 238.

The agencies agree that the aggregate demand for VMT should be roughly 
constant across alternatives, and stated this in the NPRM, where the 
differences in non-rebound VMT were on the order of 0.4%.
    NERA's modelling efforts found similar small decreases in VMT in 
regulatory alternatives where the standards are relaxed. The Alliance 
stated:

    Under all three scenarios, vehicle miles traveled (``VMT'') 
decreases relative to the augural standards. This is due primarily 
to rebound effects. Because NERA was only examining vehicles through 
MY 2029, the difference in VMT between the alternatives and the 
augural standards decreases over time, since fewer of the MY 2029 
and earlier vehicles are on the road in those later years.\1730\
---------------------------------------------------------------------------

    \1730\ Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073, 
at 11.

    NERA's model used similar assumptions as the NPRM analysis and, 
like the NPRM results, the NERA model results suggest that it is 
plausible that total VMT could decline under less stringent standards. 
A key assumption common to NERA's model and the NPRM analysis is that 
the VMT schedules are constant under all scrappage rates. However, as 
discussed in Section VI.C.1.b)(3)(b)(iv)(c), this can potentially 
overestimate total VMT in the augural case, where vehicles that were 
marginally scrapped in the proposal are kept on the road.
    Presumably, vehicles that are scrapped in the proposal, but not in 
the augural, are in more disrepair than others in the same age cohort. 
As a result, these vehicles would on average be driven less, bringing 
down the average usage of the entire age cohort. This effect could 
alter the relative size of total VMT under the regulatory alternatives, 
as Honda notes in the following comment:

    According to our calculations, if the impact of lowering the 
average cohort's utility is even 0.2% the augural standards would 
become safer than the preferred alternative. We believe that the 
agencies should consider VMT behavior change as part of an effort to 
mature and refine the scrappage model.\1731\
---------------------------------------------------------------------------

    \1731\ Honda, Honda Comment, NHTSA-2018-0067-11818, at 18.

As Honda suggests, a relatively small reduction in the average VMT 
schedules for the more stringent regulatory alternatives could result 
in a change in the direction of the safety impact. This shows the 
importance of investigating the linkage between usage and scrappage 
rates, but also shows that small changes to the total VMT assumptions 
can have meaningful impacts on the predicted effects of the analysis. 
Other commenters make similar points.
    As noted above, the difference in total non-rebound VMT in the NPRM 
analysis was only 0.4%. However, CBD notes that this relatively small 
change in VMT across the alternatives in a single year can result in a 
large number of cumulative additional miles in more stringent 
regulatory alternatives:

    While 0.4% sounds small, when the scrappage model's effect it is 
multiplied by all the VMT that NHTSA includes in its analysis, 
spanning decades, it becomes highly significant--at least 692 
billion additional VMT under the CAFE standards and 894 billion 
under the CO2 program, both relative to the preferred 
alternative.\1732\
---------------------------------------------------------------------------

    \1732\ CBD, Appendix A, NHTSA-2018-0067-12000, at 180.

Since VMT is related to many of the costs and benefits of the program, 
differences in cumulative VMT of this magnitude can have meaningful 
impacts on the incremental net benefit analysis. This point was implied 
by comments from CBD, EDF, NCAT, EAO, and in a paper published by 
academics after the issuance of the NPRM.\1733\ For this reason, the 
agencies have opted to constrain total non-rebound VMT across 
regulatory alternatives.
---------------------------------------------------------------------------

    \1733\ Bento, Antonio M., et al. ``Flawed Analyses of U.S. Auto 
Fuel Economy Standards.'' Science, vol. 362, no. 6419, 2018, pp. 
1119-21., doi:10.1126/science.aav1458.
---------------------------------------------------------------------------

    Such a constraint was suggested by EDF, IPI and other commenters. 
EDF states the following:

    A sophisticated model is not needed to correct this problem. One 
only needs to adjust the VMT added by the ``scrappage model'' so 
that it matches the VMT lost by the sales response model. Put 
another way, used vehicles would be used to the same extent as new 
vehicles since they meet the identical demand (possibly minus a 
rebound effect). \1734\
---------------------------------------------------------------------------

    \1734\ EDF, Appendix B, NHTSA-2018-0067-12108, at 49.

EDF goes on to suggest some potential issues with implementing this 
---------------------------------------------------------------------------
constraint:

    Even this adjustment would still be in favor of the proposal, as 
it assumes that all the VMT lost from fewer new vehicle sales would 
be replaced by used vehicle VMT. This assumes that travel is 
inelastic. This is

[[Page 24641]]

clearly not the case given NHTSA's position on the rebound effect. 
NHTSA must first justify the used vehicle response to any change in 
new vehicle sales. Then, in the unlikely event that this can be 
done, NHTSA must link the scrappage model to the sales response 
model to ensure that the combination of the two models does not 
increase VMT in any calendar year (and probably show a decrease, as 
the overall cost of driving will have increased).\1735\
---------------------------------------------------------------------------

    \1735\ EDF, Appendix B, NHTSA-2018-0067-12108, at 49.

The agencies disagree that lost new vehicle sales would impact the VMT 
of the new vehicles that are sold. The agencies do, however, as EDF 
notes, adjust the VMT of new vehicles to consider changes in the cost 
per mile of travel. In fact, when fuel prices increase, the agencies 
assume that owners of all existing vehicles drive less; the reduction 
will be greater when the vehicles on the road are less efficient, which 
seems consistent with what EDF suggests in the last sentence above. The 
agencies have justified the scrappage effect throughout this 
discussion, above.
    EDF identifies another reason the agencies think a constraint on 
total VMT is reasonable for purpose of the final rule analysis. The 
scrappage, sales, and VMT models each have a certain amount of 
uncertainty associated with it (the uncertainty of the scrappage model 
is discussed in Section VI.C.1.b)(3)(b)(i)(a)), so that when the three 
models are combined, the uncertainty is compounded. EDF characterizes 
these results as being inconsistent with economic theory in the comment 
below:

    We are not aware of any economic arguments which would support 
such an increase. All that can be said is that NHTSA put data from a 
variety of sources through a statistical regression and never 
bothered to see if the results were reasonable or consistent with 
its own economic theory. \1736\
---------------------------------------------------------------------------

    \1736\ EDF, Appendix B, NHTSA-2018-0067-12108, at 57.

The NPRM analysis discussed total fleet size and VMT at length; the 
agencies noted that the fleet was 1.5% bigger for the augural standard 
than the proposal, resulting in 0.4% additional non-rebound VMT in 
CY2050.\1737\ However, given the amount of uncertainty around each of 
the models, and considering that differences in total VMT can have 
meaningful impacts on the cost benefit analysis, the agencies are 
conservatively assuming for the final rule analysis that non-rebound 
VMT is constant, to constrain the outputs derived from the combination 
of the three models.
---------------------------------------------------------------------------

    \1737\ FR, Vol 83, No. 165, August 24, 2018, p.43099.
---------------------------------------------------------------------------

(v) Comments on the Evaluation of Associated Costs and Benefits
(a) Presentation and Valuation of Non-Rebound Miles
    IPI and EDF argued that it was inconsistent to exclude the costs 
and benefits of additional rebound driving but include them for the 
sales and scrappage effect. For example, EDF stated:

    [W]henever a vehicle is driven an additional mile, there is 
value associated with that travel. NHTSA completely ignores the 
value of any additional travel which occurs due to reduced 
scrappage. Including this value would not be an adequate surrogate 
for the additional repair costs required to keep older vehicles on 
the road. Just as NHTSA is now recognizing that rebound VMT is due 
to drivers' express decision to drive more, any driving of older 
vehicles in lieu of new vehicles is due to the same choice. To treat 
these identical choices in 180 degree different manners is of course 
manifestly arbitrary. \1738\
---------------------------------------------------------------------------

    \1738\ EDF, Appendix B, NHTSA-2018-0067-12108, at 58.

    The agencies agree that there is value associated with additional 
miles driven. The NPRM did not directly attribute costs for the loss of 
additional miles in the scrappage analysis when the fleet size shrank. 
The final rule analysis addresses this issue by holding non-rebound 
total VMT constant across regulatory alternatives. However, contrary to 
what EDF suggests above, the cost of additional maintenance and repair 
for otherwise-scrapped vehicles are not directly related to the 
additional miles. The cost of additional maintenance and repair is 
incurred because the value of used vehicles has increased. The increase 
in value of the used vehicles should at least offset the maintenance 
and repair costs.
    Holding aggregate non-rebound VMT constant across alternatives 
addresses IPI's and EDF's concerns that additional miles due to a 
larger fleet size were not adequately valued. However, on average newer 
vehicles tend to be safer, more efficient, more powerful, and more 
spacious than used vehicles. Because of this, driving a newer vehicle 
will be more enjoyable, and provide more utility per mile, than driving 
a used vehicle. Even disregarding trends in vehicle quality, the 
utility of a mile driven in a newer vehicle is on average higher than 
that driven in an older vehicle because the average newer vehicles in 
better condition. The regulation is responsible for the shift in the 
distribution of miles driven at each vehicle age. Including the 
additional safety risks and fuel costs accrued from more miles being 
driven by older vehicles accounts for part of the reduction in the 
utility of the average mile under more stringent standards. Quantifying 
the remaining change in utility of more miles being driven by older 
vehicles is currently beyond the scope of this rulemaking analysis and 
will require extensive future research. The agencies do not think 
excluding other sources of changes in the utility of driving 
(performance, comfort, etc.) will significant change the outcome of the 
analysis.
(b) Increase in Maintenance and Repair Costs and Used Vehicle Values
    EDF and others also commented that the agencies should include the 
value of additional maintenance and repair costs and the increase in 
value for used vehicles explicitly in the cost and benefit analysis. 
They state the following:

    ``It is important to note that NHTSA fails to account for three 
large economic impacts occurring during this process.
    1. The increase in value of the entire used vehicle fleet from 
2017-2050. This is a windfall gain for all current vehicle owners 
that is completely ignored;
    2. The cost of repairing and maintaining the older vehicles 
which are no longer scrapped;
    3. The value of the additional driving that these vehicles 
provide.
    NHTSA only counts the costs related to the additional driving 
performed by the non-scrapped vehicles. Again, NHTSA's decision to 
only include this cost maximizes monetary costs related to the 
current standards and minimizes those related to the proposal.'' 
\1739\
---------------------------------------------------------------------------

    \1739\ EDF, Appendix B, NHTSA-2018-0067-12108, at 50.

    As discussed above, in Section VI.D.1.b)(3)(a)(a), the agencies 
hold the non-rebound fleetwide VMT constant to an exogenous projection 
of aggregate VMT. This addresses EDF's third concern, above. Without a 
model of the used vehicle market it is impossible for the agencies to 
estimate the value increase of used vehicles due to a substitution 
towards used vehicles when new vehicle prices increase. However, the 
maintenance and repair costs should be less than or equal to the 
increase in vehicle value (or the current owner would not pay to 
maintain the vehicle). Not including the additional maintenance and 
repair costs should at least partially offset not including the 
increase in the value of used vehicles. The remaining increase in 
vehicle value should be a transfer between the seller and buyer of a 
used vehicle so that it should be both a cost and benefit exactly 
offsetting. Thus, the total costs and benefits are understated by the 
same amount, and including them

[[Page 24642]]

should not affect the reported net benefits of the rule.
(c) Scrappage Effects From MY2030 and Beyond
    The NPRM analysis considered cost per mile as a continuous 
variable, and new vehicle prices in discrete levels. This means that 
persistently higher new vehicle prices in more stringent standards 
would continue to suppress the scrappage rate of existing vehicles. It 
also means that higher fuel economies in more stringent scenarios would 
continue to affect the scrappage rates as well. EDF noted that the cost 
and benefit accounting that considered the costs and benefits accruing 
to the remaining lifetimes of MYs 1977-2029 included some of the costs 
of the scrappage effect due to the higher prices of MYs beyond 2030, 
but did not include the benefits of the reduced fuel economy for these 
MYs. EDF proposed that the agencies consider a CY analysis instead of 
the model year presented in the NPRM:

    [A] 2017-50 CY analysis would include the operation of 2017-2029 
MY vehicles through CY 2050. This would include the any scrappage 
effects on these vehicles through 2050, consistent with the 
inclusion of new 2050 MY vehicles in the analysis. Some of the 
operation of all the 2017-2029 MY vehicles would be excluded from 
the analysis, as these vehicles are not assumed to be scrapped in 
the Volpe Model until CY 2052-2068. Such an analysis would include 
the benefits over the clear majority of the operation of 2017-2029 
MY vehicles compared to both the shorter calendar year analysis and 
NHTSA's 1977-2029 MY analysis. It would also include the scrappage 
effects caused by 2017-2050 MY vehicles through CY 2050. Any 
scrappage effects would be applied to 2030-2050 MY vehicles, as well 
as 2017-2029 MY vehicles.\1740\
---------------------------------------------------------------------------

    \1740\ EDF, Appendix B, NHTSA-2018-0067-12108, at 22.

However, as the commenter also notes, a CY analysis would exclude some 
of the lifetime costs and benefits of improving the fuel economy of MYs 
impacted by the rule (MYs 2017-2029). For this reason, the agencies do 
not think that a CY analysis should supplant the MY perspective shown 
in the NPRM.
    EDF presents an alternative to switching to a CY analysis which 
would exclude the scrappage effects due to differences in the prices 
and fuel efficiencies of MYs not included in the cost benefit analysis 
(MY 2030 and beyond):

    An alternative that keeps the model year structure of NHTSA's 
1977-2029 MY analysis would be to modify it by removing any 
scrappage effects occurring in 2030 CY and beyond. This analysis 
would still have the disadvantage of barely including any vehicles 
which reflect full compliance with the current and proposed 
standards in 2025. However, it would at least remove the primary 
problem with NHTSA's current MY analysis. The impact of including 
the scrappage effects caused by 2030 and later MY vehicles simply 
and straightforwardly increases the VMT of used vehicles under the 
current standards.\1741\
---------------------------------------------------------------------------

    \1741\ EDF, Appendix B, NHTSA-2018-0067-12108, at 23.

The agencies note that previous analyses have not considered the costs 
and benefits of MYs beyond those which could be a response to the 
change in the considered set of standards. Part of the reason for this 
was that future standards are unknown, and without existing standards 
in place, manufacturers may choose to shift application of fuel saving 
technologies to increases in vehicle performance or safety. The CAFE 
model does not currently simulate such actions, so that including MYs 
too far into the future may overstate the costs and benefits of the 
rule.
    While the agencies disagree that excluding cost and benefits of MYs 
beyond 2030 is an issue for the cost benefit analysis, the agencies 
agree that allowing persistently higher prices and fuel economies of 
future MYs to impact the scrappage of the on-road fleet but not 
considering the costs and benefits of those MYs is inconsistent. 
However, changes to the scrappage model mitigate this issue. As noted 
in Section VI.C.1.b)(3)(b)(i)(c) and VI.C.1.b)(3)(b)(ii), updates to 
the time series strategy and the way that new vehicle fuel economy is 
modelled in the FRM scrappage model change the form of how new vehicle 
prices and fuel economy enter the equation. First, addressing the 
autocorrelation by taking the first difference of variables with first 
order integration instead of including lags of the dependent variables 
means that cost per mile variables and new vehicle prices are captured 
as changes rather than in levels. This means that constant, but higher, 
new vehicle prices in the augural standards will not continue to impact 
the scrappage rates of existing vehicles. More specifically, higher 
prices of MYs 2030 and beyond in the augural case will no longer result 
in lower scrappage rates for prior MYs. Further, since new vehicle cost 
per mile is no longer explicitly included, but rather the amount of 
fuel savings consumers of new vehicles value at the time of purchase is 
excluded from the new vehicle prices series, differences in new vehicle 
fuel economies for MYs beyond 2029 will no longer impact the scrappage 
rates of earlier MYs. This naturally takes care of the concern raised 
by several commenters that the accounting for costs and benefits due to 
changes in MYs 2030 and beyond was inconsistent due to the scrappage 
model.
(c) Estimation of the FRM Scrappage Models
(i) Framing Dynamic Scrappage Models in the Literature
(a) How Fuel Economy Standards Impact Vehicle Scrappage
    As noted above, any increase in price (net of the portion of 
reduced fuel savings valued by consumers) will increase the expected 
life of used vehicles and reduce the number of new vehicles entering 
the fleet (the Gruenspecht effect). In this way, increased fuel economy 
standards slow the turnover of the fleet and the entrance of any 
regulated attributes tied only to new vehicles. Gruenspecht tested his 
hypothesis in his 1981 dissertation using new vehicle price and other 
determinants of used car prices as a reduced form to approximate used 
car scrappage in response to increasing fuel economy standards.
    Greenspan and Cohen (1996) offer additional foundations from which 
to think about vehicle stock and scrappage. Their work identifies two 
types of scrappage: Engineering scrappage and cyclical scrappage. 
Engineering scrappage represents the physical wear on vehicles which 
results in their being scrapped. Cyclical scrappage represents the 
effects of macroeconomic conditions on the relative value of new and 
used vehicles--under economic growth the demand for new vehicles 
increases and the value of used vehicles declines, resulting in 
increased scrappage. In addition to allowing new vehicle prices to 
affect cyclical vehicle scrappage [agrave] la the Gruenspecht effect, 
Greenspan and Cohen also note that engineering scrappage seemed to 
increase where EPA vehicular-criteria pollutant emissions standards 
also increased; as more costs went towards compliance technologies, 
scrappage increased. In this way, Greenspan and Cohen identify two ways 
that fuel economy standards could affect vehicle scrappage: (1) Through 
increasing new vehicle prices, thereby increasing used vehicle prices, 
and finally, reducing on-road vehicle scrappage, and (2) by shifting 
resources towards fuel-saving technologies--potentially reducing the 
durability of new vehicles.

[[Page 24643]]

(b) Aggregate vs. Atomic Data Sources in the Literature
    One important distinction in literature on vehicle scrappage is 
between those that use atomic vehicle data (data following specific 
individual vehicles), and those that use some level of aggregated data 
(data that counts the total number of vehicles of a given type). The 
decision to scrap a vehicle is made on an individual vehicle basis, and 
relates to the cost of maintaining a vehicle, and the value of the 
vehicle both on the used car market, and as scrap metal. Generally, a 
used car owner will decide to scrap a vehicle when the value of the 
vehicle is less than the value of the vehicle as scrap metal, plus the 
cost to maintain or repair the vehicle. In other words, the owner gets 
more value from scrapping the vehicle than continuing to drive it, or 
from selling it.
    Recent work is able to model scrappage as an atomic decision due to 
the availability of a large database of used vehicle transactions. Work 
by authors including Busse, Knittel, and Zettelmeyer (2013), Sallee, 
West, and Fan (2010), Alcott and Wozny (2013), and Li, Timmins, and von 
Haefen (2009) consider the impact of changes in gasoline prices on used 
vehicle values and scrappage rates. In turn, they consider the impact 
of an increase in used vehicle values on the scrappage rate of those 
vehicles. They find that increases in gasoline prices result in a 
reduction in the scrappage rate of the most fuel efficient vehicles and 
an increase in the scrappage rate of the least fuel efficient vehicles. 
This has important implications for the validity of the average fuel 
economy values linked to model years, and assumed to be constant over 
the life of that model year fleet within this study. Future iterations 
of such studies could further investigate the relationship between fuel 
economy, vehicle usage, and scrappage, as noted in other places in this 
discussion.
    While the decision to scrap a vehicle is made atomically, the data 
available to NHTSA on scrappage rates and variables that influence 
these scrappage rates are aggregate measures. This influences the best 
available methods to measure the impacts of new vehicle prices on 
existing vehicle scrappage. The result is that this study models 
aggregate trends in vehicle scrappage, and not the atomic decisions 
that make up these trends. Many other works within the literature use 
the same data source and general scrappage construct, including those 
by Walker (1968), Park (1977), Greene and Chen (1981), Gruenspecht 
(1981), Gruenspecht (1982), Feeney and Cardebring (1988), Greenspan and 
Cohen (1996), Jacobsen and van Bentham (2015), and Bento, Roth, and 
Zhuo (2016.). These works all use aggregate vehicle registration data 
as the source to compute vehicle scrappage.
    Walker (1968) and Bento, Roth and Zhuo (2016) use aggregate data 
directly to compute the elasticity of scrappage from measures of used 
vehicle prices. Walker (1968) uses the ratio of used vehicle Consumer 
Price Index (CPI) to repair and maintenance CPI. Bento, Roth, and Zhuo 
(2016) use used vehicle prices directly. While the direct measurement 
of the elasticity of scrappage is preferable in a theoretical sense, 
the CAFE model does not predict future values of used vehicles, only 
future prices of new vehicles. For this reason, any model compatible 
with the current CAFE model must estimate a reduced form similar to 
Park (1977), Gruenspecht (1981), and Greenspan and Cohen (1996), who 
use some form of new vehicle prices or the ratio of new vehicle prices 
to maintenance and repair prices to impute some measure of the effect 
of new vehicle prices on vehicle scrappage.
(c) Historical Trends in Vehicle Durability
    Waker (1968), Park (1977), Feeney and Cardebring (1988), Hamilton 
and Macauley (1999), and Bento, Ruth, and Zhuo (2016) all note that 
vehicles change in durability over time. Walker (1968) simply notes a 
significant distinction in expected vehicle lifetimes pre- and post- 
World War I. Park (1977) discusses a `durability factor' set by the 
producer for each year, so that different vintages and makes will have 
varying expected lifecycles. Feeney and Cardebring (1988) show that 
durability of vehicles appears to have generally increased over time 
both in the U.S. and Swedish fleets using registration data from each 
country. They also note that the changes in median lifetime between the 
Swedish and U.S. fleet track well, with a 1.5-year lag in the U.S. 
fleet. This lag is likely due to variation in how the data is 
collected--the Swedish vehicle registration requires a title to 
unregister a vehicle, and therefore gets immediate responses, where the 
U.S. vehicle registration requires re-registration which creates a lag 
in reporting further discussed in Section VI.C.1.b)(3)(c)(ii)(b).
    Hamilton and Macauley (1999) argue for a clear distinction between 
embodied versus disembodied impacts on vehicle longevity. They define 
embodied impacts as inherent durability similar to Park's producer 
supplied `durability factor' and Greenspan's `engineering scrappage' 
and disembodied effects as those which are environmental, not unlike 
Greenspan and Cohen's `cyclical scrappage.' They use calendar year and 
vintage dummy variables to isolate the effects--concluding that the 
environmental factors are greater than any pre-defined `durability 
factor.' Some of their results could be due to some inflexibility of 
assuming model year coefficients are constant over the life of a 
vehicle, and also some correlation between the observed life of the 
later model years of their sample and the `stagflation' \1742\ of the 
1970's. Bento, Ruth, and Zhuo (2016) find that the average vehicle 
lifetime has increased 27 percent from 1969 to 2014 by sub-setting 
their data into three model year cohorts. To implement these findings 
in the scrappage model incorporated into the CAFE model, this study 
takes pains to estimate the effect of durability changes in such a way 
that the historical durability trend can be projected into the future; 
for this reason, the agencies include a continuous `durability' factor 
as a function of model year vintage.
---------------------------------------------------------------------------

    \1742\ Continued high inflation combined with high unemployment 
and slow economic growth.
---------------------------------------------------------------------------

(ii) Polk/IHS Registration Data
    As in the NPRM, NHTSA uses proprietary data on the registered 
vehicle population from IHS/Polk for the scrappage models. IHS/Polk has 
annual snapshots of registered vehicles counts beginning in calendar 
year (CY) 1975 and continuing until CY2017. Notably, the data 
collection procedure changed in CY2002, which requires some special 
consideration (discussed below). The data includes the following 
regulatory classes as defined by NHTSA: Passenger cars, light trucks 
(classes 1 and 2a), and medium and heavy-duty trucks (classes 2b and 
3). Polk separates these vehicles into another classification scheme: 
cars and trucks. Under their schema, pickups, vans, and SUVs are 
treated as trucks, and all other body styles are included as cars. In 
order to build scrappage models to support the model year (MY) 2021-
2026 light duty vehicle (LDV) standards, it was important to separate 
these vehicle types in a way compatible with the existing CAFE model.
(a) Choice of Aggregation Level: Body Style
    Two compatible methods existed by which the agencies could 
aggregate scrappage rates: By regulatory class or by body style. Since, 
for CAFE

[[Page 24644]]

purposes, vans/SUVs are sometimes classified as passenger cars and 
sometimes as light trucks (depending upon vehicle-specific attributes) 
and there was no simple way to reclassify some SUVs as passenger cars 
within the Polk dataset, the agencies chose to aggregate survival 
schedules by body style. This approach is also preferable because it is 
consistent with the level of aggregation of the VMT schedules. Since 
usage and scrappage rates are not independent of each other, if average 
usage rates are meaningfully different at the level of body style, it 
is likely that scrappage rates are as well.
    Once stratified into body style level buckets, the data can be 
aggregated into population counts by vintage and age. These counts 
represent the population of vehicles of a given body style and vintage 
in each calendar year. The difference between the counts of a given 
vintage and vehicle type from one calendar year to the next is assumed 
to represent the number of vehicles of that vintage and type scrapped 
in each year.
(b) Greenspan and Cohen Correction
    One issue with using snapshots of registration databases as the 
basis for computing scrappage rates is that vehicles are not removed 
from registration databases until the last valid registration expires--
for example, if registrations are valid for a year, vehicles will still 
appear to be registered in the calendar year in which they are 
scrapped. To correct for the scrappage that occurs during a calendar 
year, a similar correction as that in Greenspan and Cohen (1996) is 
applied to the Polk dataset. It is assumed that the real on-road count 
of vehicles of a given MY registered in a given CY is best represented 
by the Polk count of the vehicles of that model year in the succeeding 
calendar year (PolkCY+1). For example, the vehicles scrapped 
between CY2000 and CY2001 will still remain in the Polk snapshot from 
CY2000 (PolkCY2000), as they will have been registered at 
some point in that calendar year, and therefore exist in the database. 
Using a simplifying assumption that all States have annual registration 
requirements,\1743\ vehicles scrapped between July 1st, 1999 and July 
1st, 2000 will not have renewed registration between July 1st, 2000 and 
July 1st, 2001, and will not show up in PolkCY2001. The 
vehicles scrapped during CY2000 are therefore represented by the 
difference in count from the CY2000 and CY2001 Polk datasets: 
PolkCY2001-PolkCY2000.
---------------------------------------------------------------------------

    \1743\ In future analysis, it may be possible to work with 
State-level information and incorporate State-specific registration 
requirements in the calculation of scrappage, but this correction is 
beyond the initial scope of this rulemaking analysis. Such an 
approach would be extraordinarily complicated as States can have 
very different registration schemes, and, further, the approach 
would also require estimates of the interstate and international 
migration of registered vehicles.
---------------------------------------------------------------------------

    For new vehicles (vehicles where MY is greater than or equal to 
CY), the count of vehicles will be smaller than the count in the 
following year--not all of the model year cohort will have been sold 
and registered. For these new model years, Greenspan and Cohen assume 
that the Polk counts will capture all vehicles which were present in 
the given calendar year and that approximately one percent of those 
vehicles will be scrapped during the year. Importantly, this analysis 
begins modeling the scrappage of a given model year cohort in: CY = 
MY+2,\1744\ so that the adjustment to new vehicles is not relevant in 
the modeling because it only considers scrappage after the point where 
the on-road count of a given MY vintage has reached its maximum.
---------------------------------------------------------------------------

    \1744\ Calculating scrappage could begin at CY=MY+1, as for most 
model year the vast majority of the fleet will have been sold by 
July 1st of the succeeding CY, but for some exceptional model years, 
the maximum count of vehicles for a vintage in the Polk data set 
occurs at age 2.
---------------------------------------------------------------------------

(c) Polk Data Collection Changes
    Prior to calendar year 2002, Polk vehicle registration data was 
collected as a single snapshot on July 1st of every calendar year. All 
vehicles that are in the registration database at that date are 
included in the dataset. For calendar years 2002 and later, Polk 
changed the timing of the data collection process to December 31st of 
the calendar year. In addition to changing the timing of the data 
collection, Polk updated the process to a rolling sample. That is, they 
consider information from other data sources to remove vehicles from 
the database that have been totaled in crashes before December 31st, 
but may still be active in State registration records.
    The switch to a partially rolling dataset will mean that some of 
the vehicles scrapped in a calendar year will not appear in the dataset 
and their scrappage will wrongly be attributed to the year prior to 
when the vehicle is scrapped. While this is less than ideal, these 
records represent only some of the vehicles scrapped during crashes and 
scrappage rates due to crashes should be relatively constant over the 
2001 to 2002-time period. For these reasons, the agencies expect the 
potential bias from the switch to a partially rolling dataset to be 
limited. Thus, the Greenspan and Cohen adjustment applied does not 
change for the dataset complied from Polk's new collection procedures. 
As indicated in Figure VI-67, the scrappage counts computed from the 
old Polk snapshot series represent vehicles scrapped between July 1st 
of a given calendar year and the succeeding July 1st, and is computed 
for CY1976-2000. The new Polk snapshot series represents vehicles 
scrapped between December 31st of a given calendar year and the 
succeeding calendar year, and is computed for CY2002-2016.

[[Page 24645]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.324

    There is a discontinuity between the old and new methods so that 
the computed scrappage for calendar year 2001 represents the difference 
between the vehicle count reported in PolkCY2002 and 
PolkCY2001. PolkCY2001 represents all vehicles on 
the road as of July 1st, 2000, and PolkCY2002 represents all vehicles 
on the road as of December 31, 2001. For this one timespan, the 
scrappage will represent vehicles scrapped over a 17-month time period, 
rather than a year. For this reason, the CY2001 scrappage data point is 
dropped, and because of the difference in the time period of vehicles 
scrapped under the old and new collection schemes, an indicator for 
scrappage measured before and after CY2001 was considered; however, 
this indicator is not statistically significant, and is dropped from 
the preferred model.
(d) Updated FRM Dataset
    As noted in section II.A.1, some commenters expressed concern about 
the inability of the scrappage model to predict the scrappage rates of 
vehicles over age 20. The inability was in large part due to the 
limited data on the scrappage rates of older vehicles. NHTSA has worked 
with Polk/IHS to construct some of the historical registration 
databases using the new methodology for the purposes of other research. 
As a result, the agency has registration data using both Polk 
collection methods for CY's 2001-2012. Importantly, the old Polk 
dataset censored data on older vehicles, with CY's 1975-1993 including 
vehicles ages 0-15 and each successive CY past 1993 adding one 
additional age to the dataset--so that by 2000 ages 0-22 are included. 
The new datasets do not censor data on older vehicles, giving these 
datasets an advantage over the old datasets--for this reason, NHTSA 
uses as many years of the new data as is available.
    The NPRM analysis also used all of the available data using the new 
methodology at the time of publication (CY's 2005-2015). Since the NPRM 
was published, NHTSA has gained access to registration data using 
Polk's new methodology for CY's 2002-2005 and CY's 2016-2017. Table VI-
158 shows the calendars years of data in the NPRM and the final rule 
datasets by age, as well as the total number of data points for each 
age. There are a total of 330 and 420 data points for ages over 15 in 
the NPRM and final rule datasets, respectively. That represents almost 
a 30 percent increase in the number of data points for vehicles over 
15, and a 50 percent increase in the number of data points for the 
oldest vehicles considered in the dataset (ages 27-39). This additional 
data on older vehicles allows the new scrappage models to better 
predict the survival rates of older vehicles than the NPRM models.

[[Page 24646]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.325

(e) Models of the Gruenspecht Effect Used in Other Policy 
Considerations
    This is not the first estimation of the `Gruenspecht Effect' for 
rulemaking policy considerations. In their Technical Support Document 
(TSD) for its 2004 proposal to reduce emissions from motor vehicles, 
CARB outlined how they utilized the CARBITS vehicle transaction choice 
model in an attempt to capture the effect of increasing new vehicle 
prices on vehicle replacement rates. They considered data from the 
National Personal Transportation Survey (NPTS) as a source of revealed 
preferences and a University of California (UC) study as a source of 
stated preferences for the purchase and sale of household fleets under 
different prices and attributes (including fuel economy) of new 
vehicles.
    The transaction choice model represents the addition and deletion 
of a vehicle from a household fleet within a short period of time as a 
``replacement'' of a vehicle, rather than as two separate actions. 
CARB's final data set consists of 790 vehicle replacements, 292 
additions, and 213 deletions; they do not include the deletions, but 
assume any vehicle over 19 years old that is sold is scrapped. This 
allowed CARB to capture a slowing of vehicle replacement under higher 
new vehicle prices. That said, because their model does not include 
deletions, it does not explicitly model vehicle scrappage, but assumes 
all vehicles aged 20 and older are scrapped rather than resold. CARB 
calibrated the model so that the overall fleet size is benchmarked to 
Emissions FACtors (EMFAC) fleet predictions for the starting year; the 
simulation then produced estimates that match the EMFAC predictions 
without further calibration.
    The CARB study captures the effect on new vehicle prices on the 
fleet replacement rates, and offers some precedence for including an 
estimate of the Gruenspecht Effect. However, because vehicles that 
exited the fleet without replacement were excluded, the agencies do not 
learn the effect of new vehicle prices on scrappage rates where the 
scrapped vehicle is not replaced. New and used vehicles are 
substitutes, and therefore the agencies expect used vehicle prices to 
increase with new vehicle prices. And because higher used vehicle 
prices will lower the number of vehicles whose cost of maintenance is 
higher than their value, the agencies expect the replacements of used 
vehicles to slow, but the agencies also expect that some vehicles that 
would have been scrapped without replacement under lower new vehicle 
prices will now remain on the road because their value will have 
increased. The agencies' aggregate measures of the Gruenspecht effect 
includes changes to scrappage rates both from slower replacement rates, 
and from slower non-replacement scrappage rates.
(f) Car Allowance Rebate System (`Cash for Clunkers')
    On June 14, 2009, the Car Allowance Rebate System (CARS) became 
law, with the intent to stimulate the economy through automobile sales 
and accelerate the retirement of older, less fuel efficient and less 
safe vehicles. The program offered a $3,500 to $4,500 rebate for 
vehicles traded-in for the purchase of a new vehicle. Vehicles were 
subject to several program eligibility criteria: First, the vehicle had 
to be drivable and continuously registered and insured by the same 
owner for at least one year; second, the vehicle had to be less than 25 
years old; third, the MSRP had to be less than $45,000; and finally, 
the new vehicle purchased had to be more efficient than the trade-in 
vehicle by a specified margin. The fuel economy improvement 
requirements by body style for specific rebates are presented in Table 
VI-159.

[[Page 24647]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.326

    The program was originally budgeted for $1 billion dollars and to 
end on November 1, 2009, but that amount was spent far more quickly 
than expected and the program received an additional $1.85 billion in 
funding. Even with that additional funding, the program only lasted 
through August 25, 2009, expending $2.85 billion on 678,359 eligible 
transactions. To ensure that the replaced vehicles did not remain on 
the road, the vehicles were scrapped at the point of trade-in by 
destroying the engine. While the program resulted in the replacement of 
more vehicles and at a faster rate than expected, critics have argued 
that many of the trade-ins would have happened even if the program had 
not been in place, so that any economic stimulus to the automobile 
industry during the crisis cannot be attributable to the CARS program. 
Further, others have argued that forcing the scrappage of vehicles that 
could still remain on the road has negative environmental impacts that 
could outweigh any environmental benefits of the reduced fuel 
consumption from the accelerated retirement of these less efficient 
vehicles.
    Li, Linn, and Spiller (2010) use Canada as a counterfactual example 
to identify the portion of CARS trade-ins attributable to the policy, 
i.e., trade-ins that would not have happened anywhere if the program 
were not in place. They argue that the Canadian market is largely 
similar to the U.S. market, in part based upon the fact that 13 to 14 
percent of households purchased new vehicles one year pre-recession in 
both countries. They also argue that the economic crisis affected the 
Canadian economy in a similar manner as it affected the U.S. economy. 
While they note that Canada offered a small rebate of $300 to vehicles 
traded in during January, 2009, hey further note that only 60,000 
vehicles were traded in under that program. Using those assumptions, 
Li, et al., applied a difference-in-difference methodology to isolate 
the effect of the CARS program on the scrappage of eligible vehicles. 
Li, et al., found a significant increase in the scrappage only for 
eligible U.S. vehicles, suggesting they isolated the effect of the 
policy. They conclude that of the 678,359 trade-ins made under the 
program, 370,000 of those would not have happened during July and 
August 2009. They conclude that the CARS program reduced gasoline 
consumption by 0.9-2.9 billion gallons, at $0.89-$2.80 per gallon 
saved.
    The agencies find the evidence from Li, et al., persuasive toward 
the inclusion of a control for the CARS program during calendar year 
2009. The importance is discussed further both in the data section, 
Section VI.C.1.b)(3)(c)(ii), which provides more evidence for the 
effect of the CARS program, and in the model specifications Section 
VI.C.1.b)(3)(c)(iii), which describes the control used for the effect 
of the program. This ensures that the measurements of other determining 
factors are not biased by the exceptional scrappage observed in 
calendar year 2009.
(iii) Updated Final Rule Modeling
    The agencies contemplated all of the comments and suggestions made 
by commenters and, in response, have made several changes to final 
rule's model. First, the agencies changed the time-series strategy used 
in the model, as discussed in Section VI.C.1.b)(3)(c)(iii)(a). This 
change allows the agencies to simplify the models significantly, 
addressing commenters' concerns about potential overfitting of the 
model and difficulty of interpreting individual coefficient values 
(discussed in Section VI.C.1.b)(3)(b)(i)). Second, the agencies changed 
the modeling of the durability effect as discussed in Section 
VI.C.1.b)(3)(c)(iii)(c); this change reduces the reliance on the decay 
function and has the added benefit of addressing concerns about 
overfitting and out-of-sample projections discussed in Section 
VI.C.1.b)(3)(b)(i). Third, a portion of anticipated fuel savings from 
increased fuel economy are netted from new vehicle prices--meaning 
consumers are now assumed to value fuel economy at the time of purchase 
to a certain extent--as discussed in Section VI.C.1.b)(3)(c)(iii)(d). 
This change is in response to comments discussed in Section 
VI.C.1.b)(3)(b)(ii) and addresses inconsistent treatment of consumer 
valuation within the NPRM's analysis. Finally, the agencies consider 
the inclusion of additional or alternative variables in the scrappage 
model in response to comments discussed in Section 
VI.C.1.b)(3)(b)(iii). After extensive testing, the agencies concluded 
that these additional variables do not improve the model fits or would 
introduce autocorrelation in the error structures (see Sections 
VI.C.1.b)(3)(c)(iii)(e) and VI.C.1.b)(3)(c)(iii)(f) for further 
discussion). As such, the agencies rejected the additional terms 
suggested by commenters. Input from commenters was used to simplify the 
scrappage model, make it more consistent with modeling of new vehicle 
prices elsewhere in the analysis, and improve its predictions for the 
instantaneous scrappage rates of vehicles beyond age 20.
(a) Changes to the Time Series Strategy
    As discussed in Section VI.D.1.b)(3)(b)(i)(c), the agencies 
reconsidered the time series strategy for the final rule in response to 
comments. The first step in doing so is to test the time series 
properties of the dependent and independent variables. The agencies use 
the Augmented Dickey-Fuller (ADF) unit root test implemented in the 
`CADFtest' R package to test for stationarity.\1745\ The agencies find 
that the logistic scrappage rate is I(0), or stationary in levels. 
Since the dependent variable is stationary, there is no long-term trend 
in scrappage rates to capture. Lags of dependent variables need not be 
included, but their stationary forms should be used in the regressions. 
The following table summarizes the order of integration of each of the 
considered regressions; the

[[Page 24648]]

regression forms represent the form of the variable that is included in 
the considered models.\1746\ All the variables considered are either 
I(0) or I(1), meaning that they should be run in either levels or first 
differences, respectively. This significantly simplifies the 
regressions. Two unintended, positive outcomes of this change in time 
series strategy are that the coefficients on variables are easier to 
interpret and the models are less likely to be overfit. In this way, 
the shift to address concerns about the time series strategy (discussed 
in Section VI.D.1.b)(3)(b)(i)(c)) also addresses commenter concerns 
outlined in Section VI.D.1.b)(3)(b)(i)(a).
---------------------------------------------------------------------------

    \1745\ Lupi, Claudio (2019, September 7). Package `CAFtest.' 
Retrieved from https://cran.r-project.org/web/packages/CADFtest/CADFtest.pdf.
    \1746\ Note: Some of these variables were considered or added in 
response to comments presented in Sections I.A.1.a)(1)(b)(ii), 
I.A.1.a)(1)(b)(iii), and I.A.1.a)(1)(b)(iv), and may not be present 
in the NPRM.
[GRAPHIC] [TIFF OMITTED] TR30AP20.327

(b) Final Rule Preferred and Sensitivity Specifications
    After consideration of comments on, and subsequent peer review of, 
the NPRM analysis, the agencies updated the scrappage model 
specifications for the final rule. Section VI.C.1.b)(3)(c)(iii)(a) 
through VI.C.1.b)(3)(c)(iii)(f) discuss other considered specifications 
and variables. The equation below represents the final

[[Page 24649]]

form of the scrappage equation included in the central and sensitivity 
analysis:
[GRAPHIC] [TIFF OMITTED] TR30AP20.328

    Here, ``S'' represents the instantaneous scrappage rate in a 
period, so that the dependent variable is the logit form of the 
scrappage rates. Logit models ensure that predicted values are 
bounded--in this case between zero and one. It is not possible to scrap 
more than all the remaining vehicles, nor fewer than zero percent of 
them, which is illustrated in the graph below:
[GRAPHIC] [TIFF OMITTED] TR30AP20.329

[GRAPHIC] [TIFF OMITTED] TR30AP20.330

Solving for instantaneous scrappage yields the following:

[[Page 24650]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.331

    In the equation above, [Sigma][beta]iXi represents the right-hand 
side of the above model specification. Within the right-hand side of 
the equation, Age represents the age of the model year cohort in a 
specific calendar year, defined by the Greenspan and Cohen adjustment 
discussed in Section VI.C.1.b)(3)(c)(ii)(b). The coefficient on the 
cubic age term is assumed to be zero for the van/SUV and pickup 
specifications as this term is not necessary to capture the general 
scrappage trend for these body styles. Share Remaining represents the 
share of the original cohort remaining at the start of the period. 
These two components represent the engineering portion of scrappage--
the inherent durability of a model year and the natural life cycle of 
how vehicles scrap out of a model year cohort as the cohort increases 
with age. The determination of these specific forms is discussed in 
detail in Section VI.C.1.b)(3)(c)(iii)(g).
    New Price--FS represents the average price of new vehicles minus 30 
months of fuel savings for all body styles. The central analysis 
assumes the coefficient on the age interactions for this term are zero 
for all body styles, but a sensitivity case allows the elasticity of 
scrappage to vary with age. Fuel Price represents the real fuel prices, 
weighted by fuel share of the model year cohort being scrapped. CP100M 
represents the cost per 100 miles of travel for the specific body style 
of the model year cohort being scrapped under the current period fuel 
prices and using fuel shares for that model year cohort. These measures 
capture the response of scrappage rates to new vehicle prices, fuel 
savings, and to changes in fuel prices that make the used model year 
cohort more or less expensive to operate. Because these measures are 
all I(1), as discussed above in 0, the first difference of all of these 
variables is used in modelling. The other specific modelling 
considerations that resulted in this form of modelling the new and used 
vehicles markets are discussed in Section VI.C.1.b)(3)(c)(iii)(d).
    GDP Growth represents the GDP growth rate for the current period. 
This captures the cyclical components of the macro-economy. Section 
VI.C.1.b)(3)(c)(iii)(e) discusses how this specific measure was chosen, 
and what other measures were considered as alternative or additional 
independent variables.
    CY2009 and CY2010 represent calendar year dummies for 2009 and 2010 
when the CARS program was in effect; this controls for the impact of 
the program. [Age = 25] represents an indicator for vehicles 
25 years and older. The interaction of the calendar year dummies with 
this indicator allows for the effect of the CARS program to be 
different for vehicles under 25 versus vehicles 25 and older. Since 
only vehicles under 25 were eligible for the program (see the 
discussion of the program in Section VI.C.1.b)(3)(c)(ii)(f)), this 
flexibility is important to correctly control for the program.
    Finally, FE represents a set of model year fixed effects used to 
control for heterogeneity across different model years. This is related 
to the durability and engineering scrappage. The NPRM model did not 
include fixed effects because it fit a parametric relationship to model 
year as a continuous variable as a way to capture durability. This 
change in how the durability effect is modelled is discussed further in 
Section. Further, Section VI.C.1.b)(3)(c)(iii)(g) discusses trends in 
the fixed effects and how these are projected forward within the CAFE 
model.
(c) Modeling Durability Trends Over Time
    As noted in the NPRM, the durability of successive model years 
generally increases over time. However, this trend is not constant with 
vehicle age--the instantaneous scrappage rate of vehicles is generally 
lower for later vintages up to a certain age, but increases thereafter 
so that the final share of vehicles remaining converges to a similar 
share remaining for historically observed vintages. The NPRM 
parameterized this trend by using the natural log of the model year as 
a continuous variable interacted with a polynomial form of the age 
variable--this predicted an increasing but diminishing trend in vehicle 
durability for younger ages. The analysis for the final rule makes a 
change that allows more flexibility in durability trends. Below, the 
agencies consider the survival and scrappage patterns by body style.
    Figure VI-69 to Figure VI-71 shows the survival and scrappage 
patterns of different vintages with vehicle age for cars, SUVs/vans and 
pickups, respectively. Cars have the most pronounced durability 
pattern. Figure VI-69 shows that newer vintages scrap slower at first, 
but that scrap more heavily so that the final share remaining of cars 
is more or less constant by age 25 for all vintages.

[[Page 24651]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.332

    SUVs/vans have a less pronounced durability pattern. Model year 
1980 actually lives longer than model years 1985 and 1990. This is 
likely due to a switch of SUVs/vans to be based on car chassis rather 
than pickup chasses over time. However, through the later model years, 
the durability trend is more like that of cars. The lack of a 
continuous trend in durability of SUVs/vans make how this trend is 
captured particularly important. Below the agencies discuss a change in 
how the durability trend is modelled for the final rule, which is more 
flexible than the NPRM model.
[GRAPHIC] [TIFF OMITTED] TR30AP20.333

    There is no clear trend in durability for pickups. Like SUVs/vans, 
this makes parameterizing by using a form of vintage as a continuous 
variable problematic. Such a parametric form does not allow for each 
model year to have its own durability pattern.

[[Page 24652]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.334

    As noted above, the NPRM model used the natural log of model year 
as a continuous variable interacted with age to capture an increasing 
but diminishing trend of vehicle durability for the younger ages. 
However, enforcing a parametric form on a continuous model year 
excluded the possibility of including model year specific fixed effects 
and required that durability have a parametric trend with successive 
vintages. As seen above, SUVs/vans and pickups certainly do not follow 
such a trend, so that this constraint was too restrictive, at least for 
these body styles. The final rule analysis makes an adjustment that 
allows for an initial increase in the durability of a model year to 
persist, while including fixed effects and relaxing the parametric 
assumption.
    Instead of regressing the natural log of the vintage share in the 
remaining models, shown in Table VI-161 through Table VI-163, the 
agencies use the share remaining in the previous period as an 
independent variable. Since the logistic instantaneous scrappage rate 
is stationary (it is independent of the previous periods' logistic 
instantaneous scrappage rate), the share remaining should not be 
endogenous. The share remaining models for the final rule include model 
year specific fixed effects and project a linear trend in durability by 
fitting a regression through the fixed effects. This latter part still 
requires a parametric assumption about durability (discussed in Section 
VI.C.1.b)(3)(c)(iii)(g)), but not while jointly estimating other 
coefficients. In this way, the other coefficients should not be biased 
by projecting the durability trend forwards in the implementation of 
the scrappage regressions within the CAFE model.
BILLING CODE 4910-59-P

[[Page 24653]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.335


[[Page 24654]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.336


[[Page 24655]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.337

BILLING CODE 4910-59-C
    As Table VI-161 shows, the NPRM specification and both the constant 
and the quadratic forms of the age interaction with the share remaining 
variable to capture the durability effect show evidence of 
autocorrelation. The linear form of the interaction of age and share 
remaining does not show evidence of autocorrelation and also has the 
lowest AIC and highest adjusted R-squared. For these reasons, this is 
the preferred specification of the durability effect. Since the share 
remaining coefficient is negative and larger than the positive 
coefficient on the share remaining interacted with age, a cohort that 
has a higher share remaining at an early age will have a lower 
instantaneous scrappage rate in this period until a certain age and 
then a higher scrappage rate after that age. To find the age where the 
sign of the share remaining coefficient will switch from predicting a 
lower instantaneous scrappage rate to a higher one, the agencies must 
take the ratio of the coefficient on the share remaining variable to 
the share remaining interacted with age--this suggests that at age 19, 
the sign of the share remaining variable flips. That is, the 
instantaneous scrappage rate of cars is predicted to be lower if the 
share remaining is higher until age 18, after which a higher share 
remaining predicts a higher instantaneous scrappage rate.
    As Table VI-162 shows, the linear interaction of age and share 
remaining is the only specification of the durability effect for SUVs/
vans that do not show autocorrelation in the error structure. The 
linear interaction of age and share remaining has the lowest AIC and 
highest R-squared; for this reason, this is the preferred specification 
of the durability effect for SUVs/vans. The signs for share remaining 
and share

[[Page 24656]]

remaining interacted with age show a similar trend as that to cars. 
Taking the ratio again of the share remaining to the share remaining 
interacted with age, for ages 0 to 18 a higher share remaining predicts 
lower instantaneous scrappage, and for ages beyond 18 it predicts a 
higher instantaneous scrappage rate.
    As Table VI-163 shows, all but the NPRM specification of the 
durability effect for pickups do not show autocorrelation in the error 
structures. However, similar to cars and SUVs/vans, the linear 
interaction of age and share remaining has the lowest AIC and highest 
adjusted R-squared. For this reason, this is the preferred 
specification for all body styles. Taking the ratio of the coefficient 
on share remaining to share remaining interacted with age shows that a 
higher share remaining will predict a lower instantaneous scrappage 
rate in the next period for ages 0 through 14, but a higher 
instantaneous scrappage rate for ages 15 and older.
    Using the preferred forms of the engineering scrappage rates for 
each body style as the reference point, Section VI.C.1.b)(3)(c)(iii)(d) 
considers different forms to predict the Gruenspecht effect for each 
body style. Section VI.C.1.b)(3)(c)(iii)(e) uses the preferred 
engineering and Gruenspecht forms to consider alternative macroeconomic 
variables to predict the effects of the business cycle. Finally, 
Section VI.C.1.b)(3)(c)(iii)(f) uses the preferred engineering, 
Gruenspecht and business cycle forms to consider the inclusion of other 
additional independent variables.
(d) Modeling Impacts of New Vehicle Market on Used Scrappage Rates
    Table VI-164 through Table VI-166 show the relationship between 
car, SUV/van, and pickup scrappage rates and changes in new vehicle 
price and fuel economies. The agencies consider two methods in response 
to comments outlined in Section VI.C.1.b)(3)(b)(ii). (1) changes in 
average new vehicle prices net of 30 months of fuel savings (consistent 
with the technology selection and sales model) and (2) change in 
average new vehicle prices, change in average fuel prices, changes in 
new vehicle cost per mile and changes in new vehicle fuel consumption. 
The agencies allow the elasticity of average new vehicle prices net of 
30 months of fuel savings to vary by age by including interaction 
terms.
BILLING CODE 4910-59-P

[[Page 24657]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.338


[[Page 24658]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.339


[[Page 24659]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.340

    For all body styles, the specification of the Gruenspecht effect as 
the change in new vehicle prices net of fuel savings does not show 
signs of auto-correlated errors. However, for cars and vans/SUVs, the 
specification which separates the effect of new vehicle prices and fuel 
economy does show evidence of autocorrelation. For this reason, the 
changes in new vehicle fuel prices net of fuel savings is the preferred 
specification of the Gruenspecht effect.
    The agencies consider the interaction of the change in average new 
vehicle prices with vehicle age. This relaxes an assumption that the 
elasticity of scrappage rates to change in new vehicle prices is 
constant. For cars and

[[Page 24660]]

vans/SUVs the linear interaction of change to new vehicle prices net of 
fuel savings show evidence of autocorrelation. The quadratic 
interaction of age with change in new vehicle prices shows 
autocorrelation with cars. For this reason, the agencies consider the 
constant elasticity of scrappage rates to changes in new vehicle prices 
to be the preferred specification (as the only specification that does 
not show evidence of autocorrelation for all body styles). However, the 
agencies do consider the quadratic form of the elasticity with age as a 
sensitivity case (even though there is evidence of autocorrelation (but 
only in the car specification)). This allows the agencies to test the 
impact of relaxing the assumption around constant elasticity on CAFE 
model outcomes.
(e) Considering Alternative/Additional Macroeconomic Indicators
    Table VI-167 through Table VI-169 show alternative macroeconomic 
indicators for cars, vans/SUVs and pickups, respectively. The agencies 
consider unemployment rate and per capita personal disposable income as 
alternatives to GDP growth rate to capture the cyclical component of 
the macro economy. The unemployment rate and the per capita personal 
disposable income are both I(1), so that the first difference of each 
is the form included. For the car and van/SUV specifications, the 
specifications replacing GDP growth rate show evidence of 
autocorrelation in the error structures. For this reason, the GDP 
growth rate is the preferred specification for the cyclical components 
of instantaneous scrappage rates, as in the NPRM models.
    As discussed in Section VI.D.1.b)(3)(b)(iii)(c), some commenters 
were concerned with the exclusion of interest rates. In response, the 
agencies considered including the change in interest rates for the 
otherwise preferred specification. For vans/SUVs the model has a higher 
AIC and shows evidence of autocorrelation in the error structures. For 
pickups, the sign changes on the change in cost per mile when the 
interest rate is included, which would be an implausible result. 
Finally, the AIC for cars is nearly identical regardless as to whether 
the interest rate is included. For these reasons, the agencies continue 
to exclude the interest rate from the preferred specification.

[[Page 24661]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.341


[[Page 24662]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.342


[[Page 24663]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.343


[[Page 24664]]


(f) Considering Other Additional Variables
    Table VI-170 through Table VI-172 show specifications that consider 
additional variables not included in the preferred specifications. As 
discussed in Section VI.D.1.b)(3)(b)(iii)(a), some commenters 
criticized the fact that maintenance and repair costs were excluded 
from the scrappage models. In response to comments, and since the 
maintenance and repair costs are I(1), the agencies considered 
including the difference in maintenance and repair costs. When 
included, changes in maintenance and repair costs show the expected 
sign--when maintenance and repair costs are higher, instantaneous 
scrappage rates are predicted to be higher (as used vehicles are more 
expensive to maintain). When included, the AIC is higher for the car 
and van/SUV specifications. That is, including the change in 
maintenance and repair costs does not improve the fit of the models. 
Because of this, and because there is no obvious way to predict future 
change to maintenance and repair costs (as discussed in the NPRM), the 
preferred specification continues to exclude maintenance and repair 
costs.
    As discussed in Section VI.D.1.b)(3)(b)(iii)(b), some commenters 
criticized the exclusion of steel and iron scrap prices from the 
scrappage models. In response to comments, and since this variable is 
also I(1), the agencies considered including the change in steel and 
iron scrap prices. When included, the AIC of cars and vans/SUVs is 
higher. Further, the car specification includes evidence of 
autocorrelation in the error structures. In addition, there is no known 
projection of steel and iron scrappage prices, so that the agencies 
would have to make projections to include this variable in the 
scrappage models. Accordingly, the central case continues to exclude 
steel and iron scrap prices.
    As discussed in Section VI.D.1.b)(3)(b)(iii)(d), some commenters 
and peer reviewers suggested that controlling for aggregate measures of 
model year cohorts, such as performance, might correct some unexpected 
signs. The preferred specification already addresses these concerns. 
Further, because fixed effects are included for model years, the 
agencies cannot include aggregate model year specific attributes that 
are constant over the lifetime of the cohort. The agencies do consider 
the ratio of the average horsepower to weight of a model year cohort to 
the new vehicle cohort, as this will change along with changes to the 
horsepower to weight ratio over successive calendar years. Including 
this variable results in a higher AIC for cars and vans/SUVs and shows 
evidence of autocorrelation in the errors for these two body styles. 
For this reason, the preferred specification excludes this metric.
    The agencies also considered including new vehicles sales directly 
as a predictor of instantaneous scrappage rates. Since new vehicle 
sales are I(1), the difference in new vehicle sales is the included 
form. Including the change in new vehicle sales results in a higher AIC 
for cars and vans/SUVs. It also introduces evidence of autocorrelation 
in the error structure for the car model, and reduces the effect of the 
change in fuel prices by two orders of magnitude for vans/SUVs. It 
seems unlikely that the magnitude of the effect of fuel prices would so 
drastically vary between body styles. For these reasons, the preferred 
specifications exclude the change in new vehicles sales. The agencies 
also considered including changes in vehicle stock, but this similarly 
did not improve the fit of the scrappage models--and doing so limited 
the ability to link the sales and scrappage models as some commenters 
suggested (see Sections (b)(iv)(a) and (b)(iv)(b)).

[[Page 24665]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.344


[[Page 24666]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.345


[[Page 24667]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.346

[GRAPHIC] [TIFF OMITTED] TR30AP20.347

BILLING CODE 4910-59-C
(g) Projecting Durability in the CAFE Model
    The left graphs in Figure VI-72 through Figure VI-74 show the fixed 
effects for the preferred scrappage specifications for cars, vans/SUVs, 
and pickups, respectively. For all body styles there is a general 
downward trend in the fixed effects. This suggests an increase in the 
durability of successive model years. However, since the panel datasets 
are not balanced, there is likely potential bias for the fixed effects 
that include only certain ages. This makes projecting the durability 
increase from the fixed effects a little more complicated than merely 
fitting to all fixed effects. First, the agencies must determine what 
part of this trend is likely due to increases in vehicle durability 
(and should be projected forward) and which part of the trend may 
conflate other factors.
    The right graphs in Figure VI-72 through Figure VI-74 show the 
average observed logistic scrappage rates by model year for all ages 
where data exists. As can be seen, the average observed scrappage rates 
decline dramatically for model years after 1996 for all body styles. 
There are two reasons this trend exists. First, as Figure VI-72 through 
Figure VI-74 show, the instantaneous scrappage rate generally follows 
an inverted u-shape with respect to vehicle age. The instantaneous 
scrappage rates generally peak between ages 15 and 20 for all body 
styles. Model year 1996 is the first model year which will be at least 
age 20 at the last date of available data (calendar year 2016). This 
means that all model years newer than 1996 have likely not yet reached 
the age where the instantaneous scrappage rate will be the highest for 
the cohort. Accordingly, the fixed effects could be biased downwards 
(consistent with the sharper downward slope in the fixed effects for 
most body styles for model years beyond 1996) because of the unbalanced 
nature of the panel, and not because of an actual increase in inherent 
vehicle durability for those model years.

[[Page 24668]]

    The second reason the average logistic scrappage rates for model 
years before 1996 is more stable is because each data point in the 
average has increasingly less effect on the average as more data 
exists. For model years 1996 and older there are at least 18 data 
points (we start the scrappage at age 2, by which point effectively all 
of a model year has been sold), and each will have a smaller effect on 
the average than for newer model years with fewer observations. For 
these reasons, the average observed logistic scrappage rate is more 
constant for model years before 1996. As a result, the agencies do not 
consider the trend in fixed effects after model year 1996 to rely on 
enough historical data to represent a trend in vehicle durability, as 
opposed to a trend in the scrappage rate with vehicle age.
    In considering which model year fixed effects should be considered 
in projecting durability trends forward, another important factor is 
whether there are discrete shifts in the types of vehicles that are in 
the market or category of each body style over time. For cars, an 
increasing market share of Japanese automakers which tend to be more 
durable over time might result in fixed effects for earlier model years 
being higher. This trend is shown in the fixed effects in Figure VI-72, 
which follow a steeper trend before model year 1980.
[GRAPHIC] [TIFF OMITTED] TR30AP20.348

    For vans/SUVs, earlier model years are more likely to be built on 
truck chassis (body-on-frame construction) instead of car chassis 
(unibody construction). Since pickups tend to be more durable, the 
earlier fixed effects are likely to be lower for vans/SUVs for earlier 
model years. The 1984 Jeep Cherokee was the first unibody construction 
SUV.\1747\ As Figure VI-73 shows, the fixed effects before 1986 show 
inconsistent trends; these are likely due to changes in what was 
considered a van/SUV over time. For this reason, the agencies build the 
trend of fixed effects from model years 1986 to 1996.
---------------------------------------------------------------------------

    \1747\ https://www.autoguide.com/auto-news/2018/01/10-interesting-facts-from-the-history-of-the-jeep-cherokee.html.

---------------------------------------------------------------------------

[[Page 24669]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.349

[GRAPHIC] [TIFF OMITTED] TR30AP20.350

    While the trend for pickups and cars could be extrapolated before 
1986, the agencies opt to keep the fixed effects included constant for 
all body styles. Thus, the projections are built from model year 1986 
to model year 1996 fixed effects. Table VI-173 below, shows the linear 
regressions shown as the line on the left side of Figure VI-70 through 
Figure VI-72. The durability cap represents the last model year where 
the durability trend is assumed to persist. The agencies cap the 
durability impacts at model year 2000, as data beyond this point does 
not exist for enough ages to determine if durability has continued to 
increase since this point. The implication of this cap, is that model 
years after 2000 are assumed to have the same initial durability as 
model year 2000 vehicles. Since there is a limit to the potential 
durability of vehicles, this acts as a bound on this portion of the 
scrappage model.

[[Page 24670]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.351

    The durability projections enter the scrappage equation in the CAFE 
modelling in accordance to the following equation:
[GRAPHIC] [TIFF OMITTED] TR30AP20.352

    The intercept enters as a constant added to the predicted logistic 
of the instantaneous scrappage rate. The model year slope enters as the 
model year for all model years older than 2000 and enters as 2000 for 
all model years 2000 and newer.
    Once the predicted logistic scrappage rate is calculated in the 
CAFE model (including the projections of the fixed effect portion of 
the equation), the future population of model year cohorts can be 
predicted. The instantaneous scrappage can be calculated directly from 
S. It identifies the share of remaining vehicles in each calendar year 
that are scrapped in the next year. The population of vehicles in the 
next calendar year can be calculated as follows:

Populations MY,CY +1 = Population MY,CY *(1 -SMY,CY).

    This process is iteratively calculated at the end of the CAFE model 
simulation to determine the projected population of each model year in 
each future calendar year. This allows the calculation of vehicle miles 
travelled, fuel usage, pollutant and CO2 emissions, and 
associated costs and benefits. The CAFE model documentation released 
with this final rule further details how the scrappage model is 
projected within the simulations.
(d) Updates to the Decay Function
    The scrappage models described above fit the historical data of car 
and truck scrappage well, but when used to project the scrappage of 
future model years they over-predict the remaining cars and trucks for 
ages greater than 30 in an unrealistic manner. Nearly six percent of 
the MY2015 van/SUV fleet and eight percent of the pickup fleet is 
projected to persist until age 40. This is unrealistic, and likely due 
to the fact that the agencies do not observe enough model years for 
those ages and over-predict the impact of durability increases for 
those ages. For this reason, the agencies are using the curves with an 
accelerated decay function to predict instantaneous scrappage beyond 
age 30 for pickups and SUVs/vans. The implementation and parameter 
stricture of the decay function have not changed since the NPRM model. 
Table VI-174, below, shows the inputs used for the final rule analysis.

[[Page 24671]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.353

    The final survival rate has not changed since the NPRM, but the 
input Decay age has changed. In the NPRM, the decay function was 
specified to begin after age 20, while the decay function begins after 
age 30 in the final rule analysis. This input change was possible 
because the scrappage model for the final rule predicts shares 
remaining in line with observed historical trends through age 30, 
rather than through age 20. This improvement in the model fits for 
older ages is driven both by the shift of the modelling of the 
durability effect discussed in Section VI.D.1.b)(3)(a)(g) and the 
increase in available data on the scrappage rates of older vehicles 
discussed in Section VI.C.1.b)(3)(c)(ii)(d). Overall, this outcome 
suggests that the final rule model predicts the scrappage rates of 
older vehicle better than the NPRM model.
    As in the NPRM, the decay function is implemented in the model 
using the following conditions:
[GRAPHIC] [TIFF OMITTED] TR30AP20.354

Where:
t = (age + 1 - b15
And:
[GRAPHIC] [TIFF OMITTED] TR30AP20.355

    Here, the population for ages beyond the start age of the decay 
function depends on the population of the cohort at that start age and 
the final share expected for that body style at age 40. The rate of 
decay necessary to make the final population count equal that observed 
in the historical data is applied.
(4) The Rebound Effect in the NPRM
    The fuel economy rebound effect--a specific example of the well-
documented energy efficiency rebound effect for energy-consuming 
capital goods--refers to the tendency of motor vehicles' use (as 
measured by vehicle-miles traveled, or VMT) to increase when their fuel 
economy is improved and, as a result, the cost per mile (CPM) of 
driving declines. Amending and establishing CAFE and CO2 
standards at a lower degree of stringency than the baseline level will 
lead to comparatively lower fuel economy for new cars and light trucks, 
thus increasing the amount of fuel consumed to travel each mile. The 
resulting increase in CPM will lead to a reduction in VMT over the 
lifetime of new vehicles, an example of the rebound effect working in 
reverse. In the NPRM, the agencies assumed a fuel rebound effect of 20 
percent, meaning that a 5 percent decrease in fuel economy would result 
in a one percent decrease in the annual number of miles driven at each 
age over a vehicle's lifetime.
    Many of the comments received on different components of the CAFE 
model can be traced back to the agencies' rebound selection. The 
agencies recognize that the value selected for the rebound effect 
influences overall costs and benefits associated with the regulatory 
alternatives under consideration as well as the estimates of lives 
saved under various regulatory alternatives, and that the rebound 
estimate, along with fuel prices, technology costs, and other 
analytical inputs, is part of the body of information that agency 
decision-makers have considered in determining the final levels of the 
CAFE and CO2 standards. The agencies also note that the 
rebound effect diminishes the economic and environmental benefits 
associated with increased fuel efficiency.
    For the analysis supporting the NPRM, the agencies conducted a 
thorough re-examination of the basis for the estimate of the fuel 
economy rebound effect used to analyze the impacts of CAFE and 
CO2 emission standards for model years 2012-16 and 2017-21. 
This was prompted by three developments. First, more recent updates of 
the 2007 study by Small and Van Dender that had provided the basis for 
assuming the 10 percent rebound effect used in those previous analyses 
reported larger values. Second, projected growth in the income measure 
used in those authors' 2007 study, which was anticipated to reduce the 
magnitude of the rebound effect over the future period spanned by those 
analyses, did not occur during the decade following the 2007 study's 
publication. Finally, extensive new research on the rebound effect had 
become available since those previous

[[Page 24672]]

analyses were conducted, and while its findings were mixed, many of 
those more recent studies reported values significantly above the 
agencies' previous 10 percent estimate.
    In the NPRM, the agencies first summarized estimates of the fuel 
economy rebound effect for light-duty vehicles in the U.S. from studies 
conducted through 2011, when the agencies originally surveyed research 
on this subject. As the accompanying discussion in the proposal 
indicated, the research available through 2011 collectively suggested 
that the rebound effect was likely to fall in the range from 20 percent 
to 25 percent, although the then-recent study by Small and Van Dender 
(2007) pointed to smaller values, particularly for future years. The 
agencies then identified 16 additional studies of the rebound effect 
that had been conducted since their original survey, and the NPRM 
discussed the various approaches they used to measure the magnitude of 
the rebound effect, their data sources and estimation procedures, 
reported findings, and strengths and weaknesses of each study.
    Based on this re-examination, the agencies concluded that currently 
available evidence did not appear to support the 10 percent estimate 
relied upon in previous rules, and identified a value of 20 percent as 
more representative of the totality of evidence, including both the 
research covered by the earlier and more recent studies examined in the 
NPRM. While acknowledging the wide range of estimates reported in more 
recent research--which extended from zero to more than 80 percent--the 
agencies noted that the central tendency of recent estimates appeared 
to lie in the same 20-25 percent range suggested by their extensive 
review of earlier research. The agencies also recognized that a 20 
percent estimate differed markedly from the 10 percent estimate used in 
the regulatory analyses for the 2010 and 2012 final rules, but noted 
that it represented a return to the value NHTSA originally used to 
analyze the impacts of CAFE standards for model years prior to 2011.
(a) Comments on the Rebound Effect Used in the NPRM
    The agencies received numerous comments on the decision to revise 
their previous estimate of the rebound effect, virtually all of which 
echoed a few common arguments. First, commenters generally agreed that 
the most appropriate measure for the agencies to rely on is the current 
long-run fuel economy rebound effect for U.S., although a few suggested 
that using an estimate of its short-run value might be 
preferable.\1748\ However, many commenters argued that some of the more 
recent studies the agencies relied upon to support the revised 20 
percent estimate may have limited relevance to the appropriate measure 
for analyzing the current rule, and that the agencies should place more 
emphasis on those that commenters asserted were more appropriate to 
rely upon.
---------------------------------------------------------------------------

    \1748\ See, e.g., RFF, Comments, NHTSA-2018-0067-11789, at 30. 
For an thorough example of the arguments made for a short- to 
medium-term rebound effect, see generally IPI, Appendix, NHTSA-2018-
0067-12213, at 61.
---------------------------------------------------------------------------

    To identify the most relevant research, some commenters proposed 
applying various selection criteria to choose which studies were most 
appropriate to rely on when estimating the value of the rebound effect 
to use in this analysis. While commenters proposed using certain 
criteria as ``filters''--that is, to eliminate any studies that did not 
meet those criteria--they also suggested applying other criteria to 
emphasize studies with particular features they argued made them more 
relevant to identifying the current value of the rebound effect for the 
U.S.\1749\ Among these suggested criteria were the following:
---------------------------------------------------------------------------

    \1749\ See, e.g., IPI, Appendix, NHTSA-2018-0067-12213, at 58-
64; EDF, Analysis of the Value and Application of the Rebound 
Effect, NHTSA-2017-0069-0574, at 16-19; California Office of the 
Attorney General et al., Attachment 1, NHTSA-2017-0069-0625, at 8; 
States and Cities, Attachment 1, Docket No. NHTSA-2018-0067-11735, 
at 78; RFF, Comment, NHTSA-2018-0067-11789, at 3; CARB, Detailed 
Comments, NHTSA-2018-0067-11873, at 120; Aluminum Association, 
Comments, NHTSA-2018-0067-11952, at 5; NCAT, Appendix A, NHTSA-2018-
0067-11969, at 34; and North Carolina Department of Environmental 
Quality, Comments, NHTSA-2018-0067-12025, at 12; among others. EPA's 
Science Advisory Board shared similar policy opinions. SAB at 26-27.
---------------------------------------------------------------------------

     Exclude estimates based upon data from outside the U.S.;
     Include only estimates based upon ``more recent'' data, 
usually taken to mean those published within approximately the last 
decade;
     View estimates based on the U.S. 2009 National Household 
Travel Survey skeptically, or exclude them from consideration 
completely;
     Emphasize estimates derived from vehicle use and fuel 
economy data spanning multiple years (such as aggregate time-series or 
panel data), while according less weight to those based on a single-
year cross section (such as most household survey data);
     Emphasize estimates of the rebound effect that measure the 
response of vehicle use to variations in fuel efficiency, rather than 
in fuel cost per mile driven or fuel price per gallon;
     Emphasize estimates that rely on identification strategies 
that account for potential endogeneity in fuel economy (as would 
result, for example, if households with high levels of demand for 
travel purchase vehicles with higher fuel economy);
     Emphasize estimates based on measures of vehicle use 
obtained from odometer readings; and
     Emphasize estimates that explicitly control for purchase 
prices of new vehicles in order to account for changes in new vehicle 
prices due to CAFE standards.
    A few commenters illustrated how applying these criteria could 
reduce the large number of published studies of the rebound effect to a 
limited subset that suggested a smaller value than 20 percent.\1750\ 
Using multiple criteria to exclude or de-emphasize studies that did not 
meet all of those applied, these commenters argued that the most 
appropriate value for this analysis was closer to (or possibly even 
below) the 10-percent estimate the agencies used for the previous 
rulemaking.\1751\ However, one commenter noted that applying these 
criteria individually to exclude any estimates not meeting them had 
almost no effect on formal measures of the central tendency (the mean 
and median values) of the remaining estimates.\1752\ This commenter 
suggested that only by applying two or more of these criteria jointly 
and excluding any studies that did not meet all of those applied could 
the universe of research on the rebound effect be reduced to a subset 
supporting a lower value than the 20 percent figure the agencies used 
to analyze the NPRM.
---------------------------------------------------------------------------

    \1750\ See, e.g., Gillingham, Nera-Trinity Responses, NHTSA-
2018-0067-12403, at 16-30.
    \1751\ See supra note 1749.
    \1752\ Alliance of Automobile Manufacturers, Attachment 3, 
NHTSA-2018-0067-12386, at 15-17.
---------------------------------------------------------------------------

    Commenters also identified several additional recent studies that 
were not included in the agencies' review of recent evidence for the 
NPRM, and suggested revised interpretations of the empirical estimates 
reported in two studies that had been included (the agencies also 
clarified a third). Commenters represented these additional studies as 
generally supporting lower values than the agencies' revised 20 percent 
estimate, although this appeared to be a selective interpretation of 
some of the results they reported.\1753\ Other commenters asserted

[[Page 24673]]

that the two most commonly-demonstrated features of the rebound effect 
are that it varies directly with fuel prices and declines in response 
to rising income over time, and argued that the latter suggests that a 
declining value is likely to be more appropriate for analyzing the 
longer-term impacts of this final rule.\1754\
---------------------------------------------------------------------------

    \1753\ For example, some commenters (e.g., Gillingham, Nera-
Trinity Responses, NHTSA-2018-0067-12403, Table 2, at 24) 
represented the recent analysis of vehicle use data from Texas by 
Wenzel and Fujita as reporting a rebound effect of 8-15 percent, 
which appears to be based on those authors' estimates of the 
response of vehicle use to changes over time in fuel prices alone. 
This range appears to ignore those same authors' estimates of the 
sensitivity of vehicle use to variation in fuel costs per mile, 
which provides a more direct measure of the fuel economy rebound 
effect because it incorporates fuel economy as well as fuel prices. 
Those estimates range from 7-40 percent, with most falling in the 
interval from 15-25 percent; see generally, Wenzel and Fujita 
(2018), Table 4-12, at 38.
    \1754\ See particularly Small, NHTSA-2018-0067-7789, at 3.
---------------------------------------------------------------------------

    Some commenters suggested that the rebound effect is asymmetrical, 
meaning that drivers are more responsive to price increases than price 
decreases. These commenters asserted that the asymmetrical nature of 
the rebound effect favors a lower estimate.\1755\ Similarly, other 
commenters suggested that the rebound effect had to be lower than 20 
percent because congestion would limit additional driving.\1756\
---------------------------------------------------------------------------

    \1755\ EDF, Analysis of the Value and Application of the Rebound 
Effect, NHTSA-2017-0069-0574, Comment, 37-38.
    \1756\ For example, the South Coast Air Quality Management 
District argued that, logistically, rebound cannot exist in Southern 
California because ``any rebound effect will only worsen congestion 
in Southern California, such a result cannot be predicted.'' NHTSA-
2018-0067-11813 at 45.
---------------------------------------------------------------------------

(b) Agencies' Response to Comments on the NPRM
    In response to commenters who argued that the agencies' estimate of 
the rebound effect should be reduced, because research that 
incorporates the effects of congestion or allows asymmetrical responses 
to price changes suggests lower values, the agencies note that, for the 
final rule's analysis, those factors would be difficult and perhaps 
even inappropriate to incorporate in their analysis. In the case of 
congestion, the agencies note that their estimate of the rebound 
effect--like research on the rebound effect in general--represents a 
change in aggregate VMT, and has no clear implication about how that 
change in travel is likely to be distributed over times of the day or 
geographic locations.\1757\
---------------------------------------------------------------------------

    \1757\ The agencies' estimate of increased congestion costs 
associated with additional driving due to the rebound effect 
implicitly assumes that increased driving will be distributed 
according to current travel patterns, producing similar proportional 
increases at various hours of the day and geographic locations. Such 
an assumption is made out of necessity to model congestion and 
noise; the agencies acknowledge that the rebound effect is unlikely 
to affect vehicle use in such a uniform fashion.
---------------------------------------------------------------------------

    As for possible asymmetry in the response of vehicle use to changes 
in driving costs, the CAFE model applies a single estimate of the 
rebound effect for all changes in cost-per-mile, and cannot accommodate 
a rebound effect that varies with the magnitude or direction of changes 
in driving costs, which would be necessary to capture asymmetrical or 
non-linear responses to cost changes. The agencies also remind 
commenters that this rule will result in an increase in driving costs, 
for which the research they cite generally suggests a larger value of 
the rebound effect is appropriate. In any case, using a different 
estimate of the rebound effect to analyze impacts of raising and 
lowering standards would not promote consistency or replicability, both 
desirable characteristics of regulatory analysis.
    The agencies decided to include the previously omitted studies 
raised by commenters in their rebound analysis supporting the final 
rule, but do not feel that they suggest a value different from that 
used to analyze the proposal. Adding these studies to the list of 
recent research discussed in the NPRM, deleting one unpublished 
analysis, and revising the entries for selected studies to reflect more 
accurately the values reported by their authors produces a more 
extensive catalog of recent research, which is summarized in Table VI-
175 below.

[[Page 24674]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.356

    As evidenced in Table VI-175, studies continue to have a wide range 
of estimates, but collectively the research looks remarkably similar to 
the historical estimates. The newer studies suggest that a plausible 
range for the rebound effect is 10-50 percent. The central tendency of 
this range appears to be roughly 30 percent.
    In response to comments proposing the application of specific 
criteria to eliminate or reduce the consideration accorded to studies 
without certain features thought to increase the relevance of their 
findings, the agencies note that measuring the rebound effect is both 
conceptually and technically challenging, and that analysts have used 
many different approaches in an attempt to surmount these challenges. 
The agencies' view is that each of the studies included in its previous 
survey and in Table VI-175 above provides some useful evidence on the 
likely value of the rebound effect, and while all have some conceptual 
or theoretical weaknesses, each nevertheless provides some useful 
insights into the appropriate magnitude of the rebound effect for the 
current analysis.
    As a general approach to estimating parameters that are uncertain, 
the agencies prefer to rely on the totality of empirical evidence, 
rather than restricting the available evidence by categorically 
excluding or according less weight to that do not meet selection 
criteria that may not be widely agreed upon. From this perspective, 
analyses that rely on different measurement approaches, data sources, 
and estimation procedures all have the potential to provide valuable 
information for choosing the most representative value. The agencies 
also view sound measurement strategies and careful empirical analysis 
using reliable data as equally important features when compared to a 
study's vintage or geographic scope. Examining the widest possible 
range of research also enables useful comparisons and ``cross-checks'' 
on the estimates that individual studies report.
    Notwithstanding this more inclusive perspective, the agencies 
endorse certain of the characteristics preferred by commenters, 
although the agencies view them as indicators of a strong study, rather 
than a bright-line test of whether to accord it any weight rather than 
discarding it from consideration. Specifically, the agencies agree with 
many commenters that both the extended time span encompassed by their 
analysis of the impacts of CAFE and CO2 standards and the 
long expected lifetimes of vehicles subject to this final rule means 
that estimates of the long-run rebound effect are most relevant for 
purposes of the final rule

[[Page 24675]]

analysis.\1758\ The agencies also agree with commenters that estimates 
based upon more recent data are generally preferable, but nevertheless 
note that older studies that combine careful analysis with unusually 
reliable or novel data can offer evidence that remains useful.\1759\ 
The agencies also concur with some commenters' argument that estimates 
of the rebound effect that are derived from the relationship of vehicle 
use to fuel efficiency, rather than to fuel cost per mile or gasoline 
prices, are likely to provide more direct measures of the fuel economy 
rebound effect itself, which is the desired parameter for the purposes 
of this analysis. Finally, the agencies generally view identification 
strategies and econometric methods that account or control for 
potential endogeneity in fuel economy as likely to provide more 
reliable estimates.
---------------------------------------------------------------------------

    \1758\ Most of the vehicles affected by today's standards will 
remain on the roads for at least a decade, with a significant 
fraction surviving considerably longer. As such, long-run estimates 
are more likely to reflect the lifetime mileage accumulation of the 
new fleet than either short-run or medium-run estimates. 
Furthermore, a long-run rebound estimate better reflects the 
cumulative impact of successive CAFE and CO2 standards 
such as those adopted by the agencies beginning as early as 2010.
    \1759\ One example is the study by Greene et al. (1999), which 
used advanced econometric analysis of unusually detailed and 
reliable data on household demographic and economic characteristics, 
household members' use of individual vehicles, and fuel purchases to 
estimate the response of households' use of individual vehicles to 
their actual on-road fuel economy, and its implications for total 
household driving.
---------------------------------------------------------------------------

    In contrast, the agencies view other criteria proposed by 
commenters as unnecessarily restrictive, particularly when they are 
used to disqualify otherwise informative research from consideration. 
For instance, categorically excluding from consideration non-U.S. 
studies--which the agencies agree should be treated cautiously--seems 
likely to exclude useful evidence, particularly recognizing some of 
those studies' access to unusually reliable data on vehicle use and 
fuel economy and use of sophisticated econometric analysis. In 
addition, many foreign studies have been conducted in nations with 
income levels comparable to the U.S., and in some cases levels of auto 
ownership that are beginning to approach U.S. levels. Furthermore, 
driving habits throughout the U.S. are not homogenous. In fact, some 
regions in the U.S. may exhibit driving habits that more closely 
resemble those in some foreign nations than driving patterns in other 
regions of the U.S.\1760\
---------------------------------------------------------------------------

    \1760\ For example, drivers in Manhattan, Kansas likely respond 
to changes in fuel prices and fuel economy differently than drivers 
in Manhattan, New York.
---------------------------------------------------------------------------

    In response to some commenters' recommendation that the agencies 
more heavily weigh studies using data spanning multiple years than 
those relying on data for a single year, the agencies note that 
household surveys, the most common form of data for a single year, 
provide cross-sectional variation in vehicle use and other 
characteristics that is helpful for identifying the desired long-run 
measure of the rebound effect. Household surveys are also an important 
source of information that enable analysts to measure the response of 
individual vehicles' use to variation in their fuel economy, while also 
controlling adequately for household characteristics that affect travel 
patterns and vehicle use. Household survey data can also enable 
analysts to identify the vehicle substitution patterns within multiple-
vehicle households that are increasingly responsible for producing the 
rebound effect, while even modest-scale household surveys include many 
more observations than are typically available in aggregate time-series 
or panel data.
    These strengths of course need to be balanced against the potential 
drawbacks of relying on a one-time snapshot of households' behavior 
during a single time period. Surveys also frequently rely on owner-
reported estimates of vehicle use and usually require analysts to 
impute vehicles' fuel economy ratings from limited and sometimes 
incomplete information on the specific vehicle models and vintages that 
households report owning. One result is that estimates of the rebound 
effect derived from household survey data may be based on inaccurate 
estimates of vehicles' use and fuel economy. Assuming the errors in 
measuring these variables are random, the errors would increase the 
uncertainty surrounding the estimates of the rebound effect, but would 
not bias the estimate.
    In contrast, studies using nationwide aggregate or average measures 
of vehicle use and fuel economy or fuel cost rarely provide adequate 
independent variation to support reliable estimates of the response of 
vehicle use to variation in fuel economy, even where extended time 
series are available, while State-level measures of these variables are 
subject to potentially extreme measurement error that can compromise 
estimates of these relationships.\1761\ Moreover, controlling for the 
many other demographic and economic factors likely to affect vehicle 
use using national or even State-level aggregate data presents 
difficult challenges.
---------------------------------------------------------------------------

    \1761\ For example, State-level estimates of travel by 
individual vehicle classes such as cars and light-duty trucks often 
exhibit implausible year-to-year variability due to the measurement 
procedures states employ and the difficulty of distinguishing among 
different types of vehicles. At the same time, the potential 
geographic ``mismatch'' between State-level vehicle use and fuel 
sales complicates any effort to measure fuel efficiency or fuel 
costs at the State level.
---------------------------------------------------------------------------

    Finally, the agencies note that no single selection criterion 
proposed by commenters noticeably reduces the central tendency 
displayed by the universe of estimates of the rebound effect, and 
multiple criteria must be applied simultaneously to restrict the 
universe to a subset of studies that points toward a significantly 
lower value than the 20 percent estimate the agencies used to analyze 
the proposal. Applying multiple criteria drastically reduces the number 
of studies that remain available to guide the agencies, while at the 
same time discarding potentially valuable information provided by 
research those criteria exclude from consideration.\1762\ Doing so 
would thereby necessarily reduce the confidence that the agencies can 
have in the resulting estimate.
---------------------------------------------------------------------------

    \1762\ As an illustration, excluding non-U.S. studies reduces 
the number of recent analyses surveyed in the proposal from 15 to 8, 
while eliminating those that rely on the 2009 National Household 
Travel Survey (NHTS) discards another 5, leaving only 3.
---------------------------------------------------------------------------

    Regarding some commenters' assertion that the rebound effect is 
known to decline in response to rising income, and that this 
observation warrants using a lower value for long-term future 
evaluation of the standards' effects, the agencies note that some 
evidence based on household and vehicle use surveys suggests that the 
rebound effect increases with the level of household vehicle ownership, 
which is itself highly correlated with income. Together with forecasts 
of limited future growth in most measures of U.S. household income, 
this finding casts some doubt on whether the rebound effect is likely 
to decline over the time period spanned by the agencies' 
analysis.\1763\
---------------------------------------------------------------------------

    \1763\ For example, the widely cited IHS Markit Long-Term 
Macroeconomic Outlook for Spring 2019 projects that per Capita 
disposable personal income in the U.S. will grow at 1.6 percent 
annually over the next 30 years; see Federal Highway Administration, 
Forecasts of Vehicle Miles Traveled (VMT): Spring 2019, Table 2, 
available at https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_forecast_sum.cfm.
---------------------------------------------------------------------------

    The agencies also note that one of the studies cited in Table VI-
175 above (DeBorger et al., 2016) finds that the decline in the fuel 
economy rebound effect with income reported in the

[[Page 24676]]

earlier analysis by Small and Van Dender (2007)--on which the agencies 
relied in reducing their original estimate of the rebound effect to 10 
percent--results entirely from a reduction in drivers' sensitivity to 
fuel prices as their incomes rise, rather than from any effect of 
rising income on the sensitivity of vehicle use to fuel economy.\1764\ 
This latter measure--which DeBorger et al. find is quite small and has 
not changed significantly as incomes have risen over time--is the most 
direct measure of the fuel economy rebound effect, so their analysis 
calls into question its widely-assumed sensitivity to income.
---------------------------------------------------------------------------

    \1764\ DeBorger, B., Mulalic, I., and Rouwendal, J., ``Measuring 
the rebound effect with micro data: A first difference approach.'' 
Journal of Environmental Economics and Management, 79 (2016), at 1-
17.
---------------------------------------------------------------------------

    Finally, because there is not a clear consensus around a single 
rebound estimate within the literature, the agencies believe it is 
important to benchmark their analysis with other large scale surveys of 
the literature published by neutral observers. In one early survey, 
Greening, Greene, and Difiglio (2000) reviewed studies that estimated 
the rebound effect for light-duty vehicles in the U.S., concluding that 
those relying on aggregate time-series data found it was likely to 
range from 10-30 percent, while those using cross-sectional analysis of 
household vehicle use suggested a larger rebound effect, in the range 
of 25-50 percent.\1765\ Sorrell et al. (2009) found that the magnitude 
of the rebound effect for personal automobile travel is likely to fall 
in the 10-30 percent range, with some evidence suggesting that the 
lower end of that range might be most appropriate.\1766\
---------------------------------------------------------------------------

    \1765\ Greening, L.A., Greene, D.L. and Difiglio, C., ``Energy 
efficiency and consumption--the rebound effect--a survey.'' Energy 
Policy, Vol. 28 (2000), at 389-401.
    \1766\ Sorrell, Steve, John Dimitropoulos, and Matt Sommerville, 
``Empirical Estimates of the Direct Rebound Effect: A Review,'' 
Energy Policy 37(2009), at 1356-71.
---------------------------------------------------------------------------

    Most recently, a meta-analysis of 74 published studies of the 
rebound effect conducted by Dimitropoulos et al. (2018) estimated that 
the long-run rebound effect ranges from 22-29 percent when measured by 
the response of vehicle use to variation in fuel efficiency (the 
authors' preferred measure), from 21-41 percent when it is measured 
using the variation fuel cost per unit distance, and from 25-39 percent 
using fuel price per gallon.\1767\ The authors concluded that ``the 
magnitude of the rebound effect in road transport can be considered to 
be, on average, in the area of 20%,'' but noted that the long-run 
estimate was about 32 percent.\1768\ A subsequent published study by 
these same authors (Dimitropoulos et al. (2018)) concludes that the 
most likely estimate of the long-run rebound effect is in the range of 
26-29 percent, but could range from as low as 15 percent to as high as 
49 percent at income levels, development densities, and fuel prices 
that are currently representative of the U.S.\1769\
---------------------------------------------------------------------------

    \1767\ Dimitropoulos, Alexandros, Walid Oueslati, and Christina 
Sintek, ``The rebound effect in road transport: a meta-analysis of 
empirical studies,'' Paris, OECD Environment Working Papers, No. 
113; see esat Table 5, at 25 (and accompanying discussion).
    \1768\ Id. at 28.
    \1769\ Dimitropoulos, Alexandros, Walid Oueslati, and Christina 
Sintek, ``The Rebound Effect in Road Transport: A Meta-Analysis of 
Empirical Studies,'' Energy Economics 75 (2018), at 163-79; see esat 
Table 4, at 170, Table 5, at 172 (and accompanying discussion), and 
Appendix B, Table B.V., at 177.
---------------------------------------------------------------------------

(c) Selecting a Value of the Rebound Effect for Evaluating the Impacts 
of This Rule
    After reviewing the evidence on the rebound effect previously 
summarized in the NPRM, comments the agencies received, other recent 
studies of the rebound effect that were not summarized in the NPRM but 
suggested by commenters, and published surveys of literature, a 
reasonable case can be made to support values of the rebound effect at 
least as high as 30 percent. The totality of evidence, without 
categorically excluding studies on grounds that they fail to meet 
certain criteria, and evaluating individual studies based on their 
particular strengths, suggests that a plausible range for the rebound 
effect is 10-50 percent. The central tendency of this range appears to 
be at or slightly above its midpoint, which is 30 percent. Considering 
only those studies that the agencies believe are derived from unusually 
reliable data, employ identification strategies that are likely to 
prove effective at isolating the rebound effect, and apply rigorous 
estimation methods suggests a range of approximately 10-45 percent, 
with most of their estimates falling in the 15-30 percent range.\1770\
---------------------------------------------------------------------------

    \1770\ As indicated previously, these are the selection criteria 
proposed by commenters with which the agencies concur. In 
chronological order, the studies the agencies feel best meet those 
criteria include Greene et al. (1997), Small and Van Dender (2007) 
and subsequent updates by Hymel, Small, and Van Dender (2010, 2015), 
Linn (2016), Anjovic and Haas (2012), Gillingham (2014), and 
DeBorger et al. (2016). Other studies the agencies believe warrant 
serious consideration because they offer some or most of these same 
advantages include those by Liu et al. (2014), Knittel and Sandler 
(2018), and Wenzel and Fujita (2018).
---------------------------------------------------------------------------

    At the same time, the agencies conclude that a reasonable case can 
also be made to support values of the rebound effect falling in the 5-
15 percent range. This argument relies on using the criteria proposed 
by commenters to restrict the studies considered to include recently 
published analyses using U.S. data, and to accord the most weight to 
research that relies on measures of vehicle use derived from odometer 
readings, controls for the potential endogeneity of fuel economy, and 
estimates the response of vehicle use to variation in fuel economy 
itself, rather than to fuel cost per distance driven or fuel prices. 
This approach suggests that the rebound effect is likely in the range 
from 5-15 percent, and is more likely to lie toward the lower end of 
that range. The agencies note that estimates of very low or no rebound 
effect cited by some commenters are either misinterpretations of the 
findings reported by their authors, or do not represent measures of the 
fuel economy rebound effect.\1771\
---------------------------------------------------------------------------

    \1771\ For example, some commenters misinterpret Greene's (2012) 
inability to identify a statistically significant estimate of the 
response of vehicle use to variation in fuel economy as evidence 
that its true value is zero. Similarly, some commenters misinterpret 
the result reported by West et al. (2017) that buyers of more fuel-
efficient vehicles did not increase their driving as evidence that 
fuel economy itself has no effect on vehicle use, when--as the 
study's authors and some commenters acknowledge--it reveals instead 
that buyers regarded those vehicles as providing inferior 
transportation service and drove them less as a consequence. Because 
the agencies repeatedly insist that vehicle attributes other than 
fuel economy will not change as a consequence of this rule, those 
authors' finding is of limited or no relevance to the analysis 
supporting this rule.
---------------------------------------------------------------------------

    Finally, the agencies note that surveys of evidence on the rebound 
effect have consistently found that the most appropriate estimate falls 
in the range of 10-40 percent. These findings have remained 
surprisingly consistent over time, despite a rapidly expanding universe 
of empirical evidence that includes estimates drawn from more diverse 
settings, and reflects continuing improvements in the data they rely 
upon, an expanding range of strategies for identifying the rebound 
effect and distinguishing it from other influences on vehicle use, and 
advances in the econometric procedures analysts use to estimate its 
magnitude.
    For the aforementioned reasons, the agencies have elected to retain 
the 20 percent rebound effect used to analyze the effects of the NPRM 
on vehicle use and fuel consumption for analyzing the comparable 
effects of this final rule. As explained above and in the NPRM, older 
research suggests a rebound of 20 to 25 percent. The new research in 
Table VI-175 supports a similar--or even larger--range. Extensive 
survey studies support

[[Page 24677]]

a rebound at or above 20 percent. As such, the agencies feel 20 percent 
is a reasonable--and probably even conservative--estimate of the 
totality of the evidence. While a lower estimate may be reasonable 
under certain circumstances, the agencies are uncomfortable making the 
requisite assumptions regarding which specific criteria should be used 
to identify relevant studies and relying on a subset of the literature 
for the central analysis. However, recognizing the uncertainty 
surrounding the rebound value, the agencies also examine the 
sensitivity of those estimated impacts to values of the rebound ranging 
from 10 percent to 30 percent, both in isolation and in conjunction 
with plausible variation in other key parameters.
(5) Vehicle Miles Traveled (VMT)
    VMT directly influences many of the various effects of fuel economy 
and CO2 standards that decision-makers consider in 
determining what levels of standards to set. For example, fuel savings 
is a function of a vehicle's efficiency, miles driven, and fuel price. 
Similarly, factors like criteria pollutant emissions and fatalities are 
direct functions of VMT. In the CAFE model, VMT is the product of 
average usage per vehicle in the fleet and fleet composition, which is 
itself a function of new vehicle sales and vehicle retirement 
decisions, otherwise known as scrappage. These three components--
average vehicle usage, new vehicle sales, and older vehicle scrappage--
jointly determine total VMT projections for each alternative.
    As the following discussion explains, today's VMT analysis provides 
aggregate results comparable to other well-regarded VMT estimates. 
However, because the agencies' analysis looks at the incremental costs 
and benefits across alternatives (see Section VII), it is more 
important that the analysis capture the variation of VMT across 
alternatives than accurately to predict total VMT within a scenario. As 
such, the agencies note that today's VMT estimates are logical, 
consistent, and precise across alternatives. Furthermore, as will be 
described in further detail below, while the agencies, in response to 
comments, have decided to modify their approach to calculating VMT and 
to use different VMT estimates than those used in the NPRM, the general 
trends between alternatives are comparable.
    Commenters addressed a number of topics related to the total amount 
of estimated VMT, the incremental differences in estimated VMT between 
regulatory alternatives, and per-vehicle VMT estimates in the NPRM 
analysis. In general, commenters felt that the NPRM's VMT numbers were 
inaccurate and should not be relied on for the analysis.\1772\ Some 
commenters were more specific and argued that the total amount of 
estimated VMT projected in the NPRM started at too low a level, and 
increased too much over the years simulated. Similarly, some commenters 
argued that the agencies' estimates were too different from other 
recognized estimates and suggested that the agencies benchmark VMT 
projections to other sources to ensure both a consistent starting point 
and comparable VMT throughout the calendar years analyzed.
---------------------------------------------------------------------------

    \1772\ See, e.g., Securing America's Energy Future, NHTSA-2018-
0067-11981 at 37-38.
---------------------------------------------------------------------------

    A few commenters objected to the underlying mileage accumulation 
schedules, which form the basis for per-vehicle VMT estimates in CAFE 
Model simulations. Such commenters speculated that revisions to these 
schedules undertaken in 2016 might be the reason for discrepancies in 
total VMT. Other commenters were less concerned about how VMT was 
computed within each scenario but were apprehensive about differences 
in VMT estimates across regulatory alternatives. For instance, Honda 
argued that, ``[a]ssuming all other parameters are held constant--and 
excluding the rebound effect--it is not obvious why one scenario should 
have different total VMT than another.'' \1773\ While commenters 
generally provided few specific recommendations about the level to 
which VMT estimates should be constrained among alternatives, several 
commenters argued that VMT projections would benefit from consideration 
of travel demand modeling.
---------------------------------------------------------------------------

    \1773\ Honda, NHTSA-2018-0067-11818, at 17.
---------------------------------------------------------------------------

    Additionally, some commenters (RFF, IPI, NRDC) argued that a 
superior, and perhaps even necessary, approach would be to incorporate 
a model that considers jointly the decision to buy, use, and retire 
vehicles at the household level. As RFF posited ``a household makes 
decisions about its vehicle ownership and use jointly: people don't buy 
new vehicles or get rid of existing ones without considering how these 
actions will affect the use of their vehicles.'' \1774\ IPI further 
argued that ``[i]n sum, VMT is influenced by vehicle choice and vehicle 
choice is influenced by VMT. And a `unified model of vehicle choice and 
usage' is necessary.'' \1775\ While the agencies agree that a joint 
household consumer choice model--if one could be developed adequately 
and reliably to capture the myriad circumstances under which families 
and individuals make decisions relating to vehicle purchase, use and 
disposal--would reflect decisions that are made at the household level, 
the agencies do not agree that it is necessary, or necessarily 
appropriate, to model the national program at that scale in order to 
produce meaningful results that can be used to inform policy decisions. 
The most useful information for policymakers relates to national 
impacts of potential policy choices. No other element of this analysis 
occurs at the household level, and the error associated with allocating 
specific vehicles to specific households over the course of three 
decades would easily dwarf any error associated with the estimation of 
these effects in aggregate. The agencies have attempted to incorporate 
estimates of changes to the new and used vehicle markets at the highest 
practical levels of aggregation, and worked to ensure that these 
effects produce fleetwide VMT estimates that are consistent with the 
best, current projections given our economic assumptions. While future 
work will always continue to explore approaches to improve the realism 
of CAFE/CO2 simulation, there are important differences 
between small-scale econometric studies and the kind of flexibility 
that is required to assess the impacts of a broad range of regulatory 
alternatives over multiple decades. The agencies have read and 
evaluated the comments on the NPRM, incorporating many suggestions that 
improve the fidelity of this analysis--taking particular care to be 
conservative with the analysis. The modifications the agencies have 
made in response to these comments are described below (and in the 
RIA).
---------------------------------------------------------------------------

    \1774\ RFF, NHTSA-2018-0067-11789, at 5.
    \1775\ IPI, Appendix, NHTSA-2018-0067-12213, at 80 (internal 
citation omitted).
---------------------------------------------------------------------------

    The agencies carefully assessed all comments. To address them, the 
agencies have revised their calculation of estimated VMT in two, 
significant respects. First, in response to comments regarding the 
mileage accumulation schedules, the agencies have revised the schedules 
using panel data. Second, to deal with commenters' concerns with the 
fluctuation of estimated ``non-rebound'' VMT across regulatory 
alternatives, the agencies have created a method that constrains ``non-
rebound'' VMT across regulatory alternatives. The agencies believe 
these two changes collectively resolve the substantive issues raised by 
commenters. The total VMT for the final rulemaking (FRM) analysis now 
aligns with estimates of the Federal Highway Administration

[[Page 24678]]

(FHWA) and the only differences in VMT between alternatives is 
attributable to changes in the fleet's fuel economy. The following 
sections discuss these changes in detail.
(a) Mileage Accumulation Schedule
    To account properly for the average value of consumer and societal 
costs and benefits associated with vehicle usage under various CAFE and 
CO2 alternatives, it is necessary to estimate the portion of 
these costs and benefits that will occur each calendar year for each 
model year cohort. Doing so requires some estimate of how many miles 
the average vehicle of each body type is expected to drive at each age. 
The agencies call these ``mileage accumulation schedules.'' For this 
final rule, the agencies are modifying the mileage accumulation 
schedules, largely in response to comments.
(i) Data
    As mentioned in previous sections, NHTSA purchased a data set 
containing 70 million vehicle odometer readings from Polk in part to 
create the vehicle mileage accumulation schedules used in the NPRM. In 
the proposal, the agencies explained that Polk data was newer and 
believed to be qualitatively superior to the 2001 and 2009 National 
Household Travel Survey (NHTS) data used in prior rules.\1776\ 
Consistent with previous analyses,\1777\ the agencies used a cross-
sectional sample of the Polk data for the NPRM. Cross-sectional data is 
like a ``snapshot'' in time. Rather than tracking vehicles over a 
period, the sample contained a single odometer reading from each 
vehicle sampled. In other words, the sample contained observations of 
the total lifetime accumulation of miles (represented by its odometer 
reading) through CY2015 of all MYs still present on the road. The 
cross-sectional sample was limited in the number of vintages included 
in the sample. While the sample was suitable to capture the heaviest 
usage ages (age zero to 15 years), it contained no observations for 
vehicles older than 16 years. This required the agencies to rely on 
mileage accumulation schedules developed from other data sources to 
produce annual VMT rates for older vehicles. Furthermore, in order to 
develop a schedule of mileage accumulation by age, it was necessary to 
assume that each vehicle traveled the same number of miles each year to 
reach its odometer reading, e.g. if a MY 2007 vehicle had an odometer 
reading of 88,000 in CY2015, the analysis assumed the vehicle drove 
11,000 miles each year from CY2007 to CY2015.
---------------------------------------------------------------------------

    \1776\ See, e.g., 83 FR at 43089-90 (Aug. 24, 2018).
    \1777\ Previous rules were based on odometer data from the 2001 
National Household Travel Survey (NHTS). S. Lu, ``Vehicle 
Survivability and Travel Mileage Schedules,'' Report Number: DOT HS 
809 952 (January 2006).
---------------------------------------------------------------------------

    The agencies acknowledged that this approach missed some of the 
nuances of car ownership.\1778\ For example, vehicles are commonly part 
of multi-vehicle household fleets and their usage changes over time as 
households buy new vehicles and replace older ones. Similarly, most 
vehicles belong to multiple owners over the course of their useful 
lives, each of whom may have different patterns of usage. The most 
significant limitation of using cross-sectional data is the presence of 
an attrition bias. As a cohort ages, vehicles that have been used more 
heavily are more likely to be retired at each age than vehicles that 
are driven less. As the most heavily-driven vehicles drop out of the 
fleet, the remaining vehicles, which likely have been driven less at 
each age throughout their lives, will have lower odometer readings. 
Making the common, but necessary assumption that each vehicle is driven 
uniformly at each age results in lower miles-per-age estimates because 
of this attrition bias. In the schedules used for the NPRM, the effect 
of this bias occurred during the ages where each model year cohort 
typically scraps at the highest rates--9 to 15 years. These limitations 
led to lower estimates, which led commenters such as EDF to state 
``[g]iven that the Volpe Model VMT falls far short of confident 
measurements of gasoline consumption, these mileage accumulation 
schedules need to be increased.'' \1779\ The agencies note that many of 
these data limitations were present in previous CAFE and CO2 
analyses.\1780\
---------------------------------------------------------------------------

    \1778\ See 83 FR at 43092 (Aug. 24, 2018).
    \1779\ EDF, Appendix B (Rykowski comments), NHTSA-2018-0067-
12108, at 46.
    \1780\ See, e.g., NHTSA Final Regulatory Impact Analysis: 
Corporate Average Fuel Economy for MY 2012-MY 2016 Passenger Cars 
and Light Trucks, NHTSA-2010-0131, at 372-79.
---------------------------------------------------------------------------

    Several commenters noted the agencies' reliance on cross-sectional 
data, and urged the use of panel data instead to develop mileage 
accumulation schedules. For example, API argued that cross-sectional 
data cannot accurately capture mileage accrual and suggested ``the 
agencies re-consider the use of the [Polk] data for developing revised 
mileage accumulation schedules unless the data can capture mileage 
accumulation rates versus age on an individual-vehicle basis.'' \1781\ 
The NPRM discussed the possible use of panel data in the future and the 
benefits that doing so could provide.\1782\
---------------------------------------------------------------------------

    \1781\ API, EPA-HQ-OAR-2018-0283-4548, at 10.
    \1782\ See 83 FR at 43092 (Aug. 24, 2018).
---------------------------------------------------------------------------

    In response to these comments, the agencies created new mileage 
accumulation schedules based on panel data for this final rule. Unlike 
cross-sectional data, panel data includes a temporal element, which 
resolves the limitations imposed by cross-sectional data. The data 
source used for the final rule contains sequential readings of 
individual vehicles over time, and the vehicles are tracked at the VIN 
level. Polk accumulates readings about individual vehicles through 
state inspection programs, title changes, and maintenance events, among 
other sources. The Polk data includes observations of a specific 
vehicle's odometer readings over the course of many years, capturing 
the accumulated lifetime mileage at multiple ages. By using the 
observation date and accumulated miles (represented by the odometer 
reading), the agencies can compute the rate of driving (miles per year, 
or month) between observations for each vehicle. This is a superior 
method to assuming that the rate of accumulation, over all ages, is 
simply the ratio of odometer to age, as commenters noted. In 
particular, calculating the rates of mileage accumulation using 
successive observations of the same vehicle explicitly resolves the 
attrition bias and matches the approach to estimating driving rates 
with panel data in other studies.\1783\
---------------------------------------------------------------------------

    \1783\ See, e.g., Kenneth Gillingham, Alan Jenn, and In[ecirc]s 
M.L. Azevedo (2015), ``Heterogeneity in the Response to Gasoline 
Prices: Evidence from Pennsylvania and Implications for the Rebound 
Effect, Energy Economics,'' Volume 52, Supplement 1, 2015, Pages 
S41-S52, ISSN 0140-9883, available at https://doi.org/10.1016/j.eneco.2015.08.011.
---------------------------------------------------------------------------

    The panel dataset has another advantage over other sources: Because 
it tracks individual vehicles over time, the agencies have more precise 
information about each vehicle's useful age. In previous analyses, the 
agencies were forced to assume that ``age'' was simply equal to the 
calendar year minus the model year in which the vehicle was produced. 
For example, a MY2010 vehicle was assumed to be five years old in 2015. 
This created, as API stated, a ``discontinuity in the values between 
year 1 and year 2'' within the schedules.\1784\ It is common for 
vehicles produced in a given model year to be sold and registered over 
the course of multiple calendar years. Thus, a MY2010 vehicle assumed 
to be five years old in 2015, could have been

[[Page 24679]]

registered for the first time in CY2012 and might have a real driving 
age of three years, rather than five, simply because it sat on a 
dealership lot for a couple of years before being purchased. The Polk 
data allows us to identify the first registration date of each vehicle 
in the sample and compute its true driving age at each point in time. 
This not only improves the precision of the mileage accumulation rate 
in the first year, but in subsequent years as well. The odometer data 
used in the NPRM had another limitation: Odometer readings were grouped 
into cohorts by nameplate, for which only distributional information 
was available. It was necessary to use the mean odometer reading for 
each cohort at each age, but in cases where the distribution was 
skewed, the mean could be misleading. Making the same assumption about 
registration date, as each cohort contained information about the 
average registration date, further compounded the potential for 
distortion.
---------------------------------------------------------------------------

    \1784\ API, EPA-HQ-OAR-2018-0283-5458, at 9-10.
---------------------------------------------------------------------------

    To the extent that commenters objected to the NPRM's use of Polk 
data on the basis of it being proprietary, the agencies note that using 
proprietary data is common in rulemakings, and, specifically, Polk data 
has been used for CAFE and CO2 analyses on multiple 
occasions previously. For the 2016 final medium- and heavy-duty rule 
and Draft TAR, the agencies used Polk odometer data to develop the 
vehicle mileage accumulation schedules.\1785\ Further, the specific 
data set was cited and is available for acquisition through Polk.
---------------------------------------------------------------------------

    \1785\ See, e.g., 81 FR 73478, 73746 (Oct. 25, 2016); see also 
81 FR 49217 (Jul. 27, 2016).
---------------------------------------------------------------------------

    Recently, the 2017 National Household Travel Survey has become 
available as a possible data source to develop mileage accumulation 
schedules. While attractive from the standpoint of transparency, it 
suffers from the same flaws as data sources used to develop previous 
schedules. In particular, it represents a cross section of odometer 
readings at a single point in time, requiring the assumption that the 
rate of usage is simply reported odometer divided by vehicle/age, or an 
extrapolation of respondents' daily travel behavior into representative 
annual schedules, which commenters suggested was a poor assumption. 
Additionally, all of the odometers in the newest NHTS are self-
reported, leading to questionable reliability of the individual data 
points (and notably round numbers in many cases). Finally, the NHTS is 
intended to be a representative sample of households, but not a 
representative sample of vehicles. Research has found that creating a 
representative sample of households can represent a significant 
challenge, as past iterations of the NHTS have systematically 
oversampled high income households. The nature of the sample also 
explicitly excludes vehicles used for commercial purposes, which 
nonetheless compose a meaningful portion of the new vehicle market, 
accumulate miles of travel, and consume fuel. The data set on which the 
mileage accumulation schedule used for this final rule is based 
contains at least two readings (and frequently several) for over 70 
percent of the registered light duty vehicle population in 2016.
(ii) Methodology
    The data used to construct the schedules initially included between 
two and fifty odometer readings from each of over 251 million unique 
vehicles. While most of the readings had plausible reading dates, 
odometer counts, and implied usage rates, some of the readings appeared 
unrealistic and received additional scrutiny. The agencies used a set 
of criteria to identify and remove readings that were likely record 
errors. For example, odometer readings predating the commercial release 
of the vehicle, showing negative VMT accumulation over time, or taken 
too closely together to provide meaningful insight into annual vehicle 
usage were removed from the analysis.\1786\ Such sanitization of real 
datasets is typically necessary, and each step in the process was 
recorded and described in conformity with standard econometric 
practice.\1787\
---------------------------------------------------------------------------

    \1786\ Refer to Section VI.D.1.(5).(b) of the FRIA for a full 
accounting of the process used to clean the Polk odometer data.
    \1787\ See, e.g., Osborne, Jason W., Best Practices in Data 
Cleaning, SAGE Publications, Inc., January 2012.
---------------------------------------------------------------------------

    Similar to the NPRM, the remaining readings were sorted into five 
categories: Cars, SUV's/vans, pickups, MDHD pickups/vans, and chassis. 
The car, SUVs/vans and pickup categories match the definitions used to 
build the VMT schedules used in the NPRM, as well as those used to 
build the scrappage model. Table VI-176 shows the number of VINs, 
reading pairs, and average readings per VIN by body style.
[GRAPHIC] [TIFF OMITTED] TR30AP20.357

    *Not used in this final rule analysis, in part in response to 
comments.
    Once the dataset was cleaned, the agencies created a sample of one 
million reading pairs, where each pair represented an initial odometer/
date reading and a subsequent odometer/date reading from the same 
vehicle. Analysis

[[Page 24680]]

of the entire dataset was too computationally demanding and 
statistically unnecessary. Two conditions were created for sampling. 
The first controlled for Polk's censoring in the odometer readings 
recorded in the dataset (described below), and the second ensured the 
usage data was not biased by survival and that it represented usage 
rates over a relatively short period of time compatible with the 
beginning of the FRM analysis. Further analysis suggests that shorter 
periods between readings is still correlated with higher usage rates so 
that further filtering of the data sample was considered in the 
regression analysis. Once these filters were applied, the agencies 
considered several polynomial fits to the average odometer readings. 
These fits inform the final usage rates by age and body style used in 
this FRM analysis. The details are further described below.
    One element of the usage data (mentioned above as the first 
condition control) required the agencies to filter the dataset. The 
odometer readings recorded are censored at 250k miles.\1788\ For this 
reason, the agencies exclude readings recorded exactly as 250k miles. 
The censoring could bias estimates of usage rates if odometer readings 
and future usage rates are correlated, which they likely are. While the 
agencies hope to reconcile this limitation of the dataset in future 
work, the benefits of observing actual usage data through 30 years 
(rather than average odometer readings by model through 15 years) far 
outweigh the limitation. Still, the agencies filtered out these 
censored data points, since the actual odometer readings for such 
vehicles are likely higher than reported.
---------------------------------------------------------------------------

    \1788\ Polk codes any vehicle whose odometer exceeds 250K miles 
as 250K miles exactly, regardless of the actual odometer reading.
---------------------------------------------------------------------------

    The Polk dataset is conditional on survival so that it represents 
the usage of vehicles on the road at the time of the sample (the end of 
the first quarter of 2017). In this way, it captures the actual 
observed usage rates of vehicles surviving to their current age in the 
dataset. An issue with this is that all readings of a vehicle are 
included in the sample. If usage rates from earlier ages and survival 
are correlated, which they likely are, then including the readings for 
a 30-year-old vehicle when it was 10 years old will bias the estimated 
usage rates of 10-year-old vehicles downward because vehicles that 
survive to advanced ages tend to be used less than vehicles that are 
retired at earlier ages for the same model year. As noted above, the 
NHTS data used in the NPRM suffered from the same problem. To mitigate 
this issue, the agencies applied a second filter when sampling the data 
set: The agencies only included readings where the reading date of the 
second reading in the pair is January 2015 or later. This reduces the 
potential bias from the joint probability of usage and survival to only 
those vehicles scrapped between January 2015 and the first quarter of 
2017. This balances losing information for older, less represented ages 
by excluding too much data on these vehicles and severely biasing the 
estimates of usage by age.
    For estimates within the CAFE model the average usage is the 
relevant measure. Table VI-177 shows the average usage rates for cars 
by age as well as linear, quadratic, and cubic polynomial fits on these 
points.\1789\ The average usage rates follow a relatively smooth 
pattern, but appear to decline at an accelerating rate for the oldest 
ages. The linear equation captures this trend for older vehicles, but 
underestimates early ages. The quadratic fit shows a diminishing 
decrease in the usage of older vehicles which may overestimate their 
use. The cubic fit captures the early age usage trends and the 
accelerating decrease in the usage of older ages. For this reason, the 
agencies used the cubic curve as the basis for the new VMT schedules by 
age.
---------------------------------------------------------------------------

    \1789\ In general, the objective of a polynomial regression is 
to capture the nonlinear relationship between two variables. While 
the fit produces a nonlinear curve, it is linear in the 
coefficients. Choosing the lowest degree of the polynomial function 
that captures the inflection points in the data preserved degrees of 
freedom and ensures that applying the polynomial function to 
observations outside the range of data (as done here for ages beyond 
30) is well behaved.
[GRAPHIC] [TIFF OMITTED] TR30AP20.358


[[Page 24681]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.359

    Table VI-178 shows the observed and predicted average usage rates 
by age for SUVs/vans. All the polynomial fits predict the observed 
average usage rates reasonably well. However, the linear fit under 
predicts the usage of the oldest vehicles, and the cubic fit predicts 
higher usage rates for vehicle ages beyond age 30. The quadratic fit 
predicts reasonable usage rates for all observed and out-of-sample ages 
through age 40. For this reason, the quadratic fit was used as the 
basis for the SUV mileage schedule.

[[Page 24682]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.360


[[Page 24683]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.361

    Table VI-179 shows the observed and predicted average usage rates 
for pickups by age. The observed rates initially decline at an 
increasing rate, the decline diminishes and appears to accelerate again 
for the oldest ages. The linear fit underestimates the usage rates for 
the youngest and oldest ages and overestimates middle-aged vehicles. 
The quadratic fit reasonably predicts the observed average usage rates 
but predicts an increase in usage rates for the oldest ages out of the 
observed sample. The cubic fit reasonably predicts the observed 
averages and appears to capture the diminishing decline of usage for 
the oldest ages observed in the in-sample averages. For this reason, 
the agencies used the cubic fit as the basis for the pickup VMT 
schedules.
[GRAPHIC] [TIFF OMITTED] TR30AP20.362


[[Page 24684]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.363

    As in the NPRM, the current schedule differs by body-style to 
represent different usage profiles that the agencies observed in the 
data. While more stratification is possible, it is unlikely to provide 
much additional value. Table VI-180 shows the annual miles driven at 
each age for passenger cars, SUVs (and CUVs and minivans), and pickup 
trucks at each age of their useful life, conditional upon surviving to 
that age.

[[Page 24685]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.364

(b) Benchmarking Total VMT
    In order to assess the fuel consumption and environmental impacts 
of regulatory alternatives, it is desirable to have a representation of 
aggregate travel and fuel consumption that is both reasonable and 
internally consistent. Some commenters suggested that the aggregate 
totals presented in the NPRM deviated from other published estimates, 
and argued that the entire analysis was therefore an unreliable source 
of information for decision-makers to consider. For example, EDF 
stated, ``the NHTSA model `projects' aggregate, nationwide VMT levels 
for 2016 and 2017 that are about 20 percent lower than formal 
government estimates by EIA and FHWA.'' \1790\ EDF further stated, 
``[b]etween 2017 and 2025, fleetwide VMT grows by 3.1% per year in the 
Volpe Model, while it only grows 0.5% per year in the 2018 Annual 
Energy Outlook.'' \1791\ EDF also suggested, ``[o]ne obvious way to 
assess the accuracy of the schedules is to compare the projections of 
the Volpe Model of total fleetwide fuel consumption in a recent 
calendar year with actual gasoline sales.'' \1792\
---------------------------------------------------------------------------

    \1790\ EDF, Appendix A, NHTSA-2018-0067-12108, at 59.
    \1791\ EDF, Appendix B (Rykowski comments), NHTSA-2018-0067-
12108, at 44.
    \1792\ Id. at 43.
---------------------------------------------------------------------------

    The Federal Highway Administration (FHWA) publishes annual VMT 
estimates for the light-duty vehicle fleet, the most recent of which is 
calendar year 2017. The NPRM estimate of total light-duty VMT was 2.22 
trillion miles in calendar year 2016. The FHWA estimate for light duty 
VMT in 2016 was 2.85 trillion miles.\1793\ While the definitions of 
light-duty are not identical in the two cases (where FHWA excludes 
trucks with 10,000 lbs. GVW, the agencies' analysis excludes trucks 
with GVW greater than 8,500 lbs. from its light duty definition), that 
definitional discrepancy is not significant enough to account for the 
difference in the total VMT. While some commenters suggested that the 
agencies compare simulated fuel consumption to published estimates from 
EIA to determine the validity of our VMT assumptions, such a comparison 
requires accurate assumptions about the true on-road fuel efficiency of 
registered vehicles over forty model years in addition to their annual 
usage. Comparing simulated VMT directly to FHWA measurements requires 
fewer assumptions and is a more meaningful comparison.
---------------------------------------------------------------------------

    \1793\ See Highway Statistics 2017, Table VM-1, available at 
https://www.fhwa.dot.gov/policyinformation/statistics/2017/vm1.cfm.
---------------------------------------------------------------------------

    Substituting the updated mileage accumulation schedules for the 
NPRM schedules, and using the calendar year 2016 fleet from the NPRM, 
produces an estimate of total light duty VMT in 2016 that is about 2.85 
trillion miles--nearly identical to the FHWA estimate for 2016, despite 
the use of different estimation methods and data sources. FHWA's 
estimate of total light-duty VMT in 2017 is 2.88 trillion miles,\1794\ 
while the estimate produced by the simple product of the mileage 
accumulation schedule on the estimated on-road fleet is 2.94 trillion 
miles, a difference of about two percent. While not as close as the 
estimate for calendar year 2016, the discrepancy is still small 
considering that the estimates are obtained through entirely different 
methods. One important source of

[[Page 24686]]

discrepancy with FHWA's 2017 VMT estimate is the fact that the CAFE 
model simulation assumes all of the vehicles produced in a given model 
year are driven for the entire calendar year matching the 
vintage.\1795\ This means, for calendar year 2017, the initial year of 
the simulation used to support this rule, MY2017 vehicles are assumed 
to have been both registered and driven for the entirety of CY2017. As 
a result, it naturally overestimates the true VMT for calendar year 
2017. The analysis accounts for this discrepancy by adjusting calendar 
2017 total VMT downward by one percent, and the discrepancy in total 
VMT caused by conflating model years and calendar years dissipates over 
time.
---------------------------------------------------------------------------

    \1794\ Id.
    \1795\ The CAFE model uses an annual timestep, meaning that each 
time period represents one year. Because calendar years are 
(obviously) years, and all of the other inputs (discounting and 
inflation, macroeconomic variables, fuel prices, VMT, etc.) 
represent annual values, the timestep in the CAFE model is a 
calendar year. However, model years start prior to the calendar year 
for which they are named, and new model year sales continue (albeit 
only slightly) after their calendar year ends. In order to account 
for model year sales on their true timing relative to calendar 
years, the model would need to be restructured to use a quarterly 
timestep. While this would improve the fidelity between calendar 
year and model year for sales, obtaining quarterly projections of 
nearly every other variable in the analysis would be complicated (if 
not impossible). For this reason, the model conflates ``model year'' 
and ``calendar year'' for the analysis, even though it is a 
simplification.
---------------------------------------------------------------------------

    While the agencies have established that the years for which they 
have data are sufficiently similar to published VMT estimates, the 
question of projection still remains. FHWA, in its forecasts of VMT 
(Spring 2019),\1796\ forecasts a compound annual growth rate of 0.8 
percent for light-duty vehicles between 2017 and 2047 in its baseline 
economic outlook. However, that projection uses a different set of 
macroeconomic conditions and fleet assumptions than this analysis. To 
compare CAFE model simulations of total VMT to the FHWA projections, 
the agencies ran the FHWA model with a comparable set of assumptions to 
the greatest extent possible.\1797\ \1798\ Using similar economic 
growth assumptions, our reference case total light-duty VMT grows at a 
compound rate of 0.63 percent per year between 2017 and 2050. Using 
comparable assumptions in the FHWA model produce an annual growth rate 
of 0.66 percent. Again, these differences are remarkably low for models 
created with different methods, and lead to trivial variances, for the 
purposes of our analysis, in total VMT. The relevant annual projections 
for the baseline scenario appear in Table VI-181.
BILLING CODE 4910-59-P
---------------------------------------------------------------------------

    \1796\ See ``FHWA Forecasts of Vehicle Miles Traveled (VMT): 
Spring 2019,'' Office of Highway Policy Information, available at 
https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_forecast_sum.pdf.
    \1797\ See ``FHWA Travel Analysis Framework: Development of VMT 
Forecasting Models for Use by the Federal Highway Administration,'' 
Volpe, available at https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_model_dev.pdf.
    \1798\ In particular, we ran the FHWA VMT forecasting model with 
the same: Personal disposable income, population, fuel prices (all 
of which come from AEO2019), and simulated on-road fleet fuel 
economy in the baseline.

---------------------------------------------------------------------------

[[Page 24687]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.365


[[Page 24688]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.366

BILLING CODE 4910-59-C
(c) Preserving Total VMT Across Regulatory Alternatives
    In the NPRM, the combined effect of the sales and scrappage 
responses created small percentage differences in total VMT across the 
range of regulatory alternatives.\1799\ However, as the Environmental 
Group Coalition noted, even a 0.4 percent difference can result in 
``692 billion additional VMT under the CAFE standards and 894 billion 
under the CO2 program.'' \1800\ Since VMT is related to many 
of the costs and benefits of the program, VMT of this magnitude can 
have meaningful impacts on the incremental net benefit analysis. This 
point was made by a number of commenters who were concerned about the 
magnitude and direction of differences in VMT between regulatory 
alternatives (IPI, EDF, CBD, CARB, EPA's Science Advisory Board).\1801\
---------------------------------------------------------------------------

    \1799\ The agencies explained in the NPRM that some amount of 
this difference was due to the rebound effect, and that ``non-
rebound'' VMT between alternatives differed by as much as 0.4 
percent. See 83 FR at 43099 (Aug. 24, 2018).
    \1800\ Environmental Group Coalition, Appendix A, NHTSA-2018-
0067-12000, at 180.
    \1801\ See, e.g., id.; EDF, Appendix B (Rykowski comments), 
NHTSA-2018-0067-12108, at 42-46; IPI, Appendix, NHTSA-2018-0067-
12213; at 79; CARB, Detailed Comments, NHTSA-2018-0067-11873, at 
237-242.
---------------------------------------------------------------------------

    More generally, commenters argued that non-rebound VMT should be 
held constant across regulatory alternatives, regardless of the number 
of new vehicles sold and registered vehicles scrapped. For example, CBD 
commented that the ``total number of VMT should be determined based on 
demand for travel, not arbitrarily driven by fleet size.'' CARB added 
that fleet size can change across the alternatives ``as long as the VMT 
schedules are adjusted to account for overall travel activity that is 
distributed over a larger number of vehicles.'' \1802\ NCAT, Global, 
Auto Alliance, EDF, IPI, and Honda made similar arguments.\1803\
---------------------------------------------------------------------------

    \1802\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 238 
(internal citation omitted).
    \1803\ See, e.g., Global, Attachment A, NHTSA-2018-0067-12032, 
at A-26-A-30; NCAT, Comments, NHTSA-2018-0067-11969, at 28-32; EDF, 
Appendix A, NHTSA-2018-0067-12108, at 30; IPI, Appendix, NHTSA-2018-
0067-12213, at 80-85; Honda, NHTSA-2018-0067-12111.
---------------------------------------------------------------------------

    While commenters generally provided few specific recommendations 
about the level to which VMT should be constrained among alternatives, 
several of them argued that VMT projections would benefit from 
consideration of travel demand modeling. UCS, CBD, NCAT, and others 
suggested that the overall level of light-duty VMT in a given year 
should reflect the broader economic context in which travel 
occurs.\1804\ For example, Honda stated, ``[i]ncreasing VMT is closely 
associated with increased economic activity.'' \1805\
---------------------------------------------------------------------------

    \1804\ See, e.g., NCAT, Comments, NHTSA-2018-0067-11969, at 31-
32; Environmental Group Coalition, Appendix A, NHTSA-2018-0067-
12000, at 175-76; and, UCS, Technical Appendix, NHTSA-2018-0067-
12039, at 60-61.
    \1805\ Honda, Supplemental Analysis, NHTSA-2018-0067-1211, at 4.
---------------------------------------------------------------------------

    The agencies agree that the total demand for VMT should not vary 
excessively across alternatives and stated as much in the NPRM.\1806\ 
That said, it is reasonable to assume that fleets with differing age 
distributions and inherent cost of operation will have slightly 
different annual VMT, absent VMT associated with rebound miles; 
however, the difference could conceivably be small. To address these 
comments and to take an intentionally conservative approach, the 
agencies decided to constrain ``non-rebound'' VMT (defined more 
explicitly below) to be identical across regulatory alternatives in 
this analysis using the FHWA VMT demand model to determine the 
constraint; therefore, the only difference in total VMT between 
regulatory alternatives is the rebound miles attributable to 
differences in fuel economy resulting from the regulatory alternatives. 
Nevertheless, as explained in the NPRM and revealed in the extensive 
quantitative results published with the NPRM, setting aside the rebound 
effect, aggregate VMT as estimated in the NPRM was roughly constant 
across alternatives. Although differences may have appeared large in 
absolute terms, especially when aggregated across many calendar years 
and ignoring the underlying annual total quantities, the differences 
were nevertheless very small in relative terms--small enough to be well 
within the range of measurement or estimation error for virtually any 
of the other inputs to, or outputs of, the agencies' analysis. It is 
unclear whether a 0.4 percent change in highway travel can be measured 
with any degree of confidence.
---------------------------------------------------------------------------

    \1806\ See 83 FR at 43099 (Aug. 24, 2018).
---------------------------------------------------------------------------

    To constrain non-rebound VMT, the agencies needed to create a 
definition of non-rebound VMT and a method for calculating it. The 
agencies used the FHWA VMT forecasting model to produce a forecast of 
non-rebound VMT, to which total non-rebound VMT in every regulatory 
alternative is constrained in each year, regardless of the fleet size 
or distribution of ages in the fleet. In calendar years where total 
non-rebound VMT determined by the size of the fleet and assumed usage 
of each vehicle is lower than the constraint produced from the FHWA 
model, VMT is added to that total and allocated across vehicles to 
match the non-rebound forecast (preserving the constraint). These 
additional miles are then carried throughout the analysis as vehicles 
accrue costs and benefits. Because non-rebound VMT is being held 
constant for the FRM analysis across the set of regulatory alternatives 
in each calendar year, the only difference in VMT among the

[[Page 24689]]

alternatives in any calendar year results from differences in fuel 
economy improvement relative to MY2016 that occur as a result of the 
standards. Finally, in Section VII, the agencies calculate the changes 
in total VMT attributable to fuel economy, otherwise known as the 
rebound VMT.
(i) Defining Non-Rebound VMT
    In order to constrain non-rebound VMT, it is first necessary to 
define ``non-rebound VMT'' more precisely. The NPRM defined the rebound 
effect as the overall elasticity of travel with respect to changes in 
the cost per mile (CPM). CPM has two components. The first component of 
CPM is fuel prices--the agencies expect vehicles to be driven less if 
fuel prices go up, all else equal. The second component of CPM is fuel 
economy. Therefore, the NPRM defined the percentage change in CPM, for 
a given scenario, model year, and calendar year, as: \1807\
---------------------------------------------------------------------------

    \1807\ See 83 FR at 43091 (Aug. 24, 2018).

Equation VI-7--Full change in cost per mile of travel
[GRAPHIC] [TIFF OMITTED] TR30AP20.367

Where FP is fuel price, FE is fuel economy, and REF refers to the 
reference FE value of a given age (in particular, FE 
2016-(CY-MY), which is the FE of the MY cohort that was 
age CY-MY in CY 2016). In the equation above, FESN,MY,CY 
refers to the observed fuel economy of the MY cohort (typically 
applied at the vehicle level) for a given scenario (SN) in calendar 
year CY.

    The CAFE model uses one value, the value specified as the rebound 
effect, to measure CPM elasticity. Naturally, the CAFE model produces 
the same magnitude of change in travel for equivalent changes in fuel 
prices and fuel economy. Constructing such a projection of future VMT 
(from 2017 to 2050) that sets aside the rebound effect required 
constructing inputs that were consistent with that perspective. In 
particular, it was necessary to separate the price response associated 
with the change in fuel prices relative to the year on which the 
agencies based the mileage accumulation schedule (end of CY2016), and 
the change in VMT associated with only the improvements in fuel 
economy, relative to MY2016, that occur for future model years at the 
forecasted fuel price.
    As vehicles age, the agencies expect their VMT to decrease in the 
presence of a non-zero rebound effect if rising fuel prices over time 
increase the per-mile cost of travel, and the rebound effect represents 
the degree to which their travel is reduced for a percentage change 
increase in operating cost. It is intuitive that, as the cost of fuel 
rises over time, a vehicle with a fixed fuel economy would be driven 
less if gasoline costs $3.50/gallon than it would be if gasoline costs 
$2.50/gallon. Such a response is also consistent with economic 
principles (and literature),\1808\ and so it is included in the ``non-
rebound'' VMT that the agencies constrain across alternatives in each 
calendar year.
---------------------------------------------------------------------------

    \1808\ See, e.g., Goodwin, P., J. Dargay, and M. Hanly. 
Elasticities of road traffic and fuel consumption with respect to 
price and income: A review. Transport Reviews, 24:275-292, 2004.
---------------------------------------------------------------------------

    Similarly, the annual mileage accumulation of cohorts in the 
inherited fleet is clearly affected by fuel price, but also by 
evolution. Setting aside any fuel economy improvements in vehicles sold 
and entering the on-road fleet between 2017 and 2050, the average fuel 
economy of each age cohort is going to improve over that period. The 
travel behavior of the on-road fleet was last observed through calendar 
year 2016 in the Polk data (discussed in (a)(ii)), when a 20-year-old 
car was part of the model year 1997 cohort, and had an average fuel 
economy of 23.4 MPG. However, the fleet continually turns over. In 
2035, the 20-year-old car will be a member of the model year 2016 
cohort, and have an average fuel economy of 29.2 MPG (assumed to be the 
average fuel economy of MY2016 vehicles when they were new).\1809\ If 
fuel prices persist at 2016 levels (in real dollars), then that 25 
percent improvement in fuel economy would reduce the cost per mile of 
travel for 20-year-old vehicles relative to the observed values in 
calendar year 2016, and lead to an increase in travel demand for 
vehicles of that age. Importantly, this transition to more efficient 
age cohorts occurs in all of the regulatory alternatives. Considering 
only the fuel economy levels of vehicles that exist prior to the first 
year of simulation (2017), a secular improvement in the fuel economy of 
the on-road fleet would occur with no further improvements in fuel 
economy from new vehicles in model years 2017 to 2050. As the fleet 
turns over, its fuel efficiency will gradually resemble that of the 
model year 2016 cohort, up to the point at which each age cohort is as 
efficient as the model year 2016 cohort.\1810\
---------------------------------------------------------------------------

    \1809\ In practice, vehicles will scrap at different rates over 
time, even within a body-style. Some nameplates and manufacturers 
have reputations for longevity and individual vehicle models with 
different fuel economies may seem like better candidates for repairs 
under particular fuel price scenarios. In light of this, the fuel 
economy for a given body-style will likely not continue to be the 
sales-weighted average fuel economy when the cohort was new, even 
without accounting for degradation and changes to the on-road gap 
over time. The agencies make this assumption here out of necessity.
    \1810\ Vehicles scrap at different rates over time, and there 
are important differences by body style for both scrappage rates and 
mileage accumulation. This discussion is intended to provide 
intuition, without all of the computational nuance that exists in 
the model's implementation.
---------------------------------------------------------------------------

    The notion of ``non-rebound'' VMT is a construct necessary to 
support this regulatory analysis by controlling for VMT attributable to 
reasons other than rebound driving, but present only in theory. Using 
our symmetrical definition of rebound to represent the expected 
response to changes in CPM, regardless of whether those changes occur 
as a result of changes in fuel price or fuel economy, it is well 
established that demand for VMT responds to the cost of travel. To 
isolate the change in VMT for which the regulatory alternatives are 
responsible, the agencies have also included the VMT attributable to 
secular fleet turnover (through MY2016) in the total ``non-rebound'' 
VMT projection. In particular, this means that the conventional rebound 
definition used in previous analyses, is replaced in the ``non-
rebound'' VMT estimation with a more limited definition:

Equation VI-8--Fuel price and secular improvement component of 
elasticity

[[Page 24690]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.368

Where FP is fuel price, FE is fuel economy, and REF refers to the 
reference FE value of a given age (in particular, FEREF = 
FE 2016-(CY-MY), which is the average FE of the MY cohort 
that was age (CY-MY) in CY 2016). In Equation VI-8, 
FEMIN(2016,MY)

refers to the observed fuel economy of the model year being evaluated 
up to and including the 2016MY cohort. This construction explicitly 
accounts for the improvement in fuel economy between MY2016 and all the 
historical ages (through MY1977) with respect to the change in (real) 
fuel price relative to calendar year 2016. Thus, the VMT associated 
with the rebound effect in this analysis only accounts for changes to 
CPM that result from the amount of fuel economy improvement that occurs 
relative to MY2016. The full elasticity definition (in Equation VI-7) 
differs from that in Equation VI-8 in only one way; the fuel economy in 
the denominator of the first term is the fuel economy of the model year 
being evaluated, rather than being the minimum of the actual model year 
and model year 2016.
    Combining this demand elasticity with the endogenously estimated 
vehicle population and the mileage accumulation schedule provides an 
initial estimate of non-rebound VMT, as in Equation VI-9.

Equation VI-9--Unadjusted total non-rebound VMT in a calendar year
[GRAPHIC] [TIFF OMITTED] TR30AP20.369

    In Equation VI-9, VMT represents the non-rebound mileage 
accumulation schedule (by age, A, and body style, S), Population is the 
on-road vehicle population simulated by the CAFE Model (in calendar 
year CY, for each age, A, and body style, S), [egr] is the elasticity 
of demand for travel (the rebound effect, assumed to be -0.2 in this 
analysis).
    However, there are factors beyond the CPM that affect light-duty 
demand for VMT. The FHWA VMT forecasting model includes additional 
parameters that can mitigate or increase the magnitude of the effect of 
fuel price changes on demand for VMT. In particular, the model accounts 
for changes to per-capita personal disposable income (and U.S. 
population) over time. This means that even if fuel prices are 
increasing over the study period (as they are in the central case), and 
fleetwide fuel economy improves only through fleet turnover (as it does 
in the simulated ``non-rebound'' case), total demand for VMT can still 
grow as a result of increases in these other relevant factors. Not only 
does the forecast of non-rebound VMT continue to grow in the non-
rebound case, it does so at a faster rate than Equation VI-9 produces. 
Thus, in order to preserve non-rebound VMT in a way that represents 
expected VMT demand, the agencies must constrain non-rebound VMT in 
each alternative to match the forecast produced by the FHWA model using 
the fuel price series from the central analysis, AEO2019 Reference case 
assumptions for per-capita personal disposable income, and fleetwide 
fuel economy values produced by simulating the effect of fleet turnover 
(only) in the CAFE model.\1811\
---------------------------------------------------------------------------

    \1811\ Non_rebound_VMT_forecasting.xls in Docket No. NHTSA-2018-
0067.
---------------------------------------------------------------------------

Constraining Non-Rebound VMT
    For this final rule, total `non-rebound' VMT is calculated for each 
calendar year and reported in Section VI.D.1.b)(5)(d). In any future 
calendar year, ``non-rebound'' VMT is calculated as a product of the 
initial CY2017 total and a series of compound growth rates:

Equation VI-10--Total non-rebound VMT constraint in each calendar year
[GRAPHIC] [TIFF OMITTED] TR30AP20.370

Where CY is calendar year, r is the compound annual growth rate 
(unique to each CY), and TotalVMT is the calendar year total light-
duty VMT estimated by the CAFE Model using the annual VMT for each 
body style and age in the mileage accumulation schedule (defined in 
Table VI-180), the population of each age/style cohort in CY2017, 
and the initial difference between operating costs in 2016 and 2017. 
The compound annual growth rates, rCY, in Equation VI-10 
are derived from the inter-annual differences in the forecast of 
total non-rebound VMT that the agencies created using the FHWA 
model.

    The agencies used the FHWA forecasting model to produce two 
distinct VMT forecasts (both of which appear in Table VI-182). The 
first of these is identical to the forecast of total VMT reported in 
Table VI-181, and represents the AEO2019 Reference case assumptions 
with the exception of average on-road fuel economy, which was simulated 
using the CAFE model to simulate new vehicle fuel economy, new vehicle 
sales, and vehicle retirement under the baseline standards. The 
forecast in the second column of Table VI-182 is identical to the 
first, except that the average on-road fuel economy accounts for only 
the effect of fleet turnover on fuel economy

[[Page 24691]]

improvements (new vehicles are assumed to be only as fuel efficient as 
the MY2016 cohort, discussed above).
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.371


[[Page 24692]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.372

BILLING CODE 4910-59-C
    The third column is the non-rebound VMT constraint produced by the 
CAFE model, to which non-rebound VMT is constrained to in every 
regulatory alternative (under central analysis assumptions regarding 
fuel prices and economic growth). The non-rebound VMT constraint is 
produced endogenously by the model in each run based on the estimated 
VMT for calendar year 2017 and a series of growth rates intended to 
reproduce the general growth trend in light-duty VMT under the set of 
``non-rebound'' assumptions in the FHWA model (Equation VI-10).\1812\ 
It differs from the ``non-rebound'' forecast produced by the FHWA model 
by one to three percent in any year. This adjustment was both an 
attempt to match the FHWA model's projection of total VMT (including 
rebound) in the baseline, and an acknowledgment that differing levels 
of modeling resolution and construction are likely to produce slightly 
different projections. In general, the one to three percent difference 
in non-rebound VMT is within the range of projections based on the 
confidence intervals of the coefficients that define the FHWA 
forecasting model.
---------------------------------------------------------------------------

    \1812\ This ensures internal consistency with the set of 
assumptions provided by the user, but can lead to differences 
between the non-rebound VMT constraint in the central analysis and 
one that is generated under a different set of assumptions (as in 
the sensitivity analysis, for example).

---------------------------------------------------------------------------

[[Page 24693]]

    The fourth column in Table VI-182 represents the unadjusted ``non-
rebound'' VMT produced by the CAFE Model using Equation VI-9. The 
reader will observe that in every calendar year, this total is lower 
than the non-rebound VMT constraint. This occurs because the projected 
fuel prices in the central analysis increase much faster than the 
fleetwide fuel economy (in the non-rebound case). This increases CPM 
and, as a consequence, reduces demand for VMT based on the price 
elasticity of demand for travel (rebound effect). However, the FHWA 
model accounts for additional variables that recognize the economic 
context in which this fuel price projection occurs. In particular, the 
model accounts for changes in the U.S. (human) population and changes 
to personal disposable income over the same period. These factors act 
to attenuate the demand response to rising fuel prices, producing a 
rising demand for VMT even as the CPM rises for several years.
    In order to constrain non-rebound VMT to be identical in each year 
across regulatory alternatives, it is necessary to add VMT to the 
unadjusted total, endogenously calculated by the CAFE Model in each 
calendar year. These additional miles, denoted [Delta]miles for this 
discussion, represent the simple difference between the annual VMT 
constraint (column 3 of Table VI-182) and the unadjusted VMT defined in 
Equation VI-9 (above) in each calendar year.
[GRAPHIC] [TIFF OMITTED] TR30AP20.373

    Because each regulatory scenario produces a unique on-road fleet 
(in terms of the number of vehicles, the distribution of ages among 
them, and the resulting distribution of fuel economies), the total 
unadjusted VMT in each calendar year (given by Equation VI-9) will be 
unique to each regulatory scenario. As a corollary, 
[Delta]milescy will also be unique to each regulatory 
scenario. By distributing [Delta]milescy across the vehicle 
fleet in each calendar year, the CAFE Model scales up the unadjusted 
non-rebound VMT to equal the non-rebound VMT constraint in each 
calendar year, for each regulatory alternative. While there are a 
number of ways to reallocate [Delta]milescy across the on-
road fleet in order to match the non-rebound VMT constraint, the fact 
that unadjusted VMT is always lower suggests an obvious approach.
    The primary goal of reallocation is to adjust total non-rebound VMT 
so that it is identically equal to the VMT constraint in every calendar 
year for each regulatory alternative, while conserving the general 
trends of the mileage accumulation schedule--which represents a good 
estimate of observed usage at the start of the simulation. In 
particular, the reallocation approach should preserve the basic ideas 
that annual mileage decreases with vehicle age because newer (and more 
efficient) vehicles are more likely to be driven additional miles than 
their older counterparts, and mileage accumulation varies by body 
style. To accomplish the reallocation, the CAFE Model computes a ratio 
that varies by body style, calendar year, and regulatory alternative. 
The ratio captures the share of additional VMT that can be absorbed by 
the registered vehicle population of each body style based on their 
relative representation in the fleet, so that per-vehicle totals across 
ages remain sensible (even if the distribution of body styles should 
change over time as the new vehicle market evolves). Then this quantity 
is further scaled by the total VMT for a given body style in the 
calendar year for which [Delta]miles has been computed. The resulting 
ratio is then used to scale the unadjusted miles from Equation VI-9, so 
that the new sum of annual (non-rebound) VMT across all of the vehicles 
in the on-road fleet equals the constraint. For a single calendar year, 
CY, and a single body style, S, the scaling ratio, R, is computed as:
[GRAPHIC] [TIFF OMITTED] TR30AP20.374

    In Equation VI-12, Population, refers to the on-road vehicle 
population for a given age and body style (summed over the full range 
of ages in the simulation, where vehicles are modeled to survive for, 
at most, forty years). The fraction in the numerator calculates the 
fleet composition by body type.\1813\ As long as the unadjusted non-
rebound VMT produced by the CAFE Model is smaller than the VMT 
constraint for all years and regulatory alternatives (and it is), this 
scaling ratio allows the CAFE Model to add miles to the annual total in 
a way that preserves the basic ideas of the mileage accumulation 
schedule and achieves equality with the constraint. In particular, the 
total adjusted non-rebound VMT is then calculated as:
---------------------------------------------------------------------------

    \1813\ We also considered basing this ratio on each body style's 
share of total VMT in that calendar year. However, that approach has 
the potential to result in allocations that add (or remove) too many 
miles per vehicle, depending on the age distribution and size of 
each body style cohort. While that approach better preserves the age 
distribution of VMT within a style, capturing the differences in age 
distribution of the population in each scenario is an objective of 
the VMT accounting. In testing, the differences in approach were 
small (about 0.1 percent difference).

---------------------------------------------------------------------------

[[Page 24694]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.375

    To make each alternative match the VMT constraint, Equation VI-13 
allocates miles (in this case, adds) to each vehicle in a calendar year 
by multiplying the product of the mileage accumulation schedule (for 
that style vehicle, at that age), the %[Delta]NrbdCPM (described in 
Equation VI-8), and the elasticity (the rebound effect of -0.2) with 
the appropriate scaling ratio (defined in Equation VI-12). The 
``Allocated Miles'' in Table VI-176 are the result of this calculation 
for a passenger car in CY2020.
    Unlike some of the accounting, which focuses on the impacts to a 
model year cohort of vehicles over the course of its useful life, the 
rebound constraint and reallocation are calendar year concepts. The 
constraint represents demand for VMT absent ``rebound miles'' (defined 
more explicitly above) in a specific calendar year. Thus, this 
reallocation occurs in every calendar year, and a vehicle of a model 
year cohort will likely experience many of these reallocation events 
during its simulated useful life. The resulting survival weighted 
mileage accumulation is discussed in detail in the discussion of VMT 
Resulting From Simulation found in Section (d), but an example of the 
annual reallocation is provided here.
    In the baseline alternative, the non-rebound VMT constraint in 
CY2020 is about 3.068T miles, but the endogenously computed ``non-
rebound'' VMT is only 2.955T miles. This creates a difference, 
[Delta]miles2020, of 112.6B miles that must be added to the 
total unadjusted non-rebound VMT in calendar year 2020 and allocated 
across the on-road fleet in that year to preserve total non-rebound 
VMT. Over time, this discrepancy between the FHWA model's projection 
and the unadjusted total non-rebound VMT grows to about 230 billion 
miles. While the other classes operate identically, this example uses 
the reallocation that occurs to passenger cars to illustrate the 
mechanics of reallocation. Rising fuel prices depressing non-rebound 
VMT (relative to the mileage schedule) over time is a general trend 
that emerges for all body styles, as shown for passenger cars in Table 
VI-183.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.376


[[Page 24695]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.377

BILLING CODE 4910-59-C
    The number of miles added to each age vehicle is generally less 
than the difference between the unadjusted non-rebound VMT (for a given 
age) and the mileage schedule. Thus, adding the requisite miles to each 
age does not distort either the shape of the schedule with age, nor 
does it create annual usage estimates that are out of line with 
observed usage. The example shown here uses the baseline alternative to 
illustrate the reallocation of VMT in 2020, but this reallocation 
differs by alternative. In less stringent regulatory alternatives, new 
vehicles are less expensive; this increases new vehicle sales and 
accelerates the retirement of older vehicles (relative to the 
baseline). In those cases, the unadjusted non-rebound VMT is higher, 
[Delta]miles smaller, and corresponding allocation of [Delta]miles 
smaller--though still consistently positive.
    Commenters encouraged us to use a demand model to avoid creating 
unrealistic VMT projections that failed to account for factors that 
exogenously influence total demand for VMT, which the agencies have 
done here.\1814\ Had baseline case been used instead, regardless of 
whether it happens to be the most or least stringent alternative, as 
the non-rebound VMT constraint, both the non-rebound VMT and VMT with 
rebound would have differed meaningfully from both other government 
forecasts and from the projections produced by the demand models 
underlying those forecasts. By producing and enforcing a non-rebound 
constraint based on results from a travel demand model, the agencies 
ensure realism in the projections of total VMT under each regulatory 
alternative and ensure that the costs and benefits associated with 
rebound VMT result only from fuel economy improvements in the 
regulatory alternatives considered.
---------------------------------------------------------------------------

    \1814\ See, e.g., NCAT, Comments, NHTSA-2018-0067-11969, at 31-
32; Environmental Group Coalition, Appendix A, NHTSA-2018-0067-
12000, at 175-76; UCS, Technical Appendix, NHTSA-2018-0067-12039, at 
59; Honda, Supplemental Analysis, NHTSA-2018-0067-1211, at 4.
---------------------------------------------------------------------------

(d) VMT Resulting From Simulation
    This section has already demonstrated that total VMT projections 
from the simulation are consistent with FHWA projections of total light 
duty VMT using the same set of economic assumptions. Lifetime mileage 
accumulation is now a function of the sales model, scrappage model, 
mileage accumulation schedules (described in Table VI-180), and the 
redistribution of VMT across the age distribution of registered 
vehicles in each calendar year to preserve the non-rebound VMT 
constraint.
    The definition of ``non-rebound'' VMT in this analysis determines 
the

[[Page 24696]]

additional miles associated with secular fleet turnover and fuel price 
changes. Conversely, rebound miles measure the VMT difference due to 
fuel economy improvements relative to MY2016 (independent of changes in 
fuel price, or secular fleetwide fuel economy improvement resulting 
from the continued retirement of older vehicles and their replacement 
with newer ones). In order to calculate total VMT with rebound, the 
agencies apply the rebound elasticity to the full change in CPM and the 
initial VMT schedule, but apply the rebound elasticity to the 
incremental percentage change in CPM between the non-rebound and full 
CPM calculations to the miles applied to each vehicle during the 
reallocation step that ensured adjusted non-rebound VMT matched the 
non-rebound VMT constraint.
[GRAPHIC] [TIFF OMITTED] TR30AP20.378


Where VMTA,S is the initial VMT schedule by age and body-style, 
%[Delta]NonReboundCPM and %[Delta]CPM are defined in Equation VI-8 
and Equation VI-7, respectively, and [Delta]MilesA,S,CY is the per-
vehicle miles added by the reallocation described in Equation VI-13. 
The additional miles that are added to each vehicle in the 
reallocation step ([Delta]MilesA,S,CY) are multiplied by the 
difference between the percentage changes in CPM (full and non-
rebound, respectively) because the %[Delta]NonRbdCPM was used to 
derive the allocated miles and using the full CPM change to scale 
the allocated miles would count that change twice. Taking the 
difference avoids overestimating the total mileage in the presence 
of the rebound effect. The ``rebound miles'' will be the difference 
between Equation VI-14 and Equation VI-10 for each alternative. To 
the extent that regulatory scenarios produce comparable numbers of 
rebound miles in early calendar years, the impacts associated with 
those miles net out across the alternatives in the benefit cost 
analysis.
BILLING CODE 4910-59-P

    Table VI-184 displays the annual survival-weighted VMT at each age 
of a MY2025 vehicle, by regulatory class including and reallocation 
needed to preserve the VMT constraint and all rebound miles (using a 20 
percent rebound effect).\1815\
---------------------------------------------------------------------------

    \1815\ Annual survival-weighted VMT is calculated by dividing 
the annual VMT of a MY cohort by the total population of the cohort 
purchased. As such, Table VI-183 and Table VI-184 report different 
types of values.
[GRAPHIC] [TIFF OMITTED] TR30AP20.379


[[Page 24697]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.380

BILLING CODE 4910-59-C
    As earlier portions of this section have shown, the second decade 
of useful life now shows significantly higher utilization than the NPRM 
analysis for both passenger cars and light trucks. While the current 
lifetime accumulation is similar to the values produced in the 2012 
final rule, those values were simulated to occur under fuel prices that 
were consistently 40 percent higher than the prices in this analysis 
(when adjusted for inflation).\1816\ Under comparable prices, lifetime 
mileage accumulation would have been considerably higher.
---------------------------------------------------------------------------

    \1816\ The 2012 final rule also assumed a 10 percent rebound 
effect, which would have further affected lifetime mileage 
accumulation.
---------------------------------------------------------------------------

(e) Sales, Scrappage and VMT Integration
    The VMT construct described above, while an improvement over the 
version presented in the NPRM for the reasons explained, does not 
represent the fully integrated model of ownership, usage, and 
retirement decisions that some commenters argued would be preferred or 
even required to assess properly the impacts of CAFE/CO2 
standards. In particular, RFF commented that integrating sales, 
scrappage and VMT would ``make the analysis internally consistent and 
will account for the fact that households do not make scrappage and 
vehicle use decisions in isolation.'' \1817\ IPI concurred and expanded 
in their comment, stating `` `a

[[Page 24698]]

unified model of vehicle choice and usage' is necessary.'' \1818\
---------------------------------------------------------------------------

    \1817\ RFF, Comments, NHTSA-2018-0067-11789 at 14.
    \1818\ IPI, Appendix, NHTSA-2018-0067-12213, at 80 (internal 
citation omitted).
---------------------------------------------------------------------------

    The implication of such commenters is that the agencies have 
ignored important benefits of more stringent standards by not 
explicitly considering household decisions at the level of household 
vehicle fleet management. However, the opposite may be true. A recent 
National Bureau of Economic Research (``NBER'') paper finds that 
households engage in attribute substitution while managing the set of 
attributes in their vehicle portfolios.\1819\ In particular, the 
authors argue that attribute substitution within a household's vehicle 
portfolio may erode up to 60 percent of the intended fuel economy 
benefits of the footprint-based CAFE/CO2 standards, as the 
higher fuel economy of owned vehicles reduces demand for efficiency in 
the next bought vehicle, all else equal. This suggests that examining 
effects at the household level may not be as beneficial, or as 
meaningful, as some commenters might hope.
---------------------------------------------------------------------------

    \1819\ Archsmith, J., Gillingham, K., Knittel, C., Rapson, D. 
(Sept. 2017), Attribute Substitution in Household Vehicle 
Portfolios. NBER Working Paper No. NBER Working Paper No. 23856. 
Available at https://www.nber.org/papers/w23856 (last accessed Feb. 
4, 2020).
---------------------------------------------------------------------------

    While commenters have suggested ambitious models of dynamic 
relationships at the household level, moreover, it is not clear that 
such a model is currently possible. Capturing the heterogeneous 
preferences of households across purchase, usage, and retirement 
decisions at the same level of detail required to produce meaningful 
estimates of regulatory compliance costs is beyond the current scope of 
this analysis. While the agencies agree that expected usage influences 
the household decision of which vehicle to purchase, how long to hold 
it, and how to manage the usage and retirement of other vehicles within 
a household fleet, the agencies do not agree that such a detailed model 
is a necessary prerequisite to assess the impacts of CAFE and tailpipe 
CO2 emissions standards, nor that it is necessarily 
appropriate to do so given that the agencies are examining aggregate 
national fleetwide effects of such standards. Furthermore, in the most 
recent peer review of the CAFE Model, one reviewer remarked that while 
the sales and VMT would benefit from a household choice model, ``the 
decision to scrap a vehicle (remove it from the national in-use fleet) 
and the decision to purchase a new vehicle often are not made by the 
same household. No U.S. national-level transportation demand models 
(that this reviewer is aware of) tackle the issue with this level of 
complexity.'' \1820\
---------------------------------------------------------------------------

    \1820\ CAFE Model Peer Review, DOT HS 812 590, Revised (July 
2019), pp. B19-B29, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
---------------------------------------------------------------------------

    Each iteration of these regulatory analyses has endeavored to 
improve the accuracy and breadth of modeling to capture better the 
relevant dynamics of the markets affected by these policies. The 
agencies intend to address current limitations in future rulemakings, 
and meanwhile believe that the scope of the current analysis is 
reasonable and appropriate for informing decision-makers as to the 
effects of different levels of CAFE and tailpipe CO2 
emissions stringency.
(6) What is the mobility benefit that accrues to vehicle owners?
(s) Mobility Benefits in the NPRM Analysis
    As the proposal noted, the increase in travel associated with the 
rebound effect provides benefits that reflect the value to drivers and 
other vehicle occupants of the added--or more desirable--social and 
economic opportunities that become accessible with additional travel. 
The fact that drivers and their passengers elect to make more frequent 
or longer trips to gain access to these opportunities when the cost of 
driving declines demonstrates that the benefits they gain by doing so 
exceed the costs they incur, including the economic value of their 
travel time, fuel and other vehicle operating costs, and the economic 
cost of safety risks drivers assume. The amount by which the benefits 
of this additional travel exceeds its economic costs measures the net 
benefits drivers and their passengers experience, usually referred to 
as increased consumer surplus.
    Under the proposal, the fuel cost of driving each mile would have 
increased as a consequence of the lower fuel economy levels it 
permitted, thus reducing the number of miles that buyers of new cars 
and light trucks would drive as the well-documented fuel economy 
rebound effect operates in reverse.\1821\ The agencies' analysis of the 
proposed rule described the resulting loss in consumer surplus, and 
calculated its annual value using the conventional approximation, which 
is one half of the product of the increase in vehicle operating costs 
per vehicle-mile and the resulting decrease in the annual number of 
miles driven. Because the value of this loss depends on the extent of 
the change in fuel economy, it varied by model year, and also differed 
among the alternative standards that the NPRM considered.
---------------------------------------------------------------------------

    \1821\ Normally, the fuel economy rebound effect refers to an 
increase in vehicle use that results when increased fuel economy 
reduces the fuel cost for driving each mile.
---------------------------------------------------------------------------

    The agencies' analysis specifically recognized that the economic 
value of any additional travel prompted by the fuel economy rebound 
effect must exceed the additional fuel costs drivers incur, plus the 
economic cost of safety risks they and their passengers assume.\1822\ 
Thus, when vehicle use was projected to decline in response to lower 
fuel economy, the agencies noted that the resulting loss in benefits 
must have more than offset both the savings in fuel costs and the value 
of drivers' and passengers' reduced exposure to safety risks. In the 
accounting of benefits and costs for the preferred alternative, the 
loss of benefits associated with reduced mobility was recognized by 
reporting losses in travel benefits that exactly offset the value of 
reduced risks of being involved in both fatal and non-fatal crashes.
---------------------------------------------------------------------------

    \1822\ Although it did not attempt to estimate operating costs 
other than those for fuel or the value of drivers' and passengers' 
travel time, the benefits from any additional travel that occurs 
voluntarily must also at least compensate for these costs.
---------------------------------------------------------------------------

    In addition, the accounting reported a loss in mobility benefits 
from reduced use of new cars and light trucks, which included a 
component that exactly offset the fuel savings from reduced driving, 
together with the loss in consumer surplus that foregone travel would 
otherwise have provided. Including this first component was necessary 
to offset the fact that the savings in fuel costs had already been 
recognized elsewhere in the accounting, by deducting those savings from 
the increase in fuel costs resulting from lower fuel economy to arrive 
at the reported net increase in fuel costs. Thus, the resulting value 
of the net loss in travel benefits was exactly equal to the loss in 
consumer surplus that any travel foregone in response to higher fuel 
costs would otherwise have provided.
(b) Comments on the Agencies' Treatment of Mobility Benefits in the 
NPRM
    The agencies received only two comments referring to their 
treatment of mobility benefits in the analysis supporting the proposed 
CAFE and CO2 standards. The California Air Resources Board 
(CARB) noted that the accounting of benefits and costs resulting from 
the proposal included losses in mobility benefits that offset the 
reduction in fatality costs related to the decline in

[[Page 24699]]

new vehicle use from the fuel economy rebound effect. While CARB did 
not comment on the agencies' inclusion of losses in mobility benefits 
in their accounting, it did object to the fact that the agencies also 
reported the numerical change in fatalities that could be ascribed to 
the rebound effect, and considered the improvement in safety it 
reflected when selecting their proposed alternative.\1823\ Similarly, 
the Institute for Policy Integrity (IPI) termed the agencies' reliance 
on the estimated change in the number of fatalities as partial 
justification for selecting their preferred alternative as arbitrary, 
while at the same time arguing that the reduction in driving due to the 
rebound effect had no net welfare impact.\1824\
---------------------------------------------------------------------------

    \1823\ California Air Resources Board (CARB), NHTSA-2018-0067-
11873, at pp. 121.
    \1824\ Institute for Policy Integrity (IPI), NHTSA-2018-0067-
12213, at pp. 11. In fact, the agencies did not treat the reduction 
in driving as having no net impact on welfare, since as explained 
immediately above, the loss in consumer surplus benefits on the 
foregone driving was not accompanied by any offsetting cost savings. 
Therefore, the decline in driving in response to the rebound effect 
resulted in a net loss in welfare.
---------------------------------------------------------------------------

    In response to these comments, the agencies observe that 
considering changes in the actual number of fatalities as well as the 
welfare effects of changes in drivers' and passengers' exposure and 
valuation of the risks of being involved in fatal crashes represents a 
sound approach to assessing the impacts of proposed CAFE and 
CO2 standards. The safety implications of alternative future 
standards are clearly a legitimate and highly visible consequence for 
the agencies to consider when evaluating their relative merits, as are 
the implications of changes in the safety risks for the economic 
welfare of car and light truck users. Thus the agencies see no 
inconsistency or duplication in separately considering both factors as 
part of their assessment of alternative future standards.
(c) Mobility Benefits in the Final Rule
    The analysis supporting this final rule continues to treat losses 
in mobility benefits in the same manner the agencies previously did 
when analyzing the alternatives considered for the proposed rule. 
Because there are several subtleties in this treatment, Figure VI-75 is 
included below to clarify its details. In the figure, the demand curve 
shows the relationship of annual use of new cars (and light trucks), 
which can be thought of as their total or average annual vehicle-miles 
driven, to the cost per mile of driving.
[GRAPHIC] [TIFF OMITTED] TR30AP20.381

    The initial cost per mile OC0 consists of the per mile 
economic costs of the risks of being involved in fatal and non-fatal 
crashes, shown by the heights of Og and gd on the vertical axis, 
together with per-mile fuel costs at the baseline level of fuel 
economy, the height of segment dC0.\1825\ Annual miles 
driven at this initial per-mile cost are shown by the distance 
OM0 on the horizontal axis in Figure VI-75. When fuel 
economy declines from its baseline level under one of the regulatory 
alternatives considered, fuel costs per mile increase from 
dC0 to dC1, but the per-mile economic costs of 
crash risks (both fatal and non-fatal) are unaffected, so total costs 
per mile driven rise to OC1. In response to this increase in 
the per-mile fuel and total cost of driving, annual use declines to 
OM1.
---------------------------------------------------------------------------

    \1825\ Per-mile fuel costs are equal to the dollar price of fuel 
per gallon, divided by fuel economy in miles per gallon. For 
simplicity, this figure omits non-fuel operating costs, vehicle 
maintenance and depreciation, and the value of occupants' travel 
time. Including them would not change the analysis.
---------------------------------------------------------------------------

    The resulting loss in total benefits when vehicle use declines from 
OM0 to OM1 is the trapezoidal area 
M1acM0, but most of this loss is offset by cost 
savings from reduced driving, so the net welfare loss is considerably 
smaller. Specifically, the rectangle M1hiM0 
represents a reduction in the total economic costs of the risk that 
drivers and passengers will be involved in fatal crashes when the 
decline in driving

[[Page 24700]]

reduces their exposure to that risk. The dollar value of this area thus 
appears in the agencies' accounting of costs and benefits as both a 
benefit from that reduction in risk and an exactly offsetting loss in 
benefits from reduced mobility. The same is true of the rectangle hefi, 
the dollar value of which corresponds to both the reduction in the 
economic cost of non-fatal crash risks and an identical loss in 
mobility benefits.
    Total fuel costs for driving OM0 miles are initially the 
rectangular area dC0cf, and the decline in driving to 
OM1 that results as per-mile fuel and total driving costs 
rise changes total fuel costs to the rectangle dC1ae. 
Because these two areas share rectangle dC0be, the net 
change in fuel costs reported in the agencies' accounting consists of 
the dollar value of rectangle C0C1ab, minus that 
of rectangle ebcf. The economic value of the loss in mobility benefits 
the agencies report in their accounting is the trapezoid eacf, but part 
of that area consists of rectangle ebcf, and is thus exactly equal to 
the savings in fuel costs from reduced driving. Since this savings has 
been already incorporated in the reported change in total fuel costs, 
and it offsets part of the reported loss in mobility benefits, leaving 
only the loss in consumer surplus that travelers would otherwise have 
experienced on foregone reduced driving, the value of triangle bac, as 
the net loss in mobility benefits.\1826\
---------------------------------------------------------------------------

    \1826\ Thus the change in driving is not welfare-neutral, as IPI 
asserted in the comment cited previously; instead, it results in a 
net loss in welfare.
---------------------------------------------------------------------------

    This discussion assumes that drivers correctly estimate and 
consider--or ``internalize''--the risks of being involved in both fatal 
and non-fatal crashes that are associated with their additional 
driving. However, as is noted in the discussion of the potential 
effects of the rule on the mass of vehicles and its resulting impact on 
safety, consumers may value safety risks imperfectly. This possibility 
is accounted for in the final rule analysis by assuming the portion of 
the added safety risk that consumers internalize to be 90 percent. In 
Figure VI-75 above, this would be reflected by including a total social 
cost per mile that is higher than the C0 and C1 
values for the baseline and reduced MPG cases shown in the graphic by 
10 percent of the combined cost of fatal and non-fatal crash risks (the 
distance Od on the figure's vertical axis), while reducing the costs of 
safety risks that drivers do consider to 90 percent of the values 
shown. The higher social costs would offset a portion of the consumer 
surplus associated with additional mobility (in each case), and result 
in a small ``deadweight loss'' over the region where the social cost of 
driving exceeds the demand curve. These impacts are also fully 
accounted for in the final rule analysis.
(7) What is the sales surplus that accrues to vehicle owners?
    Buyers who would not have purchased new models with the baseline 
standards in effect but decide to do so in response to the changes in 
new vehicles' prices with less demanding standards in place will also 
experience increased welfare. Collective benefits to these ``new'' 
buyers are measured by the consumer surplus they receive from their 
increased purchases.
    At the proposed rule stage, the agencies elected to exclude the 
consumer surplus associated with new vehicle purchases because ``it is 
not entirely certain that sales of new cars and light trucks [would] 
increase in response to [the] proposed action.'' \1827\ Consumer 
surplus is a fundamental economic concept and represents the net value 
(or net benefit) a good or service provides to consumers. It is 
measured as the difference between what a consumer is willing to pay 
for a good or service and the market price. OMB circular A-4 explicitly 
identifies consumer surplus as a benefit that should be accounted for 
in cost-benefit analysis. For instance, OMB Circular A-4 states the 
``net reduction in total surplus (consumer plus producer) is a real 
cost to society,'' and elsewhere elaborates that consumer surplus 
values be monetized ``when they are significant.'' \1828\
---------------------------------------------------------------------------

    \1827\ See PRIA at 954.
    \1828\ OMB Circular A-4, at 37-38.
---------------------------------------------------------------------------

    The decision to exclude consumer surplus for new vehicles at the 
proposed rule stage was an error and inconsistent with OMB's guidance 
on regulatory analysis. The agencies are confident that lower vehicle 
prices, holding all else equal, should stimulate new vehicle sales and 
by extension produce additional consumer surplus. That preliminary 
decision was also inconsistent with other parts of the agencies' 
analysis. For instance, the agencies calculate the lost consumer 
surplus associated with reductions in driving owing to the increase in 
the cost per mile in less stringent regulatory cases, as discussed in 
Section VI.D.3. The surpluses associated with sales and additional 
mobility are inextricably linked as they capture the direct costs and 
benefits accrued by purchasers of new vehicles. The sales surplus 
captures the savings to consumers when they purchase cheaper vehicles 
and the additional mobility measures the cost of higher operating 
expenses. It would be inappropriate to include one without the other.
    The shaded area in Figure VI-76 reflects the consumer surplus 
calculated for new vehicle sales. Line C0 reflects the baseline vehicle 
cost. The final rule is expected to reduce the cost of light duty 
vehicles, as represented by dotted line C '. Consistent with other 
sections of the analysis, the agencies assume that consumers value 30 
months of fuel savings. Under the final rule, consumers are expected to 
experience higher fuel costs than they would under the baseline 
scenario, shifting costs from line C ' to line C1. The consumer surplus 
is equal to the area under the curve between Q0 and Q1.\1829\
---------------------------------------------------------------------------

    \1829\ The exact calculation is 0.5 * the increase in sales * 
the reduction in the cost of light duty vehicles net of the 
increased fuel cost.

---------------------------------------------------------------------------

[[Page 24701]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.382

(8) Implicit Opportunity Cost
    The agencies' central analysis assumes the selling price for new 
vehicles will be reduced to fully reflect manufacturers' savings in 
technology costs for complying with less stringent CAFE and 
CO2 emission standards. Specifically, new car and light 
truck prices are assumed to decline by the average savings in 
technology costs per vehicle that manufacturers would realize from 
complying with the standards this rule establishes, instead of with the 
more demanding baseline standards. The agencies' analysis assumes that 
under these final standards, attributes of new cars and light trucks 
other than fuel economy would remain identical to those under the 
baseline standards, so that changes in sales prices and fuel economy 
would be the only sources of benefits or costs to new car and light 
truck buyers. Furthermore, the agencies recognize that buyers may have 
time preferences that cause them to discount the future at higher rates 
than the agencies are directed to consider in their regulatory 
evaluations. In either case, the agencies' central analysis may 
overstate both the net private and social benefits from adopting more 
stringent fuel economy and CO2 emissions standards. For 
instance, Table VII-93 (Combined LDV Societal Net Benefits for MYs 
1975-2029, CAFE Program, 7 percent Discount Rate) shows that the CAFE 
final rule would generate $16.1 billion in total social net benefits 
using a 7 percent discount rate, but without the large net private loss 
of $26.1 billion, the net social benefits would equal the external net 
benefits, or $42.4 billion. Therefore, given that government action 
cannot improve net social benefits absent a market failure, if no 
market failure exists to motivate the $26.1 billion in private losses 
to consumers, the net benefits of these final standards are $42.2 
billion.
    As indicated earlier, EPA's Science Advisory Board urged the 
agencies to account for ``consumer preferences for performance and 
other vehicle attributes'' in their analysis.\1830\ To explore further 
the possibility that the central analysis is incomplete regarding the 
consumer benefits of other vehicle attributes, the agencies conducted a 
sensitivity analysis using a conservative estimate of this value. In 
the proposal, the agencies considered the lost value of other vehicle 
attributes in two sensitivity cases that reduced the total consumer 
benefit.\1831\ The agencies received several comments suggesting that 
the analysis of other vehicle attributes lost could be improved. For 
example, CARB commented that the ``analyses do not adequately model how 
vehicle values will change in response to improving fuel economy, or 
the competing effects of other attributes.'' \1832\ In response to 
commenters, the agencies have revised their sensitivity analyses to 
model better the impact of the standards on other vehicle attributes.
---------------------------------------------------------------------------

    \1830\ SAB at 10.
    \1831\ See PRIA at 954. See also, PRIA at 1539.
    \1832\ CARB, Detailed Comments, NHTSA-2018-0067-11873 at 189.
---------------------------------------------------------------------------

    The agencies considered, such as they did in the proposal, 
offsetting the net private costs associated with enabling more choices 
in fuel-saving technologies in a manner similar to rebound driving. 
However, the agencies believe that this approach is unnecessary, as 
such an analysis would produce nearly identical net benefits to the 
external net benefits--which the primary analysis already generates. 
Furthermore, given that consumers are free to choose more fuel-
efficient vehicles absent more stringent regulations, consumers who 
prefer certain vehicle attributes instead of fuel economy necessarily 
value those attributes more than the fuel efficiency technologies they 
voluntarily forgo. As such, a sensitivity analysis including a value 
for other vehicle attributes should more than offset the net private 
costs to consumers from the primary analysis.
    For the final rule, instead of keeping the same approach as the 
preliminary analysis, the agencies have elected to estimate consumer 
benefits of other vehicle attributes in a sensitivity case using 
similar logic to that used for the sales and scrappage models. In those 
models, the agencies assume that consumers value thirty months of 
undiscounted fuel savings. Given this assumption, it would be 
reasonable for the agencies then to assume that the value of other 
vehicle attributes must be greater than the fuel savings for the 
remaining term of the useful life of the vehicle--as these are fuel 
economy savings that consumers are clearly

[[Page 24702]]

willing to forgo. The agencies acknowledge that vehicles are typically 
sold more than once, but evidence suggests that fuel savings are 
capitalized into sales prices in the used car market.\1833\ If this is 
the case, new car purchasers would internalize the additional value on 
resale owing to fuel efficiency technologies, and the fuel savings over 
the remaining useful life less thirty months would be an appropriate 
value to use for the value of other vehicle attributes. Nevertheless, 
the agencies have elected to be conservative and, instead, opted to use 
the fuel savings over the first seventy-two months (less the first 
thirty months), which approximates the amount of time the first owner 
typically holds a new vehicle.\1834\ This value is referred to as the 
``implicit opportunity cost'' of forgoing other vehicle attributes in 
favor of increased fuel economy (or using their scarce financial 
resources to invest in savings or the purchase of other goods that they 
prefer more than fuel economy),\1835\ showing a cost savings for less 
stringent alternatives.\1836\ Unlike the sales surplus, which measures 
the consumer surplus of new vehicle buyers entering the market, the 
implicit opportunity cost contained in this sensitivity case represents 
the forgone benefits to consumers the model assumes would have 
purchased a vehicle regardless of the standards (but would prefer to 
take the upfront cost of fuel economy technologies and invest that 
money elsewhere, whether it be on different vehicle attributes or 
different goods altogether). These results are shown in Table VII-91 
through Table VII-95 (Combined LDV Societal Net Benefits (Accounting 
for Implicit Opportunity Cost) for MYs 1975-2029 CAFE Program, 3 
percent Discount Rate and 7 percent Discount Rate, as well as the C02 
Program, 3 percent Discount Rate and 7 percent Discount Rate).
---------------------------------------------------------------------------

    \1833\ For further discussion of the evidence, see section 
VI.D.2 of the preamble.
    \1834\ There are several reasons why 72 months is an appropriate 
approximation. According to a report from the Federal Reserve bank 
of Chicago the average new vehicle is owned for over 77 months as of 
2015. From the same report, the average new car financing term was 
over 67 months in 2016. (https://www.chicagofed.org/publications/working-papers/2019/2019-04; accessed: December 23, 2019). Data from 
R.L. Polk suggest that the average new car is held for 71.4 months 
(as cited in https://www.autotrader.com/car-shopping/buying-car-how-long-can-you-expect-car-last-240725). State Comptrollers and 
Treasurers referred to an IHS Markit report that the average length 
of time a consumer keeps a new car is approximately 6.6 years (78 
months). EPA-HQ-OAR-2018-0283-4153, at 2. CFA commented that new 
vehicle leases are running, on average, 68 months and new vehicles 
are being held, on average, longer than 60 months. Comments, NHTSA-
2018-0067-12005, at 76. The agencies selection of 72 months is 
comfortably within the range of these estimates, but errs towards 
the lower-end and therefore provides a conservative estimate.
    \1835\ These vehicle attributes may include any that consumers 
may value and are not explicitly modeled to be neutral across 
regulatory alternatives. For instance, trim levels, entertainment 
systems, crash avoidance technologies, etc. may be sacrificed to pay 
for higher fuel economy technology levels.
    \1836\ The implicit opportunity cost must be considered a value 
that consumers place on other vehicle attributes that is net of the 
cost of those attributes. This is the forgone consumer surplus of 
other vehicle attributes. As such it is appropriately additive to 
the technology cost/savings estimated in the primary analysis.
---------------------------------------------------------------------------

    The agencies note that the central analysis of the final rule 
features a conservative treatment of private benefits and costs that 
may bias the results in the favor of more stringent regulatory 
alternatives. This bias arises from the agencies' treatment of rebound 
driving. The agencies assume that drivers make a rational decision when 
electing to drive additional miles, which considers not only the risks 
the additional driving poses to their own lives and property, but also 
most of the risks their behavior poses to their passengers as well as 
the person and property of other road users. In such a case, drivers 
``internalize'' most of these risks, and it can be assumed that 
benefits to drivers must be more valuable to them than the risks they 
considered when deciding whether to undertake the additional driving. 
Therefore, the agencies have appropriately offset the loss in safety 
benefits, which are associated with the increased cost of driving in 
the final rule, with commensurate lost benefits of additional driving.
    In contrast, the agencies can be assured the private benefits and 
costs of fuel saving technologies (aside from the external 
environmental damages) are internalized--as there is no doubt that the 
owners of the vehicles will accrue the fuel costs/savings. The agencies 
believe it would be entirely contradictory to assert that consumers are 
rational, informed, and considerate enough to internalize the risks of 
additional driving to themselves, their passengers, as well as other 
drivers and passengers; but are not similarly rational and informed 
enough to consider the additional fuel costs of purchasing a vehicle 
without a particular fuel-saving technology. After all, existing 
regulations require that the estimated annual fuel costs of a vehicle 
are disclosed on the new vehicle a consumer intends to purchase--and no 
such disclosure exists for the risks associated with driving a rebound 
mile. The agencies' decision to offset rebound miles, but not net 
private costs stemming from enabling more choices in fuel-saving 
technologies, significantly favors more stringent alternatives.
    Another possibility, however, is that manufacturers could redirect 
some or all of their savings in technology costs to instead improve 
other attributes of cars and light trucks--passenger comfort, safety, 
carrying and towing capacity, or performance--that potential buyers 
value. For example, they could redeploy the energy efficiency 
improvements from some technologies that would otherwise have been used 
to increase fuel economy to instead improve vehicles' performance, or 
redirect spending on fuel economy technology to improve safety or 
interior comfort. Producers could also offer combinations of price 
reductions and more limited improvements in these other attributes on 
some of their models, while continuing to offer high levels of fuel 
economy on other models, and channeling their entire cost savings into 
price reductions on yet other vehicles. Individual manufacturers would 
presumably select different combinations of these strategies, each in 
an effort to realize maximum additional sales and profits.
    The agencies' analysis does not quantify specific improvements in 
other attributes manufacturers could make, or identify potential 
combinations of lower prices and improvements in other attributes they 
might offer when they face less demanding fuel economy and 
CO2 standards. Nevertheless, there is ample empirical 
evidence that tradeoffs among fuel economy and other attributes that 
buyers value are important considerations in vehicle design and 
marketing strategy, and that manufacturers commonly offer combinations 
of both higher fuel economy and improvements in other attributes when 
standards do not require them to focus exclusively on improving fuel 
economy.
    Table VI-185 summarizes empirical estimates of the tradeoffs among 
fuel economy, horsepower (for cars) or torque (for light trucks), and 
weight derived from different authors' econometric estimates of the 
``curvature'' of technology frontiers for cars and light trucks. Such 
frontiers describe the combinations of fuel economy and other 
attributes that manufacturers can provide with different levels of 
spending on vehicle design and technology, accounting for the gradual 
improvements in technology and energy efficiency that occur over time. 
The entries in the table show different authors' estimates of the 
percent increases in horsepower, torque, and weight that car and light 
truck manufacturers could instead achieve if

[[Page 24703]]

they reduced fuel economy by one percent. (Although increased weight is 
not desirable in and of itself, it is associated with features such as 
a vehicle's passenger- and cargo-carrying capacity, interior volume, 
comfort, and safety, which potential buyers do value.). It is important 
to note that these tradeoffs apply to the overall average values of 
each attribute for cars and light trucks produced during recent model 
years, rather than to the features of specific individual models.
[GRAPHIC] [TIFF OMITTED] TR30AP20.383

    For example, Table VI-185 shows that Klier & Linn estimate reducing 
the average fuel economy of cars by one percent would enable producers 
to increase their average horsepower by 0.24 percent, and Knittel's 
estimate of that tradeoff is very similar (0.26 percent). Similarly, 
those two studies estimate that reducing the average fuel economy of 
cars and light trucks by one percent would enable their weight to be 
increased by 0.34-0.39 percent, which would in turn enable 
manufacturers to make modest improvements in their passenger- and 
cargo-carrying capacity, interior volume, comfort, or safety. (Note 
that reducing average fuel economy by one percent would permit either 
power or weight to increase as indicated in the table, but not both at 
the same time.).
    The tradeoffs summarized in Table VI-185 provide some indication of 
changes in attributes other than fuel economy that manufacturers are 
likely to offer under the less demanding CAFE and CO2 
standards. For example, the agencies estimate that the baseline CAFE 
standards would have required increases in fuel economy approximately 5 
percent annually over model years 2020-26 for cars, while this rule 
reduces the required rate of increase to 1.5 percent annually. This 
less demanding standard would thus enable producers to accompany higher 
fuel economy with significant improvements in other features that new 
car buyers also value, as an alternative to simply reducing prices to 
reflect their savings in technology costs. As noted previously, they 
would do so only if they thought such a strategy would be more 
attractive to buyers, so the agencies' estimates of benefits to new car 
and light truck buyers represents the minimum improvement in utility 
they would realize.
    The historical evolution of car and light truck characteristics 
under CAFE standards may also provide some indication about how 
manufacturers are likely to respond to the less aggressive standards 
this rule establishes. Figure VI-77 and Figure VI-78 show that during 
the period when CAFE standards remained unchanged or increased slowly--
approximately 1985-2010--manufacturers gradually improved cars' and 
light trucks' average fuel economy as well as their power (or torque) 
and weight, while only modestly increasing the average interior volume 
of cars.
BILLING CODE 4910-59-P

[[Page 24704]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.384


[[Page 24705]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.385

BILLING CODE 4910-59-C
    Table VI-186 summarizes the rates of change in fuel economy and 
other attributes of cars and light trucks over that period. As it 
shows, most advances in cars' drive train technology were used to 
increase power and fuel economy, while most of the improvement in light 
trucks' energy efficiency was channeled into higher torque and weight, 
with relatively little used to improve fuel economy.
[GRAPHIC] [TIFF OMITTED] TR30AP20.386

    The last column of Table VI-186 combines the actual historical 
rates of increase in attributes other than fuel economy with the 
tradeoffs between fuel economy and other attributes shown previously in 
Table VI-185 to estimate the annual rates of increase in fuel economy 
that could have been achieved if all technological progress had been 
channeled into improving fuel economy. As it indicates, manufacturers 
could have increased the fuel economy of both cars and light trucks 
over the period spanned by Table VI-186 at almost exactly the 1.5 
percent annual rate this rule requires, if they had believed that 
sacrificing other improvements in the interest of achieving higher fuel 
economy was the most effective strategy to meet potential customers' 
demands.

[[Page 24706]]

    While this result should be regarded as illustrative, it appears to 
show that meeting even these relaxed standards may require 
manufacturers to focus on improving fuel economy instead of other 
vehicle attributes. It also suggests that meeting the more demanding 
baseline standards may have required manufacturers to make significant 
sacrifices in other attributes, rather than simply holding those other 
features at or near their current levels. Viewed from this perspective, 
while this rule might not enable manufacturers to improve other 
desirable features of cars and light trucks at the same time as they 
provide the improvements in fuel economy it requires, it may 
nevertheless prevent them from having to sacrifice other improvements 
that buyers regard as valuable in order to focus solely on complying 
with more demanding CAFE and CO2 standards.
(9) Additional Consumer Purchase Costs
    Some costs of purchasing and operating new and used vehicles scale 
with the value of the vehicle. When fuel economy standards increase the 
price of new vehicles, both taxes and registration fees increase, too, 
because they are calculated as a percentage of vehicle price. 
Increasing the price of new vehicles also affects the average amount 
paid on interest for financed vehicles and the insurance premiums for 
similar reasons. The agencies compute these additional costs as scalar 
multipliers on the MSRP of new vehicles. These costs are included in 
the consumer per-vehicle cost-benefit analysis, but, for the reasons 
described below, are not included in the societal cost-benefit 
analysis.
    It is worth noting that these costs are not included in the sales 
and scrappage models, discussed above. The agencies do not expect that 
the omission of these costs affects the sales and scrappage models 
because of how these additional costs are calculated in the modeling. 
These costs are assumed to be a fixed scalar on the average MSRP of new 
vehicles, so that their inclusion would simply scale the coefficients 
in the sales and scrappage models. While these costs have not stayed 
constant over time (particularly not over the times series from 1970 to 
today), the agencies do not have a time series dataset to accurately 
estimate these costs.
    The agencies hope to reconsider including sales taxes, registration 
fees, additional interest payments and insurance costs in the sales and 
scrappage models in future research.
(a) Sales Taxes and Registration Fees
    In the analysis, sales taxes and registration fees are considered 
transfer payments between consumers and the government and are 
therefore not considered a cost from the societal perspective. However, 
these costs do represent an additional cost to consumers and are 
accounted for in the private consumer perspective. To estimate the 
sales tax for the analysis, the agencies weighted the auto sales tax of 
each state by its population--using Census population data--to 
calculate a national weighted-average sales tax of 5.46%.\1837\
---------------------------------------------------------------------------

    \1837\ See Car Tax by State, FactoryWarrantyList.com, http://www.factorywarrantylist.com/car-tax-by-state.html (last visited June 
22, 2018). Note: County, city, and other municipality-specific taxes 
were excluded from weighted averages, as the variation in locality 
taxes within states, lack of accessible documentation of locality 
rates, and lack of availability of weights to apply to locality 
taxes complicate the ability to reliably analyze the subject at this 
level of detail. Localities with relatively high automobile sales 
taxes may have relatively fewer auto dealerships, as consumers would 
endeavor to purchase vehicles in areas with lower locality taxes, 
therefore reducing the effect of the exclusion of municipality-
specific taxes from this analysis.
---------------------------------------------------------------------------

    The agencies recognize that weighting state sales tax by new 
vehicle purchases within a state would likely produce a better estimate 
since new vehicle purchasers represent a small subset of the population 
and may differ between states. The agencies explored using Polk 
registration data to approximate new vehicle sales by state by 
examining the change in new vehicle registrations across several recent 
years. The results derived from this examination resulted in a national 
weighted-average sales tax rate slightly above 5.5%, which is almost 
identical to the rate calculated using population instead. The agencies 
opted to utilize the population estimate, rather than the registration-
based proxy of new vehicle sales, because the results were negligibly 
different and the analytical approach involving new vehicle 
registrations has not been as thoroughly reviewed.
(b) Financing Costs
    Consumers who purchase new vehicles with financing options incur an 
additional cost above the new vehicle price--interest. Based off an 
Experian data, \1838\ the analysis assumes 85% of automobiles are 
purchased through financing options. The analysis used data from Wards 
Automotive and JD Power on the average transaction price of new vehicle 
purchases, average principle of new auto loans, and the average OEM-
offered incentive as a percent of MSRP to compute the ratio of the 
average financed new auto principal to the average new vehicle MSRP for 
calendar years 2011-2016. Table VI-187 shows that the average financed 
auto principal was between 82% and 84% of the average new vehicle MSRP. 
Applying the assumption that 85% of new vehicle purchases involve some 
financing, the average share of the MSRP financed for all vehicles 
purchased, including non-financed transactions, was computed. Table-II-
34 shows that the average percentage of MSRP financed ranges between 
70% and 72%. From this, the agencies chose to assume that 70% of the 
value of all vehicles' MSRP is financed. It is likely that the share 
financed is correlated with the MSRP of the new vehicle purchased, but 
for simplification purposes, it is assumed that 70% of all vehicle 
costs are financed, regardless of the MSRP of the vehicle. The agencies 
note that this simplification does not impact the accuracy of the 
calculation of the average cost to consumers, but concede that it 
obfuscates which consumers bear the additional financing burden when 
vehicle prices increase (selection of specific vehicles is likely not 
independent of consumer characteristics). For sake of simplicity, the 
model also assumes that increasing the cost of new vehicles will not 
change the share of new vehicle MSRP that is financed; the relatively 
constant share from 2011-2016 when the average MSRP of a vehicle 
increased 10% supports this assumption. The agencies recognize that 
this is not indicative of average individual consumer transactions but 
provides a useful tool to analyze the aggregate marketplace.
---------------------------------------------------------------------------

    \1838\ A report by Experian found that 85.2% of 2016 new 
vehicles were financed, as were 85.9% of 2015 new vehicle purchases. 
Zabritski, M. State of the Automotive Finance Market: A look at 
loans and leases in Q4 2016, Experian, https://www.experian.com/assets/automotive/quarterly-webinars/2016-Q4-SAFM-revised.pdf (last 
visited June 22, 2018).

---------------------------------------------------------------------------

[[Page 24707]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.387

    From Wards Auto data, the average 48- and 60-month new auto 
interest rates were 4.25% in 2016, and the average finance term length 
for new autos was 68 months. The agencies recognize that longer 
financing terms generally include higher interest rates. The share 
financed, interest rate, and finance term length are added as inputs in 
the parameters file so that they are easier to update in the future.
    Using these inputs the model computes the stream of additional 
costs associated with financing options paid for the average financed 
purchases as follows: \1839\
---------------------------------------------------------------------------

    \1839\ As alluded to above, the principle portion of repayments 
do not represent an additional cost to consumers since it represents 
the sales price.
[GRAPHIC] [TIFF OMITTED] TR30AP20.600


    0Note: The above assumes the interest is distributed evenly over 
the period, when in reality more of the interest is paid during the 
beginning of the term. However, the incremental amount calculated as 
attributable to the standard will represent the difference in the 
annual payments at the time that they are paid, assuming that a 
consumer does not repay early. This will represent the expected 
---------------------------------------------------------------------------
change in the stream of financing payments at the time of financing.

    The above stream does not equate to the average amount paid to 
finance the purchase of a new vehicle. In order to compute this amount, 
the share of financed transactions at each interest rate and term 
combination would have to be known. Without having projections of the 
full distribution of the auto finance market into the future, the above 
methodology reasonably accounts for the increased amount of financing 
costs due to the purchase of a more expensive vehicle, on an average 
basis taking into account non-financed transactions. Financing payments 
are also assumed to be an intertemporal transfer of wealth for a 
consumer; for this reason, it is not included in the societal cost and 
benefit analysis. However, because it is an additional cost paid by the 
consumer, it is calculated as a part of the private consumer welfare 
analysis.
    It is recognized that increased financing terms, combined with 
rising interest rates, lead to longer periods before a consumer will 
have positive equity in the vehicle to trade in toward the purchase of 
a newer vehicle. This has impacts in terms of consumers either trading 
vehicles with negative equity (thereby increasing the amount financed 
and potentially subjecting the consumer to higher interest rates and/or 
rendering the consumer unable to obtaining financing) or delaying the 
replacement of the vehicle until they achieve suitably positive equity 
to allow for a trade.
(c) Insurance Costs
    More expensive vehicles will require more expensive collision and 
comprehensive (e.g., fire and theft) car insurance. Actuarially fair 
insurance premiums for these components of value-based insurance will 
be the amount an insurance company will pay out in the case of an 
incident type weighted by the risk of that type of incident occurring. 
For simplicity of this calculation, the agencies assume that the 
vehicle has the same exposure to harm throughout its lifetime. However, 
the value of vehicles will decline at some depreciation rate so that 
the absolute amount paid in value-related insurance will decline as the 
vehicle depreciates. This is represented in the model as the following 
stream of expected collision and comprehensive insurance payments:
[GRAPHIC] [TIFF OMITTED] TR30AP20.388


[[Page 24708]]


    To utilize the above framework, estimates of the share of MSRP paid 
on collision and comprehensive insurance and of annual vehicle 
depreciations are needed to implement the above equation. Wards has 
data on the average annual amount paid by model year for new light 
trucks and passenger cars on collision, comprehensive and damage and 
liability insurance for model years 1992-2003; for model years 2004-
2016, they only offer the total amount paid for insurance premiums. The 
share of total insurance premiums paid for collision and comprehensive 
coverage was computed for 1979-2003. For cars the share ranges from 49 
to 55%, with the share tending to be largest towards the end of the 
series. For trucks the share ranges from 43 to 61%, again, with the 
share increasing towards the end of the series. It is assumed that for 
model years 2004-2016, 60% of insurance premiums for trucks, and 55% 
for cars, is paid for collision and comprehensive. Using these shares 
the absolute amount paid for collision and comprehensive coverage for 
cars and trucks is computed. Then each regulatory class in the fleet is 
weighted by share to estimate the overall average amount paid for 
collision and comprehensive insurance by model year as shown in Table 
VI-188. The average share of the initial MSRP paid in collision and 
comprehensive insurance by model year is then computed. The average 
share paid for model years 2010-2016 is 1.83% of the initial MSRP. This 
is used as the share of the value of a new vehicle paid for collision 
and comprehensive in the future.
[GRAPHIC] [TIFF OMITTED] TR30AP20.389

    2017 data from Fitch Black Book was used as a source for vehicle 
depreciation rates; two- to six-year-old vehicles in 2016 had an 
average annual depreciation rate of 17.3%.\1840\ It is assumed that 
future depreciation rates will be like recent depreciation, and the 
analysis used the same assumed depreciation. Table VI-189 shows the 
cumulative share of the initial MSRP of a vehicle assumed to be paid in 
collision and comprehensive insurance in five-year age increments under 
this depreciation assumption, conditional on a vehicle surviving to 
that age--that is, the expected insurance payments at the time of 
purchase will be weighted by the probability of surviving to that age. 
If a vehicle lives to 10 years, 9.9% of the initial MSRP is expected to 
be paid in collision and comprehensive payments; by 20 years 11.9% of 
the initial MSRP; finally, if a vehicle lives to age 40, 12.4% of the 
initial MSRP.
---------------------------------------------------------------------------

    \1840\ Fitch Ratings Vehicle Depreciation Report February 2017, 
Black Book, http://www.blackbook.com/wp-content/uploads/2017/02/Final-February-Fitch-Report.pdf (last visited June 22, 2018).

---------------------------------------------------------------------------

[[Page 24709]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.390

    The increase in insurance premiums resulting from an increase in 
the average value of a vehicle is a result of an increase in the 
expected amount insurance companies will have to pay out in the case of 
damage occurring to the driver's vehicle. In this way, it is a cost to 
the private consumer, attributable to the CAFE standard that caused the 
price increase.
(10) Measuring Fuel Consumption
    The procedure the agencies use to estimate fuel consumption assumes 
that all vehicle models of the same body type--cars, SUVs and vans, and 
light trucks--and age are driven identical amounts each year. Under 
this assumption, the agencies' estimates of fuel consumption from 
increasing the fuel economy of each individual model depend only on how 
much its fuel economy is increased, and do not reflect whether its 
actual use differs from other models of the same body type. Neither do 
the agencies' estimates of fuel consumption account for variation in 
how much vehicles of the same body type and age are driven each year, 
which appears to be significant.
    This assumption may cause the agencies' estimates of fuel 
consumption from imposing stricter CAFE and CO2 standards to 
be too large. Because the distribution of annual driving is wide using 
its mean value to estimate fuel savings for individual car or light 
truck models may overstate the fuel consumption likely to result from 
tighter standards, even when the fuel economy of different models are 
correctly averaged.\1841\ This will be the case even when increases in 
fuel economy can be estimated reliably for individual models, as the 
agencies' analysis does, because the reduction in a specific model's 
fuel consumption depends on how much it is actually driven as well as 
on the increase that stricter standards require.
---------------------------------------------------------------------------

    \1841\ The correct average fuel economy of vehicles whose 
individual fuel economy differs is the harmonic average of their 
individual values, weighted by their respective use; for two 
vehicles with fuel economy levels MPG1 and 
MPG2 that are assumed to be driven identical amounts (as 
in the agencies' analysis), their harmonic average fuel economy is 
equal to 2/(1/MPG1 + 1/MPG2).
---------------------------------------------------------------------------

    To illustrate, the agencies estimate that new automobiles are 
driven about 17,000 miles on average during their first year. If the 
17,000 mile figure represents the average of two different models that 
are driven 14,000 and 20,000 miles annually, and the two initially 
achieve, respectively, 30 and 40 miles per gallon--thus averaging 35 
miles per gallon--they will consume a total of 967 gallons 
annually.\1842\ Improving the fuel economy of each model by 5 miles per 
gallon will reduce their total fuel use to 844 gallons, thus saving 123 
gallons annually.\1843\ In contrast, the agencies' would estimate total 
fuel consumption for the two vehicles using the 17,000 mile average 
figure for both, thus yielding estimated fuel savings of 128 gallons 
per year, about 5% above the correct value.\1844\
---------------------------------------------------------------------------

    \1842\ Calculated as 14,000 miles/30 miles per gallon + 20,000 
miles/40 miles per gallon = 467 gallons + 500 gallons = 967 gallons 
(all figures in this calculation are rounded to whole gallons).
    \1843\ Calculated as 14,000 miles/35 miles per gallon + 20,000 
miles/45 miles per gallon = 400 gallons + 444 gallons = 844 gallons 
(again, all figures in this calculation are rounded to whole 
gallons).
    \1844\ The agencies estimate of their combined initial fuel 
consumption would be 17,000 miles/30 miles per gallon + 17,000 
miles/40 miles per gallon, or 567 gallons + 425 gallons = 992 
gallons. After the 5 mile per gallon improvement in fuel economy for 
each vehicle, the agencies' estimate would decline to 17,000 miles/
35 miles per gallon + 17,000 miles/45 miles per gallon = 486 + 378 = 
863 gallons, yielding an estimated fuel savings of 992 gallons--863 
gallons = 128 gallons (as previously, all figures in this 
calculation are rounded to whole gallons).
---------------------------------------------------------------------------

    The magnitude of this potential overestimation of fuel savings 
increases with any association between annual driving and fuel economy, 
which seems likely to be strong. Acting in their own economic interest, 
car and light truck buyers who anticipate driving more should be more 
likely choose models offering higher fuel economy, because the number 
of miles driven directly affects their fuel costs and thus the savings 
from driving a model that features higher fuel economy.\1845\ 
Conversely, buyers who anticipate driving less are likely to purchase 
models with lower fuel economy. Such behavior--whereby buyers who 
expect to drive more extensively are likely to select models offering 
higher fuel economy--cannot be fully accounted for in today's analysis, 
because that analysis is necessarily based on

[[Page 24710]]

empirical estimates of average vehicle use. To the extent it occurs, 
the agencies are likely to consistently overstate actual fuel savings 
from requiring higher fuel economy, as well as to overstate increases 
in fuel consumption resulting from lower standards. Thus, the agencies' 
central analysis is likely to overestimate the final rule's impact on 
consumer benefits such as reduced fuel consumption and increased 
refueling time, as well as on the resulting environmental impacts of 
fuel production and use.
---------------------------------------------------------------------------

    \1845\ For example, some businesses, rental car firms, taxi 
operators, and ride sharing drivers are likely to anticipate using 
their vehicles significantly more than the average new car or light 
truck buyer. Furthermore, their choices among competing models are 
likely to be more heavily influenced by economics than by the 
preferences for other attributes that motivate many other buyers, 
making them more likely to select vehicles with higher fuel economy 
in order to improve their economic returns.
---------------------------------------------------------------------------

    A similar phenomenon may cause the agencies to overstate the value 
of fuel savings resulting from requiring higher fuel economy as well. 
As with miles driven, the agencies' analysis assumes all vehicle owners 
pay the national average fuel price at any time. However, fuel prices 
vary substantially among different regions of the U.S., and one would 
expect buyers in regions with consistently higher fuel prices to 
purchase vehicles with higher fuel economy, on average. To the extent 
they actually do so, evaluating the savings from requiring higher fuel 
economy identically in all regions using nationwide average fuel prices 
is likely to overstate their actual dollar value; similarly, assessing 
the increased fuel costs likely to result from lower standards using 
national average fuel prices is likely to overstate their true value 
insofar as car and light truck buyers facing above-average fuel prices 
choose higher-mpg models.
    As an illustration, suppose gasoline averages $3.00 per gallon 
nationwide, but a buyer who expects to drive a new car 17,000 miles 
during its first year (the same value used in the example above) faces 
a local price of $4.00 per gallon, and chooses a model that achieves 40 
mpg. That driver's cost of fuel during the vehicle's first year will 
total $1,700 (calculated at 17,000 miles/40 miles per gallon x $4.00 
per gallon). A buyer who plans to drive the same number of miles but 
faces a lower price of $2.00 per gallon and thus chooses a vehicle that 
offers only 30 mpg will have first-year fuel costs of $1,133 
(calculated as 17,000 miles/30 miles per gallon x $2.00 per gallon), so 
total annual fuel costs for these two vehicles will be $1,700 + $1,133 
= $2,633. If the fuel economy of both vehicles increases by 5 mpg, 
their actual fuel savings will be $189 and $162, or a total savings of 
$351. However, evaluating total fuel savings using the national average 
price of $3.00 per gallon yields savings of $382, thus overstating 
actual savings by about 10%. This same phenomenon would cause the 
agencies to overestimate of costs of increased fuel use when standards 
are relaxed, as with this rule.
(11) Refueling Benefit
    Increasing CAFE/CO2 standards, all else being equal, 
affect the amount of time drivers spend refueling their vehicles in 
several ways. First, they increase the fuel economy of ICE vehicles 
produced in the future and, consequentially, decrease the number of 
refueling events for those vehicles. Second, given increased production 
costs, they reduce sales of new vehicles and scrappage of existing 
ones, causing more VMT to be driven by older and less efficient 
vehicles which require more refueling events for the same amount of VMT 
driven. Finally, they may change the number of electric vehicles that 
are produced, and shift refueling to occur at a charging station, 
rather than at the pump--changing per-vehicle lifetime expected 
refueling costs. While there are multiple ways that fuel economy 
standards alter refueling costs, the proposal accounted for only the 
first. Before the inclusion of the sales and scrappage models, which 
first appeared in the NPRM analysis for the first time a CAFE/
CO2 rulemaking, the agencies did not have the means to 
capture the other two effects. While the agencies modeled sales and 
scrappage effects, they did not extend the results to refueling time. 
This oversight was noted by commenters, and the final rule model now 
includes these additional factors. The basic calculation for all three 
effects is the same: The agencies multiply the additional amount of 
time spent refueling by the value of time of passengers, which is 
assumed to be the same for all three effects.
(a) Value of Time
    The calculation of the value of time remains relatively unchanged 
from the proposal and follows the guidance from DOT's 2016 Value of 
Travel Time Savings memorandum (``VTTS Memo'').\1846\ The economic 
value of refueling time savings is calculated by applying valuations 
for travel time savings from the VTTS Memo to estimates of how much 
time is saved across alternatives.\1847\
---------------------------------------------------------------------------

    \1846\ United States Department of Transportation, The Value of 
Travel Time Savings: Departmental Guidance for Conducting Economic 
Evaluations, (2016), available at https://www.transportation.gov/sites/dot.gov/files/docs/2016%20Revised%20V.
    \1847\ VTTS Memo Tables 1, 3, and 4.
---------------------------------------------------------------------------

    IPI commented that the agencies used old data to calculate the 
refueling benefit in the proposal. Specifically, IPI pointed out that 
the data used in the proposal seemed ``to come from the 2003 version of 
[the VTTS Memo].'' \1848\ For the final rule, the analysis uses the 
most recent VTTS memo along with updated wages. The value of travel 
time depends on average hourly valuations of personal and business 
time, which are functions of annual household income and total hourly 
compensation costs to employers. As designated by the 2016 VTTS memo, 
the nationwide median annual household income, $56,516 in 2015, is 
divided by 2,080 hours to yield an income of $27.20 per hour. Total 
hourly compensation cost to employers, inclusive of benefits, in 2015$ 
is $25.40.\1849\ Table VI-190 demonstrates the agency's approach to 
estimating the value of travel time ($/hour) for both urban and rural 
(intercity) driving. This approach relies on the use of DOT-recommended 
weights that assign a lesser valuation to personal travel time than to 
business travel time, as well as weights that adjust for the 
distribution between personal and business travel.\1850\ In accordance 
with DOT guidance, wage valuations are estimated with base year 2015 
dollars and end results are adjusted to 2018 dollars.
---------------------------------------------------------------------------

    \1848\ IPI, Appendix, NHTSA-2018-0067-12213, at 51.
    \1849\ Ibid at11.
    \1850\ Business travel is higher than personal travel because an 
employer has additional expenses, e.g. taxes and benefits costs, 
above and beyond an employee's hourly wage. In the proposal, the 
agencies erroneously used the same value for personal and business 
travel, which was inconsistent with the VTTS Memo.

---------------------------------------------------------------------------

[[Page 24711]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.391

    Estimates of the hourly value of urban and rural travel time 
($14.14 and $20.40, respectively) shown in Table VI-190, must be 
adjusted to account for the nationwide ratio of urban to rural 
driving.\1851\ This adjustment, which gives an overall estimate of the 
hourly value of travel time--independent of urban or rural status--is 
shown in Table VI-191.
---------------------------------------------------------------------------

    \1851\ Estimate of Urban vs. Rural travel weights from FHWA 
December 2018 Traffic Volume Trends, Monthly Report, Table 2--
Cumulative Monthly Vehicle-Miles of Travel in Billions. Available at 
https://www.fhwa.dot.gov/policyinformation/travel_monitoring/18dectvt/page3.cfm.
[GRAPHIC] [TIFF OMITTED] TR30AP20.392


[[Page 24712]]


    Note that the calculations above consider the value of travel time 
for only one occupant. To estimate fully the average value of vehicle 
travel time per vehicle, the agencies must account for the presence of 
all additional passengers during refueling trips. The agencies 
estimated average vehicle occupancy using survey data gathered as part 
of our 2010-2011 National Automotive Sampling System's Tire Pressure 
Monitoring System (TPMS) study.\1852\ The study was conducted at 
fueling stations nationwide and researchers made observations regarding 
a variety of characteristics of thousands of individual fueling station 
visits from August, 2010 through April, 2011. Among these 
characteristics of fueling station visits, the total number of 
occupants per vehicle were observed. Average vehicle occupancy was 
calculated and multiplied by the value of travel time per occupant. As 
shown in Table VI-192, this adjustment is performed separately for 
passenger cars and for light trucks, yielding occupancy-adjusted 
valuations of vehicle travel time during refueling trips for each 
fleet. Lastly, the occupancy-adjusted value of vehicle travel time is 
converted to 2018 dollars using the GDP deflator as shown in Table VI-
193.\1853\
---------------------------------------------------------------------------

    \1852\ Docket for Peer Review of NHTSA/NASS Tire Pressure 
Monitoring System, available at https://www.regulations.gov/docket?D=NHTSA-2012-0001.
    \1853\ Bureau of Economic Analysis, NIPA Table 1.1.9 Implicit 
Price Deflators for Gross Domestic Product, available at https://apps.bea.gov/iTable/index_nipa.cfm.
[GRAPHIC] [TIFF OMITTED] TR30AP20.393

    IPI commented that the exclusion of children from the NPRM's 
refueling time analysis was inconsistent with DOT's 2016 Value of 
Travel Time Savings memorandum (``VTTS Memo''). IPI claimed that the 
VTTS Memo ``consider[ed] whether the value of travel time is different 
for parents versus children, but ultimately conclude[d] that `it must 
be assumed that all travelers' VTTS are independent and additive.' '' 
IPI also quoted language from page 13 of the VTTS Memo that 
``[a]lthough riders may be a family with a joint VTTS or passengers in 
a car pool or transit vehicle with independent values, these 
circumstances can seldom be distinguished [. . .] therefore, all 
individuals are assumed to have independent values,'' and that it is 
``inappropriate to use different income levels or sources for different 
categories of traveler.'' \1854\
---------------------------------------------------------------------------

    \1854\ See IPI, Appendix, NHTSA-2018-0067-12213, at 52-53 
(citing United States Department of Transportation (``DOT''), The 
Value of Travel Time Savings: Departmental Guidance for Conducting 
Economic Evaluations, (2016), available at https://www.transportation.gov/sites/dot.gov/files/docs/2016%20Revised%20V).
---------------------------------------------------------------------------

    IPI further asserted that excluding passengers under age 16 from 
the calculation of travel time savings was inconsistent with the best 
practices of benefit-cost analysis. IPI noted that Circular A-4 does 
not distinguish between children and adults except when monetizing 
health effects. IPI then cited Dale Whittington and Duncan MacRae as 
stating ``there is a clear consensus that children should be counted in 
cost-benefit analysis.'' Finally, IPI commented that Congress intended 
that the agencies consider the economic impact to children when setting 
standards.\1855\
---------------------------------------------------------------------------

    \1855\ See IPI, Appendix, NHTSA-2018-0067-12213, at 53-54 
(internal citations omitted).
---------------------------------------------------------------------------

    The agencies point out that the first passage from the VTTS Memo 
cited by IPI does not conclude, or even deliberate, that the VTTS of 
children is the same as adults, but instead states that the VTTS of 
children, parents and other passengers should be independent and 
additive.\1856\ Assuming that the opportunity cost of children's time 
is zero is compatible with this practice. Likewise, IPI concluded from 
the text on page 12 that it was inappropriate to use different incomes 
for children. However, IPI's analysis suffers from two errors.
---------------------------------------------------------------------------

    \1856\ See VTTS Memo at 5.
---------------------------------------------------------------------------

    First, the two quotes from page 12 reside in a section of the VTTS 
Memo

[[Page 24713]]

entitled Special Issues, which provides guidance on three distinct 
topics. The first quoted text comes from a paragraph advising how to 
treat vehicles with multiple passengers, while the second is from an 
ensuing topic about passenger incomes. It is baseless to assume that 
the conclusion of the second topic holds true for the first.
    Second, assuming IPI intended to comment that age is a ``category 
of traveler'' for which ``it is inappropriate to use different income 
levels,'' the agencies note that such an interpretation is tenuous. The 
VTTS Memo clearly recognizes that some categories of travelers should 
have different levels of income,\1857\ and provides two examples.\1858\ 
As children are not part of the workforce, they do not have wage 
incomes. Therefore, it is not wild speculation that they do not bear a 
financial opportunity cost associated with their time spent in vehicles 
during refueling.\1859\ As such, excluding children from the 
calculation of the refueling benefit is consistent with DOT's guidance.
---------------------------------------------------------------------------

    \1857\ The full text quoted by IPI reads, ``[e]xcept for 
specific distinctions, we consider it inappropriate to use different 
income levels or sources for different categories of traveler.'' 
VTTS Memo at 12 (emphasis added). The VTTS Memo further contemplates 
that it is appropriate to assign different incomes if ``estimates 
[of income are] derived by reliable and focused research [. . .] in 
specific cases.'' Id.
    \1858\ The VTTS Memo provides specific guidance on how to 
differentiate between personal and business travel, and air or high 
speed rail from other modes of transportation. See VTTS Memo at 12.
    \1859\ The TMPS study affords the agencies the opportunity to 
distinguish between adults and passengers, a luxury not available in 
every instance. Furthermore, there may be certain instances where it 
is appropriate to value the VTTS of children the same as adults, 
e.g., rules focusing primarily on the VTTS of children.
---------------------------------------------------------------------------

    Turning to IPI's comments on best practices and Congress' intent, 
the agencies agree that the benefit-cost analysis should include 
children when appropriate. The majority of the components of the CAFE 
model (e.g., safety analyses) include children. However, children are 
excluded from the analysis when it is appropriate (e.g., employment). 
For this specific valuation, it is reasonable to assume the value of a 
child's time is not equivalent to an adult's. Nonetheless, the agencies 
have examined the impact of valuing children's time as equal to adults' 
by including them in the average vehicle occupancy rates applied in the 
refueling analysis and using the full VTTS for personal travel. Results 
indicate that the effect of this issue is minor and impacts total 
benefits by about one-quarter percent. The agencies will continue to 
consider this issue in future CAFE and CO2 rulemakings. IPI 
also noted that the only portion of the TPMS publicly available was the 
``User's Coding Manual.'' Specifically, IPI argued that ``the agencies' 
failure to make available the full data and methodology used to 
calculate these average occupancy figures frustrates any meaningful 
public review.'' The agencies disagree. IPI was able to submit a 
meaningful comment about the agencies' decision to exclude children 
from the occupancy-adjusted value of vehicle travel time. Furthermore, 
commenters knew that the agencies intended to use occupancy estimates 
to calculate the refueling benefit; however, the agencies did not 
receive any alternative estimates or methodologies from commenters. 
Nonetheless, the agencies have provided reference to the docket folder 
containing peer review documents, analysis documentation, and data for 
the 2011 TPMS survey.
(b) Accounting for Improved Fuel Economy of ICE Vehicles
    The methodology for calculating the refueling benefits associated 
with improved fuel economy in new vehicles remains unchanged from the 
proposal. The CAFE model calculates the number of refueling events for 
each ICE vehicle in a calendar year. This is calculated as the number 
of miles driven by each vehicle in that calendar year divided by the 
product of that vehicle's on road fuel economy, tank size, and an 
assumption about the average share of the tank refueled at each event, 
as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.394

    The model then computes the cost of refueling as the product of the 
number of refueling events, total time of each event and value of the 
time spent on each event (computed as average salary), as below:
    The event time of a vehicle is calculated by summing a fixed and 
variable component. The fixed component is the number of minutes it is 
assumed each event takes, independent of any assumptions about tank 
size or share refueled at each event (the time it takes to get to and 
from the pump). The variable component is the ratio of the average 
number of gallons refueled for each event (the product of the tank size 
and share refueled) and the rate at which gallons flow from the pump. 
This is shown below:
[GRAPHIC] [TIFF OMITTED] TR30AP20.600

    1In order to calculate the refueling time cost, as described above, 
the CAFE model takes the following inputs: The value of time, the fixed 
time component of each refueling event, share of the tank refueled at 
each event, rate of flow of fuel from the pump, and vehicle tank size. 
The first of these is taken from DOT guidance on travel time savings. 
The fixed time component, share refueled, and rate of flow are 
calculated from survey data gathered as part of our 2010-2011 National 
Automotive Sampling System's Tire Pressure Monitoring System (TPMS) 
study.\1860\ Finally, the vehicle fuel tank sizes are taken from 
manufacturer specs for the reference fleet and historical averages are 
calculated from popular models for the existing vehicle fleet, as 
described, below, in discussion of the legacy fleet.
---------------------------------------------------------------------------

    \1860\ Docket for Peer Review of NHTSA/NASS Tire Pressure 
Monitoring System, available at https://www.regulations.gov/docket?D=NHTSA-2012-0001.
---------------------------------------------------------------------------

    The agencies estimated the amount of saved refueling time using 
survey data gathered as part of the aforementioned TPMS study. In this 
nationwide study, researchers gathered information on the total amount 
of time spent pumping and paying for fuel. From a separate sample (also 
part of the TPMS study),

[[Page 24714]]

researchers conducted interviews at the pump to gauge the distances 
that drivers travel in transit to and from fueling stations, how long 
that transit takes, and how many gallons of fuel are purchased.
    The agencies focused on the interview-based responses in which 
respondents indicated the primary reason for the refueling trip was due 
to a low reading on the gas gauge. Such drivers experience a cost due 
to added mileage driven to detour to a filling station, as well as 
added time to refuel and complete the transaction at the filling 
station. The agencies believe that drivers who refuel on a regular 
schedule or incidental to stops they make primarily for other reasons 
(e.g., using restrooms or buying snacks) do not experience the cost 
associated with detouring in order to locate a station or paying for 
the transaction, because the frequency of refueling for these reasons 
is unlikely to be affected by fuel economy improvements. This 
restriction was imposed to exclude distortionary effects of those who 
refuel on a fixed (e.g., weekly) schedule and may be unlikely to alter 
refueling patterns as a result of increased driving range. The relevant 
TPMS survey data on average refueling trip characteristics are 
presented below in Table VI-194.
[GRAPHIC] [TIFF OMITTED] TR30AP20.395

    The agencies assume that all of the round-trip time necessary to 
travel to and from the fueling station is a part of the fixed time 
component of each refueling event. However, some portion of the time to 
fill and pay is also a part of the fixed time component. Given the 
information in Table VI-194, the agencies assume that each refueling 
event has a fixed time component of 3.5 minutes. E.g., (for passenger 
cars) the sum of 2.28 minutes round trip time to/from fueling station 
and roughly 1.2 minutes to select and pay for fuel, remove/recap fuel 
tank, remove/replace fuel nozzle, etc. The time to fill the fuel tank 
is the variable time component; e.g., about 2.9 minutes for passenger 
cars (2.28 + 1.2 + 2.9 = 6.38 total minutes). However, the CAFE model 
uses a different methodology to determine the variable time component, 
which is explained below.
    Cars have average tank sizes of about 15 gallons, SUVs/vans of 
about 18 gallons, and pickups of about 27 gallons (see Table VI-195 
through Table VI-197 in discussion of the legacy fleet). It is a 
reasonable assumption that the average passenger car has a tank of 15 
gallons and the average light truck has a tank of 20 gallons (there are 
more SUVs/vans than pickups in the light truck fleet). From these 
assumptions, it is calculated that the average refueling event fills 
approximately 65 percent of the fuel tank for both passenger cars and 
light trucks. This value is used as an input in the CAFE model for all 
three body styles (cars, SUVs/vans, and pickups).
    Finally, the rate of the pump flow can be calculated either as the 
total gallons pumped over the assumed variable time component 
(approximately 3 minutes) or as the difference in the average number of 
gallons filled between light trucks and passenger cars over the 
difference in the time to fill and pay between the two classes. The 
first methodology implies a rate between 3 and 4 gallons per minute. 
Although the second methodology implies a rate of 15 gallons per 
minute, there is a legal restriction on the flow of gasoline from pumps 
of 10 gallons per minute.\1861\ Thus, the agencies assume the rate of 
gasoline pumps range between 4 and 10 gallons per minute, and use 7.5 
gallons per minute--a value slightly above the midpoint of that range--
as the average flow rate in the CAFE model.
---------------------------------------------------------------------------

    \1861\ 40 CFR 80.22(j), Regulation of Fuels and Fuel Additives--
subpart B. Controls and Prohibitions, available at https://www.law.cornell.edu/cfr/text/40/80.22.
---------------------------------------------------------------------------

    The calculations described above are repeated for each future 
calendar year that light-duty vehicles of each model year affected by 
the CAFE standards considered in this rule would remain in service for 
each regulatory alternative. The resulting cumulative lifetime 
valuations of time savings account for both the reduction over time in 
the number of vehicles of a given model year that remain in service and 
the reduction in the number of miles (VMT) driven by those that stay in 
service. After calculating the absolute value for each regulatory 
alternative using the methodology and inputs described above, the model 
calculates the incremental value relative to the baseline as the 
refueling cost or benefit for that regulatory alternative. More 
efficient vehicles have to be refueled less often and refueling costs 
per vehicle decline. In previous rules this was sufficient to account 
for the majority of any changes in cost of refueling under different 
CAFE standards as the modelling permitted, since the volumes of new 
vehicles and existing vehicles on the road was assumed to be constant 
under all possible standards. However, when sales and scrappage models 
are included the distribution of new and vehicles varies and a 
different number of miles will be driven by new and used vehicles in 
each regulatory alternative.
    IPI commented that it was inappropriate for the agencies to

[[Page 24715]]

exclude benefits from reducing the frequency of refueling events where 
the primary reason for stopping at a fuel station was not to refuel a 
vehicle. IPI argued that fuel efficiency impacts from relaxed standards 
would affect all drivers regardless of their rationale for refueling, 
by requiring either more frequent or marginally longer refueling 
events.\1862\ The agencies note that the language in the NPRM suggested 
that the agencies eliminated 40 percent of the potential benefit from 
fewer refueling stops--where 40 percent represents the fraction of 
refueling stops that were routinely scheduled or otherwise not made in 
response to a low fuel reading--and this appears to have been the 
origin of IPI's concern.\1863\ In fact, the agencies did not apply a 40 
percent discount factor to the refueling benefits; instead, the total 
number of additional refueling events that would result from 
alternative CAFE levels was calculated, and these were valued based on 
an assumption that their characteristics (e.g., vehicle occupancy) 
would match those of drivers who refueled due to a low fuel reading.
---------------------------------------------------------------------------

    \1862\ IPI, Appendix, NHTSA-2018-0067-12213, at 54-55.
    \1863\ See 83 FR 43088 (Aug. 24, 2018).
---------------------------------------------------------------------------

    To the extent that lower fuel economy affects those who refuel on a 
routine schedule or incidental to stops made primarily for other 
reasons, the per-event cost would actually be limited to the extra time 
spent pumping a slightly larger volume of fuel. However, the agencies 
note that by assuming that all extra fuel consumed under lower CAFE 
standards results in added refueling trips, the agencies are adopting a 
conservative assumption, in the sense that it maximizes the disbenefits 
of alternatives to the current standards.
    IPI also expressed concern that the agencies may have excluded the 
fuel costs and added emissions from additional miles driven in the 
course of the more frequent refueling events that would be required 
with more lenient CAFE standards, and correspondingly lower on-road 
fuel economy.\1864\ In the NPRM, the agencies asserted that these added 
costs are reflected in their overall estimates of fuel cost savings, 
while any increase in emissions is also reflected in the reported 
changes in total emissions. However, IPI noted that the agencies did 
not clearly explain how these cost savings and emissions reductions are 
actually accounted for in their methodology.
---------------------------------------------------------------------------

    \1864\ IPI, Appendix, NHTSA-2018-0067-12213, at 55.
---------------------------------------------------------------------------

    The agencies' methodology fully accounts for both of these impacts 
through its calculation of changes in the use of new cars and light 
trucks due to the fuel economy rebound effect, which captures the 
impact on their aggregate use (VMT) that results from changes in the 
fuel cost of driving each mile. Studies that estimate the rebound 
effect analyze the relationship between VMT per time period and fuel 
economy or per-mile fuel costs, using data for individual vehicles, 
fleet-wide average values, or aggregate estimates for an entire fleet. 
Regardless of the level of aggregation they employ, their measures of 
vehicle use invariably include travel for all purposes, including any 
extra miles driven in the course of refueling.
    Thus, the estimates of the rebound effect--the response of vehicle 
use to changes in fuel economy or per-mile fuel costs--inevitably 
capture any change in the number of miles driven for the purpose of 
refueling that occurs in response to higher or lower fuel economy. This 
change reflects the net effect of more or less frequent refueling trips 
required by their baseline or ``pre-rebound'' level of use, and any 
change in the number of refueling trips associated with increased or 
reduced driving in response to the rebound effect.
    As a consequence, the agencies' estimates of changes in aggregate 
fuel consumption and fuel costs incorporate--that is, are net of--the 
volume and cost of fuel consumed by changes in vehicle use that result 
from the rebound effect, including any change in driving associated 
with more or less frequent refueling. Similarly, the agencies' 
estimates of changes in emissions resulting from vehicle storage and 
use (referred to as ``tailpipe'' or ``downstream'' emissions) are 
derived by applying per-mile emission factors to changes in aggregate 
vehicle travel, so they necessarily incorporate changes in vehicle use 
for all purposes, including more or less frequent refueling.
    Furthermore, as the agencies demonstrated in the proposal with a 
practical example, the benefit associated with fewer miles spent 
refueling is less than 23[cent] per year for new vehicles. The 
cumulative impact of this benefit amounts to less than one tenth of 
percent of the costs of the rule.\1865\
---------------------------------------------------------------------------

    \1865\ See 83 FR at 43088. Also, note that the 23 cents estimate 
was derived for a less stringent alternative than today's standards 
and included taxes which would have been removed had the agencies 
calculated this number separately.
---------------------------------------------------------------------------

    Because all of the alternative standards evaluated in this 
rulemaking would permit lower fuel economy levels than under the 
baseline standard, per-mile driving costs would be higher and total 
vehicle use would decline in response. Although some (perhaps most) new 
vehicles would require more frequent refueling, the agencies' estimates 
of the change in aggregate use of new vehicles reflects (i.e., is net 
of) any increase in driving associated with more frequent refueling 
stops. As a result, the agencies' estimates of changes in total fuel 
consumption, aggregate fuel costs, and emissions resulting from the 
lower fuel economy levels that relaxing CAFE standards would permit 
reflect the net reduction in use of new cars and light trucks due to 
the fuel economy rebound effect, after considering any additional miles 
that would be driven in the course of more frequent refueling stops.
(c) Including the Legacy Fleet
    Under more stringent regulatory alternatives, more miles will be 
driven by older and less efficient vehicles, and the effect is to 
reduce or eliminate any refueling benefit from increasing the fuel 
efficiency of new vehicles. Failing to include the existing fleet makes 
the costs of refueling artificially lower under more stringent 
standards because new vehicle sales are lower and not only because new 
vehicles are more efficient. This update to the calculation of the 
absolute refueling costs corrects this oversight present in the NPRM 
cost-benefit analysis by calculating fleet-wide absolute refueling 
costs before considering the incremental change relative to the 
baseline.
    For other portions of the CAFE model, the agencies track the legacy 
vehicles by body style and vintage, using average measures for fuel 
economy, horsepower and curb weight. To estimate refueling costs for 
these vehicles, measures of average fuel tank sizes by body style and 
vintage are needed. The agencies are unaware of any data that directly 
estimates this value, but an estimate can be derived from publicly 
available data on fuel tank sizes of 17 high-volume nameplates with 
long histories. The tank sizes are averaged by body style, and these 
historical values are used as estimates of the average by body style 
and vintage. The vehicles included, their fuel tank sizes, and the 
averages are reported in Table VI-195 through Table VI-197 for cars, 
vans/SUVs, and pickups, respectively. The averages are used to 
represent the fuel tank sizes by vintage and vehicle body style. The 
agencies used the fuel tank sizes from Table VI-195 to Table VI-196 to 
determine the number of refueling events and time spent refueling to

[[Page 24716]]

compute refueling costs using the methodology described above.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.396


[[Page 24717]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.397


[[Page 24718]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.398


[[Page 24719]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.399


[[Page 24720]]


BILLING CODE 4910-59-C
(d) Including Electric Vehicle Recharging
    In addition to adding the refueling costs associated with the 
``legacy fleet,'' this update adds the cost to recharge electric 
vehicles to the total refueling costs. Excluding the time spent 
recharging ignores a real cost borne by owners of electric vehicles, 
one which was noted by multiple commenters. For example, Ariel Corp. 
and VNG.co LLC commented that, ``EVs require significant changes in 
consumer fueling behavior given the need to park at recharging points 
for long periods of time.'' \1866\
---------------------------------------------------------------------------

    \1866\ Ariel Corp. and VNG.co LLC, Comment, NHTSA-2018-0067-
7573, at 13.
---------------------------------------------------------------------------

    In order to do so, it is important to first understand how many 
electric vehicle charging events will require the driver to wait and 
for how long. The answer to this question depends on the range of the 
electric vehicle and the length of the trip.\1867\ For trips shorter 
than the range, the driver can recharge the vehicle at times that will 
not require them to be actively waiting and thus there is no recharging 
cost. Only for trips where the vehicle is driven more miles than the 
range will the driver have to stop at mid-trip, a time that is assumed 
to be inconvenient, to recharge the vehicle at least enough to reach 
the intended destination.
---------------------------------------------------------------------------

    \1867\ While the range of EVs is dependent on a number of 
factors, such as that grade, acceleration, and weather, the agencies 
take a conservative approach and assume a best-case scenario.
---------------------------------------------------------------------------

    The agencies use trip data from the National Household 
Transportation Survey (NHTS) to estimate the frequency and expected 
length of trips that exceed the range of the electric vehicle 
technologies in the simulation (200 and 300 mile ranges).
    The NHTS data is collected from a representative random sample of 
U.S. households. The survey collects data on individual trips by mode 
of transportation. A trip is defined by the starting and ending point 
for any personal travel, so that vehicle trips will capture any time a 
car is driven. The survey includes identification numbers for 
households, individuals, and vehicles, and mode of transportation 
(including the body style of the vehicle for vehicle trips), and the 
date of the trip. Although some trips made in the same day may allow 
for convenient charging in between trips, the agencies assume that 
travel in the same day exceeding the range will involve the driver 
waiting for the vehicle to charge. Thus, the total number of miles 
driven by the same vehicle in a single day is summed, and it is assumed 
that charging stations are not conveniently available to the driver in 
between.
    Some of the trips in the NHTS have missing information about the 
duration or length of the trip; these trips are excluded from the 
dataset. The agencies subset the dataset into three body styles--cars, 
vans/SUVs, and pickups--consistent groupings with how the VMT schedules 
and scrappage rates are estimated. The agencies exclude data on taxis 
and rental cars as the body style of the vehicle for these trips is not 
specified (they make up only 0.3 percent of the dataset, so their 
exclusion is unlikely to alter the estimate). Table VI-198, below, 
shows the resulting quantiles of the distribution of daily travel for 
all vehicles considered in the final dataset. This will include 
multiple days of travel for the same vehicle if more than one day of 
trip data is recorded in the NHTS.
[GRAPHIC] [TIFF OMITTED] TR30AP20.400

[GRAPHIC] [TIFF OMITTED] TR30AP20.401

    The data in Table VI-198 shows that excluding taxis and rentals may 
be the best choice even if their body styles were known. For taxi 
trips, only the number of trips an individual driver makes in a day is 
known. The number of trips that the taxi cab itself makes in a day is 
unknown. As can be seen, the distribution of ``daily'' travel is to the 
left for taxis because not all trips for those vehicles are reported. 
Thus, including these vehicles would incorrectly skew the daily travel 
rates downwards.
    The distribution of trip lengths for rental cars, on the other 
hand, is generally to the right of trips taken privately-owned 
vehicles. This is likely because individuals are travelling longer 
distances when they are on vacation or otherwise out-of-town. It seems 
likely that individuals renting cars for longer trips will not choose 
electric vehicles for such temporary travel. Thus,

[[Page 24721]]

including these trips in the dataset would likely overestimate the 
number of mid-trip charging events necessary for ordinary travel in a 
way that will not match what actually occurs.
    From the final body style datasets, the agencies are able to 
calculate two measures that allow for the construction of the value of 
recharging time. First, the expected distance between trips that exceed 
the range of 200-mile and 300-mile BEVs (BEV200 and BEV300, 
respectively) is calculated. This is calculated as the quotient of the 
sum of total miles driven by each individual body style and the total 
number of trips exceeding the range, as shown below:
[GRAPHIC] [TIFF OMITTED] TR30AP20.402

    This calculates the expected frequency of enroute recharging 
events, or the amount\1868\ of miles traveled per inconvenient 
recharging event. This is used later used to calculate the total 
expected time to recharge a vehicle.
---------------------------------------------------------------------------

    \1868\ The denominator counts the number of incontinent 
recharging events by body style. It is not a measurment of VMT.
---------------------------------------------------------------------------

    The second measure needed to calculate the total expected 
recharging time is the expected share of miles driven that will be 
charged in the middle of a trip (causing the driver to wait and lose 
the value of time). In order to calculate this measure the difference 
of the trip length and range is summed, conditional on the trip length 
exceeding the range for each body style. This figure is then divided by 
the sum of the length of all trips for that body style. See the 
equation below:
[GRAPHIC] [TIFF OMITTED] TR30AP20.403

    The calculated frequency of inconvenient charging events and share 
of miles driven that require the driver to wait for BEV's with 200 and 
300-mile ranges are presented in Table VI-199, below. As the table 
shows, cars are expected to require less frequent inconvenient charges 
and a smaller share of miles driven will require the driver to charge 
the vehicle in the middle of a trip. Pickups and vans/SUVs have fairly 
similar measures, with vans and SUVs requiring slightly more 
inconvenient charging than pickups.
[GRAPHIC] [TIFF OMITTED] TR30AP20.404

    The measures presented in Table VI-199, above, can be used to 
calculate the expected time drivers of electric vehicles of a given 
body style and range will spend recharging at a time that will require 
them to wait. First the agencies calculate the expected number of 
refueling events for a vehicle of a given style and range in a given 
calendar year. This is shown below as the expected miles driven by a 
vehicle in a given calendar year divided by the charge frequency of a 
vehicle of that style and range (from Table VI-199).

[[Page 24722]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.405

    Next the agencies calculate the number of miles charged for a 
vehicle of a given style\1869\ and range in a specific calendar year. 
This is the product of the number of miles driven by the vehicle and 
the share of miles driven that require an inconvenient charge for a 
vehicle of that style and range (from Table VI-199), as presented 
below:
---------------------------------------------------------------------------

    \1869\ Note that [Sigma]Trip [epsi] Style Trip Length and Miles 
CY,Veh are different values. MilesCY,Veh is the estimated amount of 
VMT predicted by VMT while [Sigma]Trip [epsi] Style Trip Length is 
the sum of trips observed by the NHTS study.
---------------------------------------------------------------------------

    Then, the expected time that a driver of an electric vehicle of a 
given style and range will spend waiting for the vehicle to charge is 
calculated. This is the product of the fixed amount of time it takes to 
get to the charging station and the number of recharging events plus 
the quotient of the expected miles that will require inconvenient 
charging over an input assumption of the rate of which a vehicle of 
that style and range can be charged in a given calendar year (expressed 
in units of miles charged per hour). The fixed amount of time it takes 
to get to a charging station is set equal to the average time it takes 
for an ICE vehicle to get to a gas station for a refueling event, as 
discussed above.\1870\ This is shown below:
---------------------------------------------------------------------------

    \1870\ The agencies note that this is a conservative estimate. 
Gas stations vastly outnumber publicly available recharging stations 
and are often in more convenient locations.
[GRAPHIC] [TIFF OMITTED] TR30AP20.406

    The expected time that a driver will wait for their vehicle to 
charge can then be multiplied by the value of time estimate, as is done 
with gasoline, diesel, and E85 vehicles (see description above of the 
current approach to accounting for refueling time costs).
    It is worth a final note to talk about how plug-in hybrids are 
treated in the modelling (which remains unchanged from the NPRM). 
Presumably, plug-in hybrids that are taken on a trip that exceeds their 
electric range will be driven on gasoline and the driver will recharge 
the battery at a time that is convenient. For this reason, the electric 
portion of travel should be excluded from the refueling time 
calculation. The gasoline portion of travel is treated the same as 
other gasoline vehicles so that when the tank reaches some threshold, 
the vehicles is assumed to be refueled with the same fixed event time 
and the same rate of refueling flow.
    The NPRM calculation of refueling benefits did not account for the 
impacts of fleet turnover--specifically the impact on ``legacy'' fleet 
vehicles and new electric vehicles. However, when the quantities of 
vehicles on the road varies between scenarios it becomes important to 
calculate the refueling costs for all vehicles since fuel economy and 
tank sizes (and therefore range before refueling) vary with vintage. 
This updated analysis adds these elements to the calculation of the 
refueling time and costs and is thus a more accurate estimation of the 
refueling benefit.
(12) Energy Security
    By amending existing standards, the final rule is expected to 
increase domestic consumption of gasoline by a relatively minimal 
amount relative to the baseline standards finalized in 2012, producing 
a correspondingly small increase in the Nation's demand for crude 
petroleum, a commodity that is traded actively in a worldwide market. 
Specifically, the agencies project that this rule will increase 
gasoline consumption by cars and light trucks produced during model 
years 1978 through 2029 by 3.1 percent.\1871\ Although the U.S. 
accounts for a sufficient (albeit diminishing) share of global oil 
consumption that the resulting increase in global petroleum demand will 
exert some upward pressure on worldwide prices, the rule is projected 
to increase global petroleum demand by less than one half of one 
percent from 2017 through 2050, so its effects on global prices is 
likely to be minimal.
---------------------------------------------------------------------------

    \1871\ This includes fuel consumed by cars and light trucks 
produced during model years 1978-2017 that are on the road today 
during their remaining lifetimes, as well as fuel consumed by cars 
and light trucks projected to be manufactured during model years 
2018-2029 over their entire lifetimes.
---------------------------------------------------------------------------

    U.S. consumption and imports of petroleum products has three 
potential effects on the domestic economy that are often referred to 
collectively as ``energy security externalities,'' and increases in 
their magnitude are sometimes cited as possible social costs of 
increased U.S. demand for petroleum.m First, any increase in global 
petroleum prices that results from higher U.S. gasoline demand will 
cause a transfer of revenue to oil producers worldwide from consumers 
of petroleum, because consumers throughout the world are ultimately 
subject to the higher global price that results. Although this transfer 
is simply a shift of resources that produces no change in global 
economic welfare, the financial drain it produces on the U.S. economy 
is sometimes cited as an external cost of increased U.S. petroleum 
consumption, because consumers of petroleum products are unlikely to 
consider it.
    As the U.S. approaches self-sufficiency in petroleum production 
(the nation is expected to become a net exporter of petroleum by 2020), 
this transfer is increasingly from U.S. consumers of refined petroleum 
products to U.S. petroleum producers, so it not only leaves welfare 
unaffected, but even ceases to be a financial burden on the U.S. 
economy.\1872\ In fact, as the U.S. becomes a net petroleum exporter, 
any transfer from global consumers to petroleum producers would become 
a financial benefit to the U.S. economy. Nevertheless, uncertainty in 
the nation's long-term import-export balance makes it difficult to 
project precisely how these effects might change in response to 
increased consumption.
---------------------------------------------------------------------------

    \1872\ The United States became a net exporter of oil on a 
weekly basis several times in late 2019, and EIA's AEO 2019 projects 
that will do so on a sustained, long-term basis by 2020; see EIA, 
AEO 2019 Reference Case, Table 21 https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=wttntus2&f=4.
---------------------------------------------------------------------------

    Higher U.S. petroleum consumption can also increase domestic 
consumers' exposure to oil price shocks and thus

[[Page 24723]]

increase potential costs to all U.S. petroleum users (including those 
outside the light duty vehicle sector, whose consumption would be 
unaffected by today's final rule) from possible interruptions in the 
global supply of petroleum or rapid increases in global oil prices. 
Because users of petroleum products are unlikely to consider the effect 
of their increased purchases on these risks, their economic value is 
often cited as an external cost of increased U.S. consumption. Finally, 
some analysts argue that domestic demand for imported petroleum may 
also influence U.S. military spending; because the increased cost of 
military activities would not be reflected in the price paid at the gas 
pump, this is often alleged to represent a third category of external 
costs form increased U.S. petroleum consumption.
    Each of these three costs could rise incrementally--albeit by a 
very limited magnitude--as a consequence of increases in U.S. petroleum 
consumption--likely to result from the final rule. This section 
describes the extent to which each cost is expected to increase as a 
result of this action, whether it represents a significant economic 
cost (or simply a transfer of resources), and how the agencies have 
measured each cost and incorporated it into their analysis.
(a) U.S. Petroleum Demand and Its Effect on Global Prices
    Figure VI-79 illustrates the effect of the increase in U.S. fuel 
and petroleum demand anticipated to result from reducing CAFE and 
CO2 standards on global demand for petroleum and its market 
price. The marginal increase in domestic demand can be represented as 
an outward shift in the U.S. demand curve for petroleum from its 
position at DUS,0 with the baseline standards for future 
model years in effect, to DUS,1 with the final rule 
standards replacing them. Because global demand is simply the sum of 
what each nation would purchase at different prices, the outward shift 
in U.S. demand causes an identical shift in the global demand schedule, 
as the figure shows.\1873\
---------------------------------------------------------------------------

    \1873\ The figure exaggerates the U.S. share of total global 
consumption, which currently stands at 20 percent, for purposes of 
illustration.
[GRAPHIC] [TIFF OMITTED] TR30AP20.407

    The global supply curve for petroleum slopes upward, reflecting the 
fact that it is progressively costlier for oil-producing nations to 
explore for, extract, and deliver additional supplies of oil to the 
world market.\1874\ Thus the upward shifts in the U.S. and world demand 
schedules cause an increase in the global price for oil, from 
P0 to P1 in the figure. U.S. purchases of 
petroleum increase from QUS,0 to QUS,1, but the 
resulting increase in global consumption from QG,0 to 
QG,1 will be slightly smaller than the increase in U.S. 
demand and purchases, because the amount of petroleum other nations 
purchase will decline slightly in response to its higher price. 
Spending on petroleum by U.S. buyers who purchase the additional oil 
will increase by the area QUS,0acQUS,1, the 
product of its new, higher price P1 and the increase in U.S. 
consumption, QUS,1-QUS,0, while spending by U.S. 
consumers whose purchases remain unchanged will increase by the product 
of their previous purchases QUS,0 and the price increase 
P1-P0, or the area P1abP0.
---------------------------------------------------------------------------

    \1874\ The figure depicts the relationship between the global 
supply of petroleum and its worldwide price during a single time 
period. The global supply curve for petroleum has been shifting 
outward over time in response to increased investment in 
exploration, the ability of refineries to utilize feedstocks other 
than conventional petroleum, and technological innovations in 
petroleum extraction. The combination of these developments may also 
have reduced its upward slope, meaning that global supply now 
increases by more in response to increases in the world price than 
it once did.
---------------------------------------------------------------------------

    CARB asserted in their comments, that the NPRM analysis was biased

[[Page 24724]]

against the baseline standards because the fuel prices in the NPRM were 
based on a unique run of DOE's NEMS model that included the 
baseline.\1875\ They argued that the proposal would have reduced fleet 
average fuel economy, leading to increased demand and subsequently 
higher fuel prices faced by consumers. As a result, the additional fuel 
costs associated with the proposal (relative to the baseline) should 
have been even higher than estimated because the fuel price faced by 
drivers in that scenario would have been higher than in the baseline. 
However, while the difference between the baseline and preferred 
alternative could create differences in fleet fuel economy in a manner 
that could influence prices at the pump, those differences are likely 
to be small. In response to CARB's comments, the agencies conducted 
additional runs with NEMS to compare the fuel price under the baseline 
standards and the fuel price under the proposed standards. Through 
2050, the fuel price difference between the alternatives was never 
higher than two percent. The standards being finalized in this rule are 
considerably closer to the baseline than were those in the proposal.
---------------------------------------------------------------------------

    \1875\ NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    SAFE commented that the United States is a ``price-taker'' in the 
global market and ``must accept the prevailing global oil price since 
it lacks sufficient market power to influence decisively this price.'' 
\1876\ This comment, however, is directly at odds with both the 
economics of the world oil market shown in Figure VI-79 above and other 
comments asserting that the increase in U.S. gasoline demand resulting 
from this rule will increase U.S. and global petroleum demand, thus 
increasing world oil prices. In response to the comment from SAFE, the 
agencies utilized a forecast of fuel prices in today's analysis that 
considers the effect of the revised standards on global petroleum 
demand and prices. This assumption slightly increases the cost of 
forgone fuel savings in the preferred alternative, compared to their 
value under the assumption that U.S. demand cannot change global prices 
and the nation acts as a price-taker.
---------------------------------------------------------------------------

    \1876\ NHTSA-2018-0067-11981.
---------------------------------------------------------------------------

    In Figure VI-79, the increase in the price of oil from 
P0 to P1 will mean that global consumers who 
previously purchased the quantity of oil QG,0 at its lower 
price will now pay more for that same amount. Specifically, previous 
purchasers will pay the additional area P1deP0, 
whose value is the increase in price P1-P0 
multiplied by the volume they originally bought, QG,0. Of 
this increase in revenue to oil producers, the rectangular area 
P1abP0--which as indicated above is the product 
of the increase in price P1-P0 and previous U.S. 
purchases QUS,0, and thus measures the increase in spending 
by previous U.S. consumers--is simply transferred from U.S. consumers 
to global oil suppliers.\1877\ The remaining fraction of increased 
payments to producers, the rectangular area adeb, whose value is the 
product of the price increase P1-P0 and previous 
purchases by other nations, which were QG,0-
QUS,0, is a transfer from consumers outside the U.S. to 
global oil producers.
---------------------------------------------------------------------------

    \1877\ Note that global oil suppliers include domestic as well 
as US-owned foreign suppliers.
---------------------------------------------------------------------------

    The total increase in global spending--including the additional 
spending by U.S. consumers as well as by those in other nations--on the 
amount of oil they previously purchased is simply a transfer of revenue 
from consumers of petroleum products to oil producers. This transfer 
can be described as a ``pecuniary'' externality, since it describes the 
effect of the price increase on wealth allocation, but is considered 
separately from any effects on quantity produced and consumed. Some of 
the increase in payments by U.S. consumers for the petroleum products 
they originally consumed may be made to foreign-owned oil producers, 
and thus represents a financial drain on the U.S. economy, while the 
remainder is received by domestic producers and thus remains within the 
U.S. economy.\1878\
---------------------------------------------------------------------------

    \1878\ Neither transfer, however, has an effect on domestic or 
global economic welfare.
---------------------------------------------------------------------------

    To an increasing extent, however, the additional payments by U.S. 
consumers that result from upward pressure on the world oil price are a 
transfer entirely within the Nation's economy, because a growing 
fraction of domestic petroleum consumption is supplied by U.S. 
producers. The U.S. is projected to become a net exporter of petroleum 
in 2020--and in fact became a net exporter in September 2019--and as 
the Nation moves toward that status, an increasing share of any higher 
costs paid by U.S. consumers of petroleum products becomes a gain to 
U.S. oil producers.\1879\ When the U.S. becomes self-sufficient in 
petroleum supply--which is now anticipated to occur in the year this 
final rule publishes--the entire value of increased payments by U.S. 
petroleum users that results from relaxing CAFE and CO2 
standards will have the same effect as if it were simply a transfer 
within the U.S. economy. As a consequence, the financial burden that 
transfers from U.S. consumers to foreign producers places on the U.S. 
economy will disappear.
---------------------------------------------------------------------------

    \1879\ The U.S. Energy Information Administration EIA estimates 
that the United States exported more total crude oil and petroleum 
products in September and October of 2019, and expects the United 
States to continue to be a net exporter. See Short Term Energy 
Outlook November 2019, available at https://www.eia.gov/outlooks/steo/archives/nov19.pdf.
---------------------------------------------------------------------------

    Over almost the entire time period spanned by the analysis of this 
final rule, any increase in domestic spending for petroleum caused by 
the effect of higher U.S. fuel consumption and petroleum use on world 
oil prices is expected on balance to be a transfer within the U.S. 
economy and thus produce no drain on domestic economic resources. For 
this reason--and because in any case such transfers do not create real 
economic costs or benefits--increased U.S. spending on petroleum 
products that results from increased U.S. fuel demand and any resulting 
upward pressure on petroleum prices stemming from this action is not 
included among the economic costs accounted for in this final rule.
(b) Macroeconomic Costs of U.S. Petroleum Consumption
    In addition to influencing global demand and prices, U.S. petroleum 
consumption imposes further costs that are unlikely to be reflected in 
the market price for petroleum, or in the prices paid by consumers of 
refined products such as gasoline.\1880\ Petroleum consumption imposes 
external economic costs by exposing the U.S. economy to increased risks 
of rapid increases in prices triggered by global events that may also 
disrupt the supply of imported oil, and U.S. consumers of petroleum 
products are unlikely to take such costs into account when making their 
decisions about how much to consume.
---------------------------------------------------------------------------

    \1880\ See, e.g., Bohi, D.R. & W. David Montgomery (1982), Oil 
Prices, Energy Security, and Import Policy Washington, DC--Resources 
for the Future, Johns Hopkins University Press; Bohi, D.R., & M.A. 
Toman (1993), ``Energy and Security--Externalities and Policies,'' 
Energy Policy 21:1093-1109; and Toman, M.A. (1993). ``The Economics 
of Energy Security--Theory, Evidence, Policy,'' in A. V. Kneese and 
J.L. Sweeney, eds. (1993), Handbook of Natural Resource and Energy 
Economics, Vol. III, Amsterdam--North-Holland, pp. 1167-1218.
---------------------------------------------------------------------------

    Sudden interruptions in oil supply and rapid increases in its price 
can impose significant economic costs, because they raise the costs of 
producing all commodities whose manufacturing and distribution consumes 
petroleum, thus temporarily reducing the level of output that the U.S. 
economy can produce using its available supplies of labor and capital. 
The magnitude of any reduction in

[[Page 24725]]

economic output depends on the extent and duration of the increases in 
prices for petroleum products that result from a disruption in global 
oil supplies, as well as on whether and how rapidly prices return to 
their pre-disruption levels--which in turn depends largely on the rest 
of the world's capability to respond to interruptions by increasing 
production elsewhere. Even if prices for oil return completely to their 
original levels, however, economic output will be at least temporarily 
reduced from the level that would have been possible with uninterrupted 
oil supplies and stable prices, so the U.S. economy will bear some 
transient losses it cannot subsequently recover.
    Supply disruptions and price increases caused by global political 
events tend to occur suddenly and unexpectedly, so they can also force 
businesses and households to adjust their use of petroleum products 
more rapidly than if the same price increase occurred gradually. Rapid 
substitutions between energy derived from oil and other forms of 
energy, as well as between energy and other inputs, and other changes 
such as adjusting production levels and downstream prices, can be 
costly for businesses to make. As with businesses, sudden changes in 
energy prices and use are also difficult for households to adapt to 
quickly or smoothly, and doing so may impose at least temporary costs 
or losses in utility for the various adjustments they make.
    Interruptions in oil supplies and sudden increases in petroleum 
prices are both uncertain prospects, and the costs of the disruptions 
they can cause must be weighted or adjusted by the probability that 
they will occur, as well as for their uncertain duration. The agencies 
estimate this expected cost of such disruptions by combining the 
probabilities that price increases of different magnitudes and 
durations will occur during the future period spanned by their analysis 
with the costs of reduced U.S. economic output and abrupt adjustments 
to sharply higher petroleum prices. Any change in the probabilistic 
``expected value'' of such costs that can be traced to higher U.S. fuel 
consumption and petroleum demand stemming from this final rule to 
establish less demanding fuel economy standards is considered to be an 
external cost of the adopting it.
    A variety of mechanisms exist to ``insure'' against higher 
petroleum prices and reduce their costs for adjusting to sudden price 
increases, including making purchases or sales in oil futures markets, 
adopting energy conservation measures, diversifying the fuel economy 
levels within the set of vehicles owned by the household, locating 
where public transit provides a viable alternative to driving, and 
installing technologies that permit rapid fuel switching. Growing 
reliance on such measures, coupled with continued improvements in 
energy efficiency throughout the economy, has certainly reduced the 
vulnerability of the U.S. economy to the costs of oil shocks in recent 
decades.
    Thus, there is now considerable debate about the magnitude and 
continued relevance of potential economic damages from sudden increases 
in petroleum prices. The petroleum intensity of the U.S economy has 
declined considerably and global oil prices are dramatically lower than 
when analysts first identified and quantified the risks they create to 
the U.S. economy. Further, not only has the Nation dramatically 
increased its own petroleum supply, but other new global supplies have 
emerged as well, both of which reduce the potential impact of 
disruptions that occur in unstable or vulnerable regions where oil is 
produced.
    As a consequence, the potential macroeconomic costs of sudden 
increases in oil prices are now likely to be considerably smaller than 
when they were original identified and estimated. Research by the 
National Research Council (2009) argued that non-environmental 
externalities associated with dependence on foreign oil are small, and 
perhaps trivial.\1881\ Research by Nordhaus and by Blanchard and Gali 
have also questioned how harmful to the economy oil price shocks have 
been, noting that the U.S. economy actually expanded immediately after 
the most recent oil price shocks, and that there was little evidence of 
higher energy prices being passed through to higher wages or 
prices.\1882\
---------------------------------------------------------------------------

    \1881\ National Research Council, Hidden Costs of Energy--
Unpriced Consequences of Energy Production and Use, National Academy 
of Sciences, Washington, DC (2009).
    \1882\ Nordhaus argues that one reason for limited vulnerability 
to oil price shocks is that monetary policy has become more 
accommodating to the price impacts, while another is that U.S. 
consumers and businesses may determine that such movements are 
temporary and abstain from passing them on as inflationary price 
increases in other parts of the economy. He also notes that changes 
in productivity in response to recent oil price increases are have 
been extremely modest, observing that ``energy-price changes have no 
effect on multifactor productivity and very little effect on labor 
productivity.'' at p. 19. Blanchard and Gali (2010) contend that 
improvements in monetary policy, more flexible labor markets, and 
the declining energy intensity of the U.S. economy (combined with an 
absence of concurrent shocks to the economy from other sources) 
lessened the impact of oil price shocks after 1980. They find that 
``the effects of oil price shocks have changed over time, with 
steadily smaller effects on prices and wages, as well as on output 
and employment . . . The message . . . is thus optimistic in that it 
suggests a transformation in U.S. institutions has inoculated the 
economy against the responses that we saw in the past.'' at p. 414; 
See William Nordhaus, ``Who's Afraid of a Big Bad Oil Shock?'' 
Available at http://aida.econ.yale.edu/~nordhaus/homepage/
Big_Bad_Oil_Shock_Meeting.pdf; and Blanchard, Olivier and Jordi 
Gali, J., ``The Macroeconomic Effects of Oil price Shocks--Why are 
the 2000s so Different from the 1970s?,'' in Gali, Jordi and Mark 
Gertler, M., eds., The International Dimensions of Monetary Policy, 
University of Chicago Press, February (2010), pp. 373-421, available 
at http://www.nber.org/chapters/c0517.pdf.
---------------------------------------------------------------------------

    Since these studies were issued in 2009 and 2010, the petroleum 
intensity of the U.S. economy has continued to decline while domestic 
energy production has increased in ways and to an extent that experts 
failed to predict, so that the U.S. became the world's largest producer 
in 2018.\1883\ The U.S. shale oil revolution has both established the 
potential for energy independence and placed downward pressure on 
prices. Lower oil prices are also a result of sustained reductions in 
U.S. consumption and global demand resulting from energy efficiency 
measures, many undertaken in response to previously high oil prices.
---------------------------------------------------------------------------

    \1883\ See U.S. Energy Information Administration EIA, Today in 
Energy August 20, 2019, available at https://www.eia.gov/todayinenergy/detail.php?id=40973; Today in Energy September 12, 
2018, available at https://www.eia.gov/todayinenergy/detail.php?id=37053.
---------------------------------------------------------------------------

    Reduced petroleum intensity and higher U.S. production have 
combined to produce a decline in U.S. petroleum imports--to 
approximately 20 percent of domestic consumption in 2017--which permits 
U.S. supply to act as a buffer against artificial or natural 
restrictions on global petroleum supplies due to military conflicts or 
natural disasters. In addition, the speed and relatively low 
incremental cost with which U.S. oil production has increased suggests 
that both the magnitude and (especially) the duration of future oil 
price shocks may be limited, because U.S. production offers the 
potential for a large and relatively swift supply response.
    And while some risk of price shocks certainly still exists, even 
the potential for a large and swift U.S. production response may be 
playing a role in limiting the extent of price shocks attributable to 
external events. The large-scale attack on Saudi Arabia's Abqaiq 
processing facility--the world's largest crude oil processing and 
stabilization plant--on September 14, 2019 caused ``the largest single-
day [crude oil] price increase in the past decade,'' of between $7 and 
$8 per

[[Page 24726]]

barrel, according to EIA.\1884\ The Abqaiq facility has the capacity to 
process 7 million barrels per day, or about 7 percent of global crude 
oil production capacity. EIA declared, however, that by September 17, 
only three days after the incident:
---------------------------------------------------------------------------

    \1884\ https://www.eia.gov/todayinenergy/detail.php?id=41413.

    Saudi Aramco reported that Abqaiq was producing 2 million 
barrels per day, and they expected its entire output capacity to be 
fully restored by the end of September. In addition, Saudi Aramco 
stated that crude oil exports to customers will continue by drawing 
on existing inventories and offering additional crude oil production 
from other fields. Tanker loading estimates from third-party data 
sources indicate that loadings at two Saudi Arabian export 
facilities were restored to the pre-attack levels. Likely driven by 
news of the expected return of the lost production capacity, both 
Brent and WTI crude oil prices fell on Tuesday, September 17.\1885\
---------------------------------------------------------------------------

    \1885\ Id.

    Thus, the largest single-day oil price increase in the past decade 
was largely resolved within a week, and assuming very roughly that 
average crude oil prices were $70/barrel in September 2019 (slightly 
higher than actual), an increase of $7/barrel would represent a 10 
percent increase as a result of the Abqaiq attack. Contrast this with 
the 1973 Arab oil embargo, which lasted for months and raised prices 
350 percent.\1886\ Saudi Arabia could have experienced increased 
revenue resulting from higher prices following the Abqaiq attack, but 
instead moved rapidly to restore production and tap reserves to control 
the risk of resulting price increases. In doing so, the Saudis likely 
recognized that sustained, long-term price increases would reduce their 
ability to control global supply (and thus prices and their own 
revenues) by relying on their lower cost of production.\1887\
---------------------------------------------------------------------------

    \1886\ See Jeanne Whalen, ``Saudi Arabia's oil troubles don't 
rattle the U.S. as they used to,'' Washington Post, September 19, 
2019, available at https://www.washingtonpost.com/business/2019/09/19/saudi-arabias-oil-troubles-dont-rattle-us-like-they-used/.
    \1887\ See, e.g., ``Dynamic Delivery: America's Evolving Oil and 
Natural Gas Transportation Infrastructure,'' National Petroleum 
Council (2019) at 18, available at: https://dynamicdelivery.npc.org/downloads.php.
---------------------------------------------------------------------------

    Some commenters asserted that U.S. shale oil resources cannot serve 
as ``swing supply'' to provide stability in the face of a sudden, 
significant global supply disruption (Jason Bordoff, 
SAFE).1888 1889 Despite its greater responsiveness to price 
changes, commenters argued that lead time to bring new shale resources 
to market (6-12 months) is inferior to ``true spare capacity'' (like 
Saudi Arabia's large oil fields) because it cannot be deployed quickly 
enough to mitigate the economic consequences resulting from rapidly 
rising oil prices. Bordoff, however, also notes that shale oil 
projects' lead times are still shorter--and possibly much shorter--than 
conventional oil resource development. So, while new U.S. oil resources 
may take some time to respond to supply disruptions, they are 
nevertheless likely to provide a stabilizing influence on supply.
---------------------------------------------------------------------------

    \1888\ NHTSA-2018-0067-11981.
    \1889\ NHTSA-2018-0067-10718.
---------------------------------------------------------------------------

    This is especially true for price increases that occur more slowly. 
When Beccue and Huntington updated their 2005 estimates of supply 
disruption probabilities in 2016,\1890\ they found that the probability 
distribution was generally flatter--suggesting that supply disruptions 
of most potential magnitudes were less likely to occur under today's 
market conditions than they had estimated previously in 2005. In 
particular, Beccue and Huntington find that supply disruptions of 
between two and four million barrels per day are significantly less 
likely than their previous estimates suggested. Although their recent 
study also estimated that larger supply disruptions (nine or more 
million barrels per day) are now slightly more likely to occur than in 
previous estimates, disruptions of that magnitude are extremely 
unlikely under either set of estimates.
---------------------------------------------------------------------------

    \1890\ Beccue, Phillip, Huntington, Hillard, G., 2016. An 
Updated Assessment of Oil Market Disruption Risks: Final Report. 
Energy Modeling Forum, Stanford University.
---------------------------------------------------------------------------

    Based on this review of the literature, the agencies concede that 
shale resources may not be able to stabilize oil markets fully to 
prevent a price increase associated with a large supply disruption 
elsewhere in the world. However, if supply disruptions are small 
enough, or move slowly enough, U.S. resources may be an adequate 
stabilizer.
    The agencies reviewed further research that emphasizes the 
continued threat to the U.S. economy posed by the potential for sudden 
increases in global petroleum prices.\1891\ For example, Ramey and Vine 
(2010) note ``remarkable stability in the response of aggregate real 
variables to oil shocks once we account for the extra costs imposed on 
the economy in the 1970s by price controls and a complex system of 
entitlements that led to some rationing and shortages.'' \1892\ In 
contrast, another recent study found that while the likely effects of 
sudden oil price increases have become smaller over time, the declining 
sensitivity of petroleum demand to prices means that any future 
disruptions to oil supplies will have larger effects on petroleum 
prices, so that on balance their economic impact is likely to remain 
significant.\1893\
---------------------------------------------------------------------------

    \1891\ Hamilton (2012) reviewed the empirical literature on oil 
shocks and concluded that its findings are mixed, noting that some 
recent research (e.g., Rasmussen and Roitman, 2011) finds either 
less evidence for significant economic effects of oil price shocks 
or declining effects (Blanchard and Gali 2010), while other research 
finds evidence of their continuing economic importance. See 
Hamilton, J. D., ``Oil Prices, Exhaustible Resources, and Economic 
Growth,'' in Handbook of Energy and Climate Change available at 
http://econweb.ucsd.edu/~jhamilto/handbook_climate.pdfhttp://
econweb.ucsd.edu/~jhamilto/handbook_climate.pdf.
    \1892\ Ramey, V. A., & Vine, D. J. ``Oil, Automobiles, and the 
U.S. Economy--How Much have Things Really Changed?'' National Bureau 
of Economic Research Working Paper 16067 (June 2010). Available at 
http://www.nber.org/papers/w16067.pdf.
    \1893\ Baumeister, C. and G. Peersman (2012), ``The role of 
time-varying price elasticities in accounting for volatility changes 
in the crude oil market,'' Journal of Applied Econometrics 28 no. 7, 
November/December 2013, pp.1087-1109.
---------------------------------------------------------------------------

    Some commenters (SAFE, CARB, Fuel Freedom Foundation, IPI) 
expressed skepticism that the United States could become a net 
petroleum exporter in the future without the continuation of the 
baseline standards. They cautioned that the global oil market is 
inherently uncertain, and Bordoff cautioned that America's shale 
resources may not last as long, or be as easy to develop, as they 
currently appear.\1894\ If the U.S. does not become a net exporter of 
petroleum as anticipated, any wealth effects from a high price of oil 
would continue to accrue to foreign owners of oil reserves. In 
addition, several of these commenters (CARB, SAFE, Bordoff, Zozana) 
argued that, regardless of whether or not the U.S. becomes a net 
petroleum exporter, its levels of petroleum consumption make it still 
vulnerable to price shocks arising in the global oil market.
---------------------------------------------------------------------------

    \1894\ NHTSA-2018-0067-10718.
---------------------------------------------------------------------------

    The agencies believe that the United States lacks the power 
(significantly) to control the global oil price and as a consequence 
remains vulnerable to the effects of oil price spikes, regardless of 
our own oil output. Geopolitical factors influence the global oil 
price--unstable regimes are often unreliable suppliers, large suppliers 
attempt strategically to manage supply to influence price or retain 
market share, and international negotiations around politically 
sensitive topics can influence the production behavior of firms in oil-
rich nations. All of these factors, as well as wars and natural 
disasters, can influence the

[[Page 24727]]

global supply and the market price for oil.
    In this analysis, any increase in the expected value of potential 
costs from economy-wide disruptions caused by sudden price increases 
that results from higher U.S. fuel and petroleum demand is accounted 
for separately from the direct cost for increased purchases of 
petroleum products. Consumers of petroleum products are unlikely to 
consider their contributions to these costs when deciding how much 
energy to consume, because those costs will be distributed widely 
throughout the economy, falling largely on businesses and households 
other than those whose decisions impose them. Thus, they represent an 
external (or ``social'') cost that users of petroleum energy such as 
transportation fuel are unlikely to internalize fully, and the agencies 
analysis includes the estimated increase in these costs among of the 
social costs stemming from the final rule. While increased U.S. 
petroleum production may impose some limits on their potential 
magnitude, their underlying source continues to be domestic petroleum 
use rather than imports.
    Although the vulnerability of the U.S. economy to oil price shocks 
depends on aggregate petroleum consumption rather than on the level of 
oil imports, variation in U.S. oil imports may itself have some effect 
on the frequency, size, or duration of sudden oil price increases. The 
expected value of the resulting economic costs would also depend partly 
on the fraction of U.S. petroleum use that is supplied by imports. 
While total U.S. petroleum consumption is the primary determinant of 
potential economic costs to the Nation from rapid increases in oil 
prices, the estimate of these costs that have been relied upon on in 
past regulatory analyses--and in this analysis--is nevertheless 
expressed per unit (barrel) of imported oil. When they are converted to 
a per-gallon basis, they thus apply to fuel that is either imported in 
refined form, or refined domestically from imported crude petroleum.
    Table VI-200 reports the per-barrel estimates of external costs 
from potential oil price shocks this analysis uses to estimate the 
increase in their total value likely to result from this final rule. 
These values differ from those used in previous analysis of CAFE and 
CO2 standards. In their comments on the NPRM, SAFE pointed 
out recent studies that have updated the estimates of the oil security 
premium since the study--on which the agencies relied upon in the 
NPRM--had been published. They depend in part on projected future oil 
prices, the elasticities of consumption with respect to price, income, 
and U.S. GDP. Since the NPRM values were last updated by the agencies, 
all of these factors have evolved in directions that would reduce the 
magnitude of the oil security premium, so continuing to use the NPRM 
values would have overestimated the increase in expected costs to the 
U.S. economy from potential oil price shocks calculated in this 
analysis, perhaps significantly.\1895\
---------------------------------------------------------------------------

    \1895\ The costs reported in Table VI-188 also depend on the 
probabilities or expected frequencies of supply interruptions or 
sudden price shocks of different sizes and durations. The most 
recent reassessment of the probabilities on which these estimates 
are based (which were originally developed in 2005) was conducted in 
2016; see Beccue, Phillip C. and Hillard G. Huntington, An Updated 
Assessment of Oil Market Disruption Risks--Final Report EMF SR 10, 
Stanford University Energy Modeling Forum (February 5, 2016) 
available at https://emf.stanford.edu/publications/emf-sr-10-updated-assessment-oil-market-disruption-risks.
---------------------------------------------------------------------------

    Specifically, the global petroleum prices projected in EIA's Annual 
Energy Outlook 2018 Reference Case range from 33-57 percent below those 
used to develop the estimates used in the NPRM and reported in Table 
VI-200. U.S. petroleum consumption and imports are now projected to be 
3-8 percent and 20-27 percent lower than the forecast values used to 
construct the NPRM estimates in the table. Finally, total petroleum 
expenditures are now projected to average 1.5-2.4 percent of U.S. GDP, 
in contrast to the 3.8-4.0 percent shares reflected in those values. 
Each of these differences suggests that the values in the NPRM 
overstated the current magnitude of potential costs to the U.S. economy 
from the risk of petroleum price shocks, and together they suggest that 
this overstatement may be significant. Indeed, the values used to 
support this final rule analysis are sourced from a recent paper by 
Brown.\1896\ Brown updates the underlying parameters used to estimate 
the oil security premium and finds a range of $0.60-$3.45 per barrel of 
imported oil, with a mean of $1.26 per barrel. The study, which was 
cited by SAFE, determines that the U.S. is less much less sensitive to 
oil price shocks than earlier estimates imply.\1897\ The values used in 
today's rule reflect that conclusion.
---------------------------------------------------------------------------

    \1896\ See Brown, Stephen P.A., New estimates of the security 
costs of U.S. oil consumption, Energy Policy, Volume 13, 2018, Pages 
171-192.
    \1897\ Another report cited by SAFE, Krupnick, et. al, similarly 
conclude that the macroeconomic cost of oil price shocks has 
diminished and that the oil security premium is lower than the 
majority of the existing literature would suggest. See Krupnick, 
Alan, Morgenstern, Richard, Balke, Nathan, Brown, Stephen P.A., 
Herrera, Ana Maria, and Mohan, Shashank, ``Oil Supply Shocks, US 
Gross Domestic Product, and the Oil Security Premium,'' Resources 
for the Future, November 2017, available at: https://media.rff.org/documents/RFF-Rpt-OilSecurity.pdf (last accessed 01/2020).
    \1898\ In order to convert per-barrel costs into per-gallon 
costs, we make the common assumption (used throughout the analysis) 
that each barrel of petroleum produces 42 gallons of motor gasoline.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24728]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.408


[[Page 24729]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.409

BILLING CODE 4910-59-C
    Because they are expressed per barrel of petroleum that is imported 
(either in already-refined form as gasoline, or as crude petroleum to 
be refined domestically), applying these estimates requires the 
agencies to project of any changes in U.S. petroleum imports that are 
likely to result from the higher level of fuel consumption anticipated 
to occur as a result of this final rule. As discussed in detail in 
Section VI.D.3.c(b)(i) of this final rule, the agencies have elected to 
retain their previous assumptions that 50 percent of any increase in 
fuel consumption attributable to the rule will be accounted for through 
imports in refined form, and that 90 percent of the remaining increase 
would be refined domestically from imported petroleum. As a 
consequence, the oil security premiums shown in Table VI-200 are 
considered to be an external cost associated with 95 percent of the 
increase in gasoline consumption projected to result from this final 
rule.\1899\
---------------------------------------------------------------------------

    \1899\ The 95 percent figure is calculates at 50 percent plus 90 
percent of the remaining 50 percent, or 50 percent plus 45 percent.
---------------------------------------------------------------------------

(c) Potential Effects of Fuel Consumption and Petroleum Imports on U.S. 
Military Spending
    A third potential effect of increasing U.S. demand for petroleum is 
an increase in U.S. military spending to secure the supply of oil 
imports from potentially unstable regions of the world and protect 
against their interruption. If an increase in fuel consumption that 
results from reducing CAFE and CO2 standards lead to higher 
military spending to protect oil supplies, this increase in outlays 
would represent an additional external or social cost of the agencies' 
action. Such costs could also include increased costs to maintain the 
U.S. Strategic Petroleum Reserve (SPR), because it is intended to 
cushion the U.S. economy against disruptions in the supply of imported 
oil or sudden increases in the global price of oil.
    While several commenters argued that current U.S. military 
expenditures are uniquely attributable to securing U.S. supplies of 
petroleum from unstable regions of the globe--the Middle East, in 
particular--should be considered as a cost of this action (CARB, SAFE, 
Zonana), they seemed to confuse those costs with the marginal impact of 
increased oil consumption (relative to the baseline) on U.S. military 
activity and its costs. However, the agencies disagree with commenters 
that incremental changes to domestic consumption of oil for light-duty 
transportation could meaningfully change the scope or scale of the U.S. 
Department of Defense mission in the Persian Gulf region. Instead, they 
side with the Fuel Freedom Foundation, which noted in its comment, 
``[i]ncrementally decreasing petroleum consumption does not 
significantly

[[Page 24730]]

decrease the military spending to protect and ensure its flow around 
the world.'' \1900\
---------------------------------------------------------------------------

    \1900\ NHTSA-2018-0067-12016.
---------------------------------------------------------------------------

    SAFE estimated a per-gallon cost of military externalities 
associated with U.S. dependence on petroleum products, and imported 
petroleum specifically.\1901\ Their low estimate of $0.28/gallon 
assumes $81 billion per year for protection of the global petroleum 
supply and divides those costs by the number of gallons consumed by 
U.S. drivers. In contrast, a similar analysis by Crane et al. stated, 
``our analysis addresses the incremental cost to the defense budget of 
defending the production and transit of oil. It does not argue that a 
partial reduction of the U.S. dependence on imported oil would yield a 
proportional reduction in U.S. spending that is focused on this 
mission. The effect on military cost from such changes in petroleum use 
would be minimal.'' \1902\ The agencies thus do not believe that any 
incremental petroleum consumption that may result from this final rule 
will influence any fraction of U.S. defense spending that can be 
ascribed to protecting the global oil network.
---------------------------------------------------------------------------

    \1901\ NHTSA-2018-0067-11981.
    \1902\ Crane, K., A. Goldthau, M. Toman, T. Light, S.E. Johnson, 
A. Nader, A. Rabasa, & H. Dogo, Imported Oil and U.S. National 
Security, Santa Monica, CA, The RAND Corporation (2009) available at 
https://www.rand.org/pubs/monographs/MG838.html.
---------------------------------------------------------------------------

    Eliminating petroleum imports (to both the U.S. and its national 
security allies) entirely might permit the Nation to scale back its 
military presence in oil-supplying regions of the globe to the extent 
that such interventions are driven by narrow concerns for oil 
production rather than other geopolitical considerations, but there is 
little evidence that U.S. military activity and spending in those 
regions have varied over history in response to fluctuations in the 
Nation's oil imports, or are likely to do so over the future period 
spanned by this analysis. Figure VI-80 shows that military spending as 
a share of total U.S. economic activity has gradually declined over the 
past several decades, and that any temporary--although occasionally 
major--reversals of this longer-term decline have been closely 
associated with U.S. foreign policy initiatives or overseas wars.
[GRAPHIC] [TIFF OMITTED] TR30AP20.410

    Figure VI-81 superimposes U.S. petroleum consumption and imports on 
the history of military spending shown in the previous figure. Doing so 
shows that variation in U.S military spending throughout this period 
has had little association with the historical pattern of domestic 
petroleum purchases, changes in which instead primarily reflected the 
major increases in global petroleum prices that occurred in 1978-79, 
2008, and 2012-13. More important, Figure VI-81 also shows that U.S. 
military spending varied almost completely independently of the 
nation's imports of petroleum over this period. This history suggests 
that U.S. military activities--even in regions of the world that have

[[Page 24731]]

historically represented vital sources of oil imports--serve a far 
broader range of security and foreign policy objectives than simply 
protecting oil supplies. Thus, reducing the nation's consumption or 
imports of petroleum is unlikely by itself to lead to reductions in 
military spending.
    SAFE further argued in its comments that the America's involvement 
in wars in the Persian Gulf region, starting with the first Gulf War 
and continuing through the Iraq War, has been a direct consequence of 
our dependence upon oil. In particular, they state that ``[w]hile there 
is debate over the precise role of oil in America's wars in the greater 
Middle East, several retired military members of SAFE's ESLC and other 
defense budget experts that were consulted for this report believe the 
connection is clear.'' \1903\ However, neither today's action, nor the 
baseline standards, has the ability to change the historical wealth 
transfer that created powerful nations in the Middle East. Attributing 
the cost of the Iraq War, for example, to oil dependence does not 
directly support an assertion that a marginal reduction in oil 
dependence could have reduced the cost of that conflict.
---------------------------------------------------------------------------

    \1903\ NHTSA-2018-0067-11981.
    [GRAPHIC] [TIFF OMITTED] TR30AP20.411
    
    Further, the agencies were unable to find a record of the U.S. 
government attempting to calibrate U.S. military expenditures, force 
levels, or deployments to any measure of the Nation's petroleum use and 
the fraction supplied by imports, or to an assessment of the potential 
economic consequences of hostilities in oil-supplying regions of the 
world that could disrupt the global market.\1904\ Instead, changes in 
U.S. force levels, deployments, and spending in such regions appear to 
have been governed by purposeful foreign policy initiatives,

[[Page 24732]]

unforeseen political events, and emerging security threats, rather than 
by shifts in U.S. oil consumption or imports.\1905\
---------------------------------------------------------------------------

    \1904\ Crane et al. (2009) analyzed reductions in U.S. forces 
and associated cost savings that could be achieved if oil security 
were no longer a consideration in military planning, and disagree 
with this assessment. After reviewing recent allocations of budget 
resources, they concluded that ``. . . the United States does 
include the security of oil supplies and global transit of oil as a 
prominent element in its force planning'' at p. 74 (emphasis added). 
Nevertheless, their detailed analysis of individual budget 
categories estimated that even eliminating the protection of foreign 
oil supplies completely as a military mission would reduce the 
current U.S. defense budget by approximately 12-15 percent. See 
Crane, K., A. Goldthau, M. Toman, T. Light, S.E. Johnson, A. Nader, 
A. Rabasa, & H. Dogo, Imported Oil and U.S. National Security., 
Santa Monica, CA, The RAND Corporation (2009) available at https://www.rand.org/pubs/monographs/MG838.html.
    \1905\ Crane et al. (2009) also acknowledge the difficulty of 
reliably allocating U.S. military spending by specific mission or 
objective, such as protecting foreign oil supplies. Moore et al. 
(1997) conclude that protecting oil supplies cannot be distinguished 
reliably from other strategic objectives of U.S. military activity, 
so that no clearly separable component of military spending to 
protect oil flows can be identified, and its value is likely to be 
near zero. Similarly, the U.S. Council on Foreign Relations (2015) 
takes the view that significant foreign policy missions will remain 
over the foreseeable future even without any imperative to secure 
petroleum imports. A dissenting view is that of Stern (2010), who 
argues that other policy concerns in the Persian Gulf derive from 
U.S. interests in securing oil supplies, or from other nations' 
reactions to U.S. policies that attempt to protect its oil supplies. 
See Crane, K., A. Goldthau, M. Toman, T. Light, SE Johnson, A. 
Nader, A. Rabasa, and H. Dogo, Imported Oil and U.S. National 
Security., Santa Monica, CA, The RAND Corporation (2009) available 
at https://www.rand.org/pubs/monographs/MG838.html; Moore, John L., 
E.J. Carl, C. Behrens, and John E. Blodgett, ``Oil Imports--An 
Overview and Update of Economic and Security Effects,'' 
Congressional Research Service, Environment and Natural Resources 
Policy Division, Report 98, No. 1 (1997), pp. 1-14; Council on 
Foreign Relations, ``Automobile Fuel Economy Standards in a Lower-
Oil-Price World,'' November 2015; and Stern, Roger J. ``United 
States cost of military force projection in the Persian Gulf, 1976-
2007,'' Energy Policy 38, no. 6 (June 2010), pp. 2816-25, https://www.sciencedirect.com/science/article/pii/S0301421510000194?via%3Dihub.
---------------------------------------------------------------------------

    The agencies thus conclude that U.S. military activity and 
expenditures are unlikely to be affected by even relatively large 
changes in consumption of petroleum-derived fuels by light duty 
vehicles. Certainly, the historical record offers no suggestion that 
U.S. military spending is likely to adjust significantly in response to 
the increase in domestic petroleum use that would result from reducing 
CAFE and CO2 standards.
    Nevertheless, it is possible that more detailed analysis of 
military spending might identify some relationship to historical 
variation in U.S. petroleum consumption or imports. A number of studies 
have attempted to isolate the fraction of total U.S. military spending 
that is attributable to protecting overseas oil supplies.\1906\ These 
efforts have produced varying estimates of how much it might be reduced 
if the U.S. no longer had any strategic interest in protecting global 
oil supplies. However, none has identified an estimate of spending that 
is likely to vary incrementally in response to changes in U.S. 
petroleum consumption or imports.
---------------------------------------------------------------------------

    \1906\ These include Copulos, M R. ``America's Achilles Heel--
The Hidden Costs of Imported Oil,'' Alexandria VA--The National 
Defense Council Foundation, September 2003-1-153, available at 
http://ndcf.dyndns.org/ndcf/energy/NDCFHiddenCostsofImported_Oil.pdf; Copulos, M R. ``The Hidden Cost 
of Imported Oil--An Update.'' The National Defense Council 
Foundation (2007) available at http://ndcf.dyndns.org/ndcf/energy/NDCF_Hidden_Cost_2006_summary_paper.pdf; Delucchi, Mark A. & James 
J. Murphy. ``US military expenditures to protect the use of Persian 
Gulf oil for motor vehicles,'' Energy Policy 36, no. 6 (June 2008), 
pp. 2253-64; and National Research Council Committee on Transitions 
to Alternative Vehicles and Fuels, Transitions to Alternative 
Vehicles and Fuels (2013).
---------------------------------------------------------------------------

    Nor have any of these studies tracked changes in spending that can 
be attributed to protecting U.S. interests in foreign oil supplies over 
a prolonged period, so they have been unable to examine whether their 
estimates of such spending vary in response to fluctuations in domestic 
petroleum consumption or imports. The agencies conclude from this 
review of research that U.S. military commitments in the Persian Gulf 
and other oil-producing regions of the world contribute to worldwide 
economic and political stability, and insofar as the costs of these 
commitments are attributable to petroleum use, they are attributable to 
oil consumption throughout the world, rather than simply U.S. oil 
consumption or imports.
    It is thus unlikely that military spending would rise in response 
to any increase in U.S. imports that did result from this final rule. 
As a consequence, the analysis of alternative CAFE and CO2 
emission standards for future model years applies no increase in 
government spending to support U.S. military activities as a potential 
cost of allowing new cars and light trucks to achieve lower fuel 
economy and thus increasing domestic petroleum use.
    Similarly, while the ideal size of the Strategic Petroleum Reserve 
from the standpoint of its potential stabilizing influence on global 
oil prices may be related to the level of U.S. petroleum consumption or 
imports, its actual size has not appeared to vary in response to either 
of those measures. The budgetary costs for maintaining the SPR are thus 
similar to U.S. military spending in that, while they are not reflected 
in the market price for oil (and thus do not enter consumers' decisions 
about how much to use), they do not appear to have varied in response 
to changes in domestic petroleum consumption or imports.
    As a consequence, the analysis does not include any potential 
increase in the cost to maintain a larger SPR among the external or 
social costs of the increase in gasoline and petroleum consumption 
likely to result from reducing future CAFE and CO2 
standards. This view aligns with the conclusions of most recent studies 
of military-related costs to protect U.S. oil imports, which generally 
conclude that savings in military spending are unlikely to result from 
incremental reductions in U.S. consumption of petroleum products on the 
scale of those that would resulting from adopting higher CAFE or 
CO2 standards.
(13) Social Cost of Carbon
    In the proposal, the agencies projected costs resulting from fuel 
consumption and emissions of CO2 using estimates of 
anticipated climate-related economic damages within U.S. borders per 
ton of CO2 emissions, which the agencies referred to as the 
domestic social cost of carbon (domestic SC-CO2). The 
domestic SC-CO2 estimates, which were originally developed 
by EPA for an earlier regulatory analysis, represent the monetary value 
of damages to the domestic economy likely to be caused by future 
changes in the climate that result from incremental increases in 
CO2 emissions during a given year.\1907\ The agencies did 
not consider climate-related damage costs resulting from emissions of 
other greenhouse gases (GHGs), such as methane or nitrous oxide, in 
their analysis supporting the proposal.
---------------------------------------------------------------------------

    \1907\ For a description of the procedures EPA used to develop 
these values, see U.S. Environmental Protection Agency, Regulatory 
Impact Analysis for the Proposed Emission Guidelines for Greenhouse 
Gas Emissions from Existing Electric Utility Generating Units; 
Revisions to Emission Guideline Implementing Regulations; Revisions 
to New Source Review Program, EPA-452/R-18-006, August 2018 (https://www.epa.gov/sites/production/files/2018-08/documents/utilities_ria_proposed_ace_2018-08.pdf), Section 4.3, at 4-2 to 4-7. 
The sources and potential magnitude of uncertainties surrounding the 
SC-CO2 estimates are described in Chapter 7 of that same 
document, at 7-1 to 7-10.
---------------------------------------------------------------------------

    Climate-related damages caused by emissions of CO2 and 
other GHGs include changes in agricultural productivity, adverse 
effects on human health, property damage from increased flood risk, and 
changes in costs for managing indoor environments in commercial and 
residential buildings (such as costs for heating and air conditioning), 
among other possible damages.
    The agencies described the SC-CO2 estimates used in the 
NPRM analysis as interim values developed under Executive Order 13783, 
which are to be used in regulatory analyses until revised values that 
incorporate recommendations from NAS can be developed.\1908\ E.O. 13783 
directed

[[Page 24733]]

agencies to ensure that estimates of the social cost of greenhouse 
gases used in regulatory analyses are consistent with the guidance 
contained in OMB Circular A-4, ``including with respect to the 
consideration of domestic versus international impacts and the 
consideration of appropriate discount rates.'' \1909\
---------------------------------------------------------------------------

    \1908\ The guidance followed by EPA in developing the SC-
CO2 values used in the NPRM analysis appears in President 
of the United States, Executive Order 13783, ``Promoting Energy 
Independence and Economic Growth,'' March 28, 2017, Federal 
Register, Vol. 82, No. 61, Friday, March 31, 2017, 16093-97. 
(https://www.govinfo.gov/content/pkg/FR-2017-03-31/pdf/2017-06576.pdf) The recommendations of the National Academies are 
reported in National Academies of Sciences, Engineering, and 
Medicine, Valuing Climate Damages: Updating Estimation of the Social 
Cost of Carbon Dioxide, Washington, DC, January 2017. Revised values 
incorporating this guidance have not yet been developed.
    https://www.nap.edu/catalog/24651/valuing-climate-damages-updating-estimation-of-the-social-cost-of.
    \1909\ E.O. 13783, at 16096.
---------------------------------------------------------------------------

    Circular A-4 states that analysis of economically significant 
regulations ``should focus on benefits and costs that accrue to 
citizens and residents of the United States,'' and the agencies 
followed this guidance by using estimates of the SC-CO2 that 
included only domestic economic damages. In response to Circular A-4's 
further guidance that regulatory analyses ``should provide estimates of 
net benefits using [discount rates of] both 3 percent and 7 percent,'' 
the agencies presented estimates of the proposed rule's economic 
impacts--including the costs of climate damages likely to result from 
increased CO2 emissions--that incorporated both discount 
rates. The PRIA included a detailed discussion of the analyses used to 
construct estimates of the domestic SC-CO2 using these 
discount rates.\1910\
---------------------------------------------------------------------------

    \1910\ See NHTSA and EPA, PRIA, Chapter 8, Appendix A.
---------------------------------------------------------------------------

    The estimates of the domestic SC-CO2 the agencies used 
in their analysis supporting the proposal increased over future years, 
partly because emissions during future years are anticipated to 
contribute larger incremental costs. Future values of the SC-
CO2 also increase because U.S. GDP is growing over time, and 
many categories of climate-related damage are estimates as proportions 
of GDP. The agencies' estimates of the domestic SC-CO2 for 
emissions occurring in the year 2020 were $1 and $8 (in 2016$) per 
metric ton of CO2 emissions using 7 and 3 percent discount 
rates, and these values were projected to increase to $2 and $10 (again 
in 2016$) by the year 2050.
    As the agencies indicated in the NPRM, the SC-CO2 
estimates are subject to several sources of uncertainty. In accordance 
with guidance provided by OMB Circular A-4 for treating uncertainty in 
regulatory analysis, the PRIA included a detailed discussion of how the 
analysis used to develop the interim SC-CO2 estimates 
incorporated sources of uncertainty that could be quantified. It also 
demonstrated how considering the uncertainty introduced by applying 
discount rates over extended time horizons could affect the estimated 
values.\1911\ To reflect this uncertainty, the analysis supporting the 
proposed rule examined the sensitivity of its estimated costs and 
benefits to using higher values for the SC-CO2 ($9-14 per 
metric ton), which were derived using a lower ``intergenerational'' 
discount rate of 2.5 percent.\1912\
---------------------------------------------------------------------------

    \1911\ See PRIA, Chapter 8, Appendix A.
    \1912\ PRIA, Tables 13-8 and 13-9, at 1547-50.
---------------------------------------------------------------------------

(a) Comments on the NPRM Value for the SC-CO2
    The agencies received extensive comments on the values of the SC-
CO2 used in the NPRM analysis. Broadly, these comments 
stressed the following concerns:
     Using a domestic value for SC-CO2 systemically 
underestimates the benefits of adopting stricter standards.
     The agencies' SC-CO2 omits potential costs due 
to foreign social and political disruptions caused by climate change 
that can affect the U.S.
     The 7 percent discount rate used in the agencies' main or 
central analysis is inappropriate because it represents an opportunity 
cost of capital rather than a rate of time preference for current 
versus future consumption opportunities, and climate change will affect 
future consumption.
 (b) Domestic vs. Global Value for SC-CO2
    Many commenters asserted that it was inappropriate for the agencies 
to use a domestic SC-CO2 value for analyzing benefits or 
costs from changing required levels of fuel economy in the NPRM 
analysis, primarily because doing so could lead regulatory agencies to 
adopt measures that provide inadequate reductions in emissions and 
protection from potential climate change.
    As noted in the NPRM and above, the SC-CO2 estimates the 
agencies used to estimate climate-related economic costs from adopting 
less demanding fuel economy and CO2 emission were developed 
in response to the issuance of E.O. 13783. The agencies remind 
commenters that E.O. 13783 directed federal agencies to ensure that 
estimates of the social cost of greenhouse gases used in their 
regulatory analyses are consistent with the guidance contained in OMB 
Circular A-4, ``including with respect to the consideration of domestic 
versus international impacts and the consideration of appropriate 
discount rates.'' \1913\ Circular A-4 states that analysis of 
economically significant proposed and final regulations ``should focus 
on benefits and costs that accrue to citizens and residents of the 
United States.'' \1914\ The agencies adhered closely to this guidance 
in evaluating the economic costs and benefits in the proposal and this 
final rule by using the domestic value of the SC-CO2 in our 
central analysis.
---------------------------------------------------------------------------

    \1913\ Executive Order 13,783, at 16096.
    \1914\ White House Office of Management and Budget, Circular A-
4: Regulatory Analysis, September 17, 2003, at 15. (https://www.whitehouse.gov/sites/whitehouse.gov/files/omb/circulars/A4/a-4.pdf).
---------------------------------------------------------------------------

    Commenters argued that Circular A-4 allows the agencies to use a 
global SC-CO2 in their central analysis. For example, IPI et 
al. commented that ``Circular A-4's reference to effects `beyond the 
borders' confirms that it is appropriate for agencies to consider the 
global effects of U.S. greenhouse gas emissions.'' \1915\ While the 
agencies agree that Circular A-4 authorizes the agencies to consider 
foreign impacts in certain circumstances, the agencies would also like 
to note that Executive Order 13783 stipulates ``when monetizing the 
value of changes in greenhouse gas emissions resulting from 
regulations, including with respect to the consideration of domestic 
versus international impact [. . .] agencies shall ensure [. . .] any 
such estimates are consistent with the guidance contained in OMB 
Circular A-4.'' \1916\ Using a global SC-CO2 in our central 
analysis would be inconsistent with Circular A-4's directive that any 
non-domestic effects calculated ``should be reported separately.'' 
\1917\ As such, if the agencies had used a global SC-CO2, 
this rulemaking would be compelled by Circular A-4 to separate the SC-
CO2 into domestic and foreign components, and to include 
only the former in our central analysis.
---------------------------------------------------------------------------

    \1915\ IPI et al., DEIS Joint SCC Comments, NHTSA-2017-0069-
0559, at 20.
    \1916\ Executive Order 13,783, at 16096.
    \1917\ Specifically, OMB Circular A-4 directs federal agencies 
as follows: ``Where you choose to evaluate a regulation that is 
likely to have effects beyond the borders of the United States, 
these effects should be reported separately.'' OMB Circular A-4, at 
15.
---------------------------------------------------------------------------

    Furthermore, today's analysis will likely have global impacts 
beyond climate change. For example, freeing manufacturers who compete 
in the U.S. domestic automobile market from burdensome fuel efficiency 
standards may enable them to dedicate time and resources to becoming 
more competitive in global markets, and is thus likely to affect 
product innovation and performance throughout the global auto

[[Page 24734]]

market.\1918\ It would be inconsistent to report the global SC-
CO2 while ignoring other global costs and benefits. The 
agencies do not have a method for analyzing the comprehensive impacts 
of CAFE and CO2 standards--including their many likely 
impacts beyond climate change--on a global scale, and did not receive 
any suggestions about how to conduct such an analysis from commenters. 
Because it would be inconsistent to quantify only climate change and 
none of these other potential global-scale impacts, the agencies have 
decided to focus their attention on domestic impacts, which are more 
readily measurable.
---------------------------------------------------------------------------

    \1918\ Some commenters assert that weakening U.S. fuel economy 
standards could make domestic auto companies less competitive in 
international markets, since several other nations have also adopted 
similar standards. For reasons discussed Section VIII.B.6. of this 
rule, however, the agencies find these comments unpersuasive.
---------------------------------------------------------------------------

    Several commenters argued that the agencies are still obligated to 
report the global impacts of carbon. For example, the North Carolina 
Department of Environmental Quality commented that ``by omitting any 
analysis of the global social cost of carbon, [the agencies] failed to 
adhere to OMB's Circular A-4.'' \1919\ The agencies note Circular A-4 
grants agencies discretion to choose which impacts to report. However, 
to be fully informed of the gamut of potential effects of today's rule, 
the agencies have included two sensitivity cases analyzing the impacts 
of the standards using a global SC-CO2.
---------------------------------------------------------------------------

    \1919\ North Carolina Department of Environmental Quality, 
Comments, NHTSA-2018-0067-12025, at 39.
---------------------------------------------------------------------------

 (c) Scope of Domestic Climate Damages
    Some commenters asserted that even if the agencies are required to 
use a domestic SC-CO2, the specific value employed by the 
agencies underestimated the domestic impacts of climate change. They 
argued the agencies failed to incorporate economic costs associated 
with social or economic disruptions caused by climate change in regions 
of the world that were more vulnerable to its effects, but that could 
``spill over'' to impose damages to the U.S. via their effects on 
migration patterns, international trade flows, or other mechanisms that 
connect nations. Other commenters argued that E.O. 13783 does not 
prohibit the agencies from using the estimates or practices developed 
by the IWG to develop new estimates of the SC-CO2, and 
asserted that the IWG's methods and resulting estimates continue to 
represent the best available practices.
    However, all of the IWG's estimates measure the global SC-
CO2, and as discussed previously, E.O. 13783, in conjunction 
with Circular A-4, directs the agencies to use a domestic SC-
CO2 which precludes the use of the IWG estimates. To develop 
interim estimates of the domestic SC-CO2 that were 
consistent with the IWG's procedures, EPA used the same three climate 
economic models the IWG employed previously to calculate the domestic 
SC-CO2. Two of those three models directly estimate the U.S. 
domestic SC-CO2, which represents the economic costs 
resulting from climate change that are likely to be borne within U.S. 
borders.\1920\ The third model the IWG used previously does not 
estimate the domestic SC-CO2 directly, but EPA approximated 
domestic U.S. costs from future climate change as 10 percent of its 
estimate of their global value, based on results from a companion model 
developed by the same author.\1921\ Thus the agencies believed that the 
SC-CO2 values they used in the NPRM analysis represented the 
most reliable estimates of domestic economic costs from future climate 
change that were available for use in evaluating the proposal.
---------------------------------------------------------------------------

    \1920\ The Policy Analysis of the Greenhouse Effect (PAGE) model 
is described in Hope, C., ``The marginal impact of CO2 
from PAGE2002: An integrated assessment model incorporating the 
IPCC's five reasons for concern,'' The Integrated Assessment 
Journal, Vol. 6 No. 1 (2006), at 19-56; and Hope, C., ``Optimal 
carbon emissions and the social cost of carbon under uncertainty,'' 
The Integrated Assessment Journal Vol. 8, No. 1 (2008), at 107-22. 
The Climate Framework for Uncertainty, Negotiation, and Distribution 
(FUND) model is documented in Tol, Richard, ``Estimates of the 
damage costs of climate change. Part I: benchmark estimates,'' and 
``Estimates of the damage costs of climate change. Part II: Dynamic 
estimates.'' Environmental and Resource Economics Vol 21 (2002), at 
47-73 and 135-60.
    \1921\ The third model is the Dynamic Integrated model of 
Climate and the Economy (DICE), described in Nordhaus, William, 
``Estimates of the Social Cost of Carbon: Concepts and Results from 
the DICE-2013R Model and Alternative Approaches.'' Journal of the 
Association of Environmental and Resource Economists, Vol. 1, No. 2 
(2014), at 273-312 (https://www.jstor.org/stable/pdf/10.1086/676035.pdf). The 10 percent figure is based on the results from a 
regional version of that model (RICE 2010), as described in 
Nordhaus, William D. 2017, ``Revisiting the social cost of carbon,'' 
Proceedings of the National Academy of Sciences of the United 
States, 114 (7), at 1518-23, Table 2. (https://pdfs.semanticscholar.org/f83b/3a7431e0ae2d4e8be3d0ee5f3787a802c34c.pdf?_ga=2.211824467.636056015.1572384992-158339427.1562696454).
---------------------------------------------------------------------------

    The agencies were unable to develop an estimate of the domestic 
value for SC-CO2 that incorporated any of these alleged 
spillover effects, due both to their speculative nature and to the 
absence of credible empirical estimates of their potential magnitude. 
Nor did commenters provide credible explanations for how such 
spillovers might arise, or reliable empirical estimates of their 
potential magnitude.
(d) Discount Rate Used To Construct the SC-CO2 Value
    Many commenters also objected to the agencies use of an SC-
CO2 value that incorporated a 7 percent discount rate in the 
NPRM analysis. Some of these comments reflected a misperception that 
the agencies used such a value in their main or central analysis, when 
in fact it was only used in a sensitivity analysis case as described 
below. Other comments appeared to object to the agencies' use of an SC-
CO2 value incorporating a 7 percent discount rate even as a 
sensitivity case.
    E.O. 13783 directed agencies to ensure that any estimates of the 
social cost of CO2 and other greenhouse gases they used for 
purposes of regulatory analyses are consistent with OMB Circular A-4's 
guidance ``with respect to the consideration of. . .appropriate 
discount rates.'' \1922\ In turn, Circular A-4 refers agencies to OMB's 
earlier guidance on discounting contained in its Circular A-94, noting 
that ``[a]s a default position, OMB Circular A-94 states that a real 
discount rate of 7 percent should be used as a base-case for regulatory 
analysis.'' \1923\ OMB continues to use the 7 percent rate to estimate 
the average pre-tax rate of return to private capital investment 
throughout the U.S. economy. Because it is intended to approximate the 
opportunity cost of capital, it is the appropriate discount rate for 
evaluating the economic consequences of regulations that affect 
private-sector capital investments.
---------------------------------------------------------------------------

    \1922\ E.O. 13,783, at 16096.
    \1923\ OMB Circular A-4, at 33.
---------------------------------------------------------------------------

    At the same time, however, OMB's guidance on discounting also 
recognizes that some federal regulations are more likely to affect 
private consumption decisions made by households and individuals, such 
as when they affect prices or other attributes of consumer goods. In 
these cases, Circular A-4 advises that a lower discount rate is likely 
to be more appropriate, and that a reasonable choice for such a lower 
rate is the real consumer (or social) rate of time preference. This is 
the rate at which individual consumers discount future consumption to 
determine its present value to them.
    OMB estimated that the rate of consumer time preference has 
averaged 3 percent in real or inflation-adjusted terms over an extended 
period, and continues to use that value. In summary, Circular A-4 
reiterates the guidance provided in OMB's earlier Circular A-

[[Page 24735]]

94 that ``[f]or regulatory analysis, you should provide estimates of 
net benefits using both 3 percent and 7 percent.'' \1924\
---------------------------------------------------------------------------

    \1924\ OMB Circular A-4, at 34.
---------------------------------------------------------------------------

    Finally, OMB's guidance on discounting indicates that it may be 
appropriate for government agencies to employ an even lower rate of 
time preference when their regulatory actions entail tradeoffs between 
improving the welfare of current and future generations. Recognizing 
this situation, Circular A-4 advises if the ``rule will have important 
intergenerational benefits or costs [an agency] might consider a 
further sensitivity analysis using a lower but positive discount rate 
in addition to calculating net benefits using discount rates of 3 and 7 
percent.'' \1925\
---------------------------------------------------------------------------

    \1925\ OMB Circular A-4, at 36.
---------------------------------------------------------------------------

    The agencies adhered closely to each of these provisions of OMB's 
guidance on discounting future climate-related economic costs in their 
analysis supporting the NPRM. Specifically, their central analysis 
relied exclusively on a SC-CO2 value that was constructed by 
applying a 3 percent discount rate to future climate-related economic 
damages. This value ranged from $6 per metric ton in 2015 to nearly $11 
per metric ton (both figures in 2016$) by the end of the analysis 
period, the year 2050.
    Throughout the NPRM central analysis, costs resulting from 
increased emissions of CO2 were also discounted from the 
year when those increases in emissions occurred to the present using a 
3 percent rate, even when all other future costs and benefits were 
discounted at a 7 percent rate. Thus the agencies' central analysis for 
the NPRM did not use SC-CO2 values for future years that 
were constructed by applying a 7 percent rate to discount distant 
future climate-related economic damages, and did not use a 7 percent 
rate to discount costs of increased CO2 from the years when 
they were projected to occur to 2018 (the base year used in the 
analysis).
    Notwithstanding concerns raised by commenters about including a 
sensitivity analysis that used a higher discount rate, OMB's guidance 
clearly directs the agencies to report estimates of the present value 
of the economic costs resulting from increased CO2 emissions 
that reflect discount rates of both 3 and 7 percent. Thus to supplement 
their central analysis, which as indicated previously employed a 3 
percent discount rate throughout, the agencies also reported an 
estimate of the economic costs of increased CO2 emissions 
based on a value for the SC-CO2 that was constructed using a 
7 percent discount rate as a sensitivity case, which they termed the 
``Low Social Cost of Carbon'' sensitivity analysis.\1926\ The values 
for the SC-CO2 used in the Low Social Cost of Carbon 
sensitivity analysis varied from $1 per metric ton in 2015 to $3 per 
metric ton (both figures in 2016$) by the end of the analysis period. 
Using these values reduced the loss in total economic benefits 
resulting from the proposed alternative by 1.1 percent, thus increasing 
its net benefits by slightly less than 2 percent.\1927\
---------------------------------------------------------------------------

    \1926\ PRIA, Table 13-1, at 1531-34.
    \1927\ PRIA, Tables 13-8 and 13-9, at 1547-50. Using a lower 
value for the SC-CO2 had opposite effects on the 
proposal's total and net economic benefits, because its net benefits 
represented the difference between the loss in benefits and the 
savings in costs that would result from adopting the proposed rule, 
compared to the baseline of adopting the Augural standards.
---------------------------------------------------------------------------

    For the proposal, the agencies also included a second sensitivity 
analysis using a value for the SC-CO2 that reflected a lower 
``intergenerational'' discount rate of 2.5 percent, which is within the 
1 to 3 percent range for discount rates that have previously been 
applied to economic costs and benefits that span multiple generations, 
as reported in OMB guidance.\1928\ Because using a lower discount rate 
results in a higher value for the SC-CO2, this analysis was 
termed the ``High Social Cost of Carbon'' sensitivity case.\1929\ The 
values for the SC-CO2 used in this additional sensitivity 
analysis varied from $8 per metric ton in 2015 to $14 per metric ton 
(both figures in 2016$) in 2050, the last year of the analysis. Using 
these higher values increased the magnitude of the estimated loss in 
economic benefits resulted from adopting the proposed rule (versus 
retaining the Augural standards) by 0.5 percent from that estimated in 
the central analysis, thus reducing its net benefits by 1.0 
percent.\1930\ Thus it appeared that when used to construct alternative 
estimates of the SC-CO2, the range of discount rates 
specified in OMB Circular A-4 had little or no effect on the estimated 
total benefits of the proposed rule, and the sensitivity analyses 
conducted in support of this Final Rule confirm this result.\1931\
---------------------------------------------------------------------------

    \1928\ OMB Circular A-4, at 36.
    \1929\ PRIA, Table 13-1, at 1531-34.
    \1930\ PRIA, Tables 13-8 and 13-9, at 1547-50. As in the Low 
Social Cost of Carbon sensitivity case, using a higher value for the 
SC-CO2 had opposite effects on the total and net economic 
benefits, because its net benefits were the difference between the 
sacrifice in benefits and the savings in costs from adopting the 
proposed rule, where both were measured against the baseline of 
adopting the Augural standards.
    \1931\ See section VII.B. of this Final Rule for results of the 
``High Social Cost of Carbon'' sensitivity case.
---------------------------------------------------------------------------

(e) SC-CO2 for the Final Rule
    After carefully considering the concerns raised by commenters, the 
agencies decided to leave the SC-CO2 values unchanged for 
the final rule. This means the SC-CO2 estimate used in this 
analysis is still a domestic value that was constructed using a 3 
percent discount rate, and that costs from increased CO2 
emissions are discounted from the year those emissions occur to the 
present using a 3 percent rate. The agencies have again included ``High 
Social Cost of Carbon'' and ``Low Social Cost of Carbon'' sensitivity 
analyses, which continue to use domestic SC-CO2 values that 
incorporate alternative discount rates of 2.5 percent and 7 percent.
    The agencies have also added two sensitivity cases using global 
values for the SC-CO2, which reflect discount rates of 3 
percent and 7 percent. Finally, the agencies have also included an 
additional sensitivity case that incorporates estimates of the domestic 
climate damage costs caused by emissions of the GHGs methane 
(CH4) and nitrous oxide (N2O). Like the SC-
CO2 values used in this analysis, the estimates of the 
domestic values for SC-CH4 and SC-N2O are interim 
estimates developed by EPA for use in regulatory analyses conducted 
under the guidelines specified in E.O. 13783 and OMB Circular A-4, and 
incorporate a 3 percent discount rate.
(14) External Costs of Congestion and Noise
(a) Values Used To Analyze the Proposal
    As explained in the proposal, changes in vehicle use affect the 
levels and economic costs of traffic congestion and highway noise 
associated with motor vehicle use.\1932\ Congestion and noise costs are 
``external'' to the vehicle owners whose decisions about how much, 
where, and when to drive more--

[[Page 24736]]

or less--in response to changes in fuel economy result in these costs. 
Therefore, unlike changes in the costs incurred by drivers for fuel 
consumption or safety risks they willingly assume, changes in 
congestion and noise costs are not offset by corresponding changes in 
the travel benefits drivers experience.\1933\
---------------------------------------------------------------------------

    \1932\ The proposal estimated changes in congestion and noise 
costs associated with the overall change in vehicle use, which 
included changes in the use of new cars and light trucks associated 
with the fuel economy rebound effect as well as with changes in the 
use of older vehicles resulting from the effect of CAFE and 
CO2 standards on turnover in the car and light truck 
fleets. As discussed in more detail elsewhere in this final rule, 
the current analysis assumes that total vehicle use (VMT) differs 
between the baseline and regulatory alternatives only because of 
changes in the use of cars and light trucks produced during the 
model years affected by this rule that occur in response to the fuel 
economy rebound effect.
    \1933\ The potential contribution of increased vehicle use to 
the costs of injuries and property damage caused by motor vehicle 
crashes may also be partly external to drivers who elect to travel 
more in response to the fuel economy rebound effect. However, these 
costs are dealt with directly and in more detail than the external 
costs of congestion and noise, in section VI.C.2. below.
---------------------------------------------------------------------------

    Congestion costs are limited to road users; however, since road 
users include a significant fraction of the U.S. population, changes in 
congestion costs are treated as part of the rule's economic impact on 
the broader U.S. economy instead of as a cost or benefit to private 
parties. Costs resulting from road and highway noise are even more 
widely dispersed, because they are borne partly by surrounding 
residents, pedestrians, and other non-road users, and for this reason 
are also considered as a cost to the U.S. economy as a whole.
    To estimate the economic costs associated with changes in 
congestion and noise caused by differences in miles driven, the 
analysis supporting the NPRM used estimates of per-mile congestion and 
noise costs from increased automobile and light truck use that were 
originally developed by FHWA as part of its 1997 Highway Cost 
Allocation Study.\1934\ The agencies previously employed these same 
cost estimates in the 2010, 2011, and 2012 final rules.
---------------------------------------------------------------------------

    \1934\ Federal Highway Administration, 1997 Highway Cost 
Allocation Study, Chapter V, Tables V-22 and V-23, available at 
https://www.fhwa.dot.gov/policy/hcas/final/five.cfm.
---------------------------------------------------------------------------

    The marginal congestion cost estimates reported in the 1997 FHWA 
study were intended to measure the costs of increased congestion 
resulting from incremental growth in travel by different types of 
vehicles (including autos and light trucks), and the delays it causes 
to drivers, passengers, and freight shipments. As explained in the 1997 
FHWA study, the distinction between marginal and average costs is 
extremely important in considering congestion costs on a per-vehicle-
mile basis. Average congestion costs on a section of highway are 
calculated as the total congestion costs experienced by all vehicles, 
divided by total vehicle miles. In contrast, marginal congestion costs 
are calculated as the increase in congestion costs resulting from an 
incremental increase in vehicle miles.
    Marginal congestion costs are significantly higher than average 
congestion costs because each additional vehicle that enters a crowded 
roadway slows travel speeds only slightly, thus adding only modestly to 
the average travel time of vehicles already on the road. During 
congested conditions, however, this modest increase is experienced by a 
very large number of vehicles, so the resulting increase in total delay 
experienced by all travelers using the road can be extremely large. As 
a consequence, the increases in total delay and congestion costs 
associated with additional driving are more than proportional to 
changes in VMT that cause them.\1935\
---------------------------------------------------------------------------

    \1935\ Such ``non-linearity'' is a common feature of complex 
systems, such as computing or juggling. Each additional element 
added to a computation, or ball to a cascade, makes performing the 
task more difficult than the last addition.
---------------------------------------------------------------------------

    The FHWA study's estimates of marginal noise costs reflected the 
variation in noise levels resulting from incremental changes in travel 
by autos, light trucks, and other vehicles, and the annoyance and other 
adverse impacts caused by noise. These included adverse impacts on 
pedestrians and residents of the surrounding area, as well as on 
vehicle occupants themselves.
    To calculate the incremental costs of congestion and noise, the 
agencies multiplied FHWA's ``middle'' estimates of marginal congestion 
and noise costs per mile of auto and light truck travel in urban and 
rural areas by the annual increases in driving attributable to the 
standards to yield increases in total congestion and noise externality 
costs. Because the proposal, and other alternatives that were 
considered, reduced the stringency of CAFE and CO2 standards 
for model years 2021-2026, resulting in lower fuel economy for new cars 
and light trucks produced during those years, the fuel economy rebound 
effect resulted in fewer miles driven relative to the baseline, thus 
generating savings in congestion and noise costs relative to their 
levels under the baseline. Similarly, each of those alternatives also 
reduced the total amount of travel by the used vehicle fleet, 
generating additional savings in these costs.
(b) Comments on the NPRM Values
    The agencies received few comments on the estimates of congestion 
and noise costs they used to analyze the economic impacts of the 
proposal. Almost all of these comments focused on the appropriateness 
of the estimated magnitude of the fuel economy rebound effect they used 
to estimate the change in use of new cars and light trucks or the 
plausibility of the reduction in driving by used vehicles, rather than 
to the unit costs estimates themselves. These included comments from 
ICCT and CARB.\1936\
---------------------------------------------------------------------------

    \1936\ ICCT, Comment, NHTSA-2018-0067-11741 at 121; CARB, 
Comment, NHTSA-2018-0067-11873 at 316.
---------------------------------------------------------------------------

    One individual commenter did suggest that recent growth in traffic 
levels, resulting in part from increased use of home delivery services 
for online purchases, has increased congestion and resulting 
delays.\1937\ Although this commenter is correct, traffic growth is not 
strictly a recent phenomenon, and longer-term growth in vehicle use--
combined with comparatively modest increases in road and highway 
capacity--has contributed to increasing congestion levels. Because 
congestion increases more than proportionately to growing traffic 
volumes, this suggests that FHWA's estimates of congestion costs--now 
more than two decades old--are likely to understate the contribution of 
continuing increases in vehicle use to congestion, resulting delays to 
vehicle occupants and freight shipments, and their associated costs. 
Because noise levels also increase non-linearly with the volume of 
traffic using roads and highways, FHWA's 1997 estimates of marginal 
noise costs may also understate current values.
---------------------------------------------------------------------------

    \1937\ Richard Carriere, NHTSA-2018-0067-12216.
---------------------------------------------------------------------------

(c) Values Used To Analyze the Final Rule
    The agencies are retaining the same methodology employed in the 
NPRM to estimate congestion and noise costs for the final rule. Like 
other nominal estimates used throughout the analysis, the agencies have 
updated the FHWA estimates to account for current economic and highway 
conditions. The major determinants of marginal congestion costs imposed 
by additional travel include baseline traffic volumes, which determine 
current travel speeds and how they would change in response to further 
increases in travel, together with vehicle occupancy and the value of 
occupants' travel time. These last two factors interact to determine 
the average hourly value of delays to vehicles, which is by far the 
largest component of the total cost of delays that occur under 
congested travel conditions.\1938\ Because travel speeds measure the 
duration of congestion-related delays, while the

[[Page 24737]]

value of vehicle occupants' time determines their hourly cost, the 
effects of changes in these variables on overall congestion costs is 
approximately additive, as long as changes in the two are relatively 
modest.
---------------------------------------------------------------------------

    \1938\ Fuel consumption and other operating costs can also 
increase during travel in congested conditions, but their 
relationships to the frequent changes in speed that typically occur 
in congested travel is less well understood, and in any case, they 
vary by far smaller amounts than the value of vehicle occupants' 
travel time.
---------------------------------------------------------------------------

    The agencies approximated the effect of growth in traffic volumes 
on travel speeds and congestion-related delays by increasing congestion 
costs in proportion to the increase in annual vehicle-miles of travel 
per lane-mile on major U.S. highways that occurred between 1997 and 
2017.\1939\ Next, they estimated the increase in the value of travel 
time per vehicle-hour over that same period by combining growth in the 
value of travel time per person-hour--estimated in accordance with DOT 
guidance \1940\--with the increase in average vehicle occupancy by 
persons 16 years of age and older (the same measure of occupancy used 
to estimate the value of refueling time elsewhere in this 
analysis).\1941\ The agencies applied the increases in congestion-
related delays and the hourly value of travel time to FHWA's 1997 
estimates of marginal congestion costs to update those original values 
to reflect current conditions. The updated values of external 
congestion costs are $0.154 per vehicle-mile of increased travel by 
cars and $0.138 per vehicle-mile for light trucks (expressed in 
constant 2018 dollars), and these values are assumed to remain constant 
throughout the analysis period.
---------------------------------------------------------------------------

    \1939\ Traffic volumes, as measured by the annual number of 
vehicle-miles traveled per lane-mile of roads and highways 
nationwide, rose by 53 percent between 1997 and 2017. Calculated 
from FHWA, Highway Statistics, 1998 and 2018, Tables VM-1 and HM-48, 
available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
    \1940\ See U.S. Department of Transportation, ``Revised 
Departmental Guidance for the Valuation of Travel Time in Economic 
Analysis,'' 2016, at 5-6 and Table 1 at 13.
    \1941\ The average hourly value of travel time increased by 82 
percent between 1997 and 2017; see U.S. Department of 
Transportation, ``Departmental Guidance for the Valuation of Travel 
Time in Economic Analysis,'' April 9, 1997, Table 4, and U.S. 
Department of Transportation, ``Benefit-Cost Analysis Guidance for 
Discretionary Grant Programs,'' December 2018, Table A-3. From 1995 
to 2017, the average number of light-duty vehicle occupants 16 years 
of age and older increased by 18 percent; values were tabulated from 
FHWA, Nationwide Personal Transportation Survey, 2005 and 2017, 
using online table designer available at https://nhts.ornl.gov/ and 
https://nhts.ornl.gov/index9.shtml.
---------------------------------------------------------------------------

    Similarly, the agencies revised the FHWA estimate of marginal noise 
costs by adjusting for inflation--since the 1994 base year used to 
express values in the FHWA study. Because marginal noise costs are so 
small--less than $0.001 per mile of travel for both cars and light 
trucks--this change did not have a significant impact on the agencies' 
estimates of benefits and costs from the final rule.
(15) Labor Utilization Assumptions
    In previous joint CAFE/CO2 rulemakings, the agencies 
considered employment impacts on the automobile manufacturing industry, 
but many of the considerations were qualitative. In the NPRM, the 
agencies presented and took comment on a methodology to quantify 
roughly the direct labor utilization impacts. The agencies recognize 
there is significant uncertainty in any forward-looking 
characterization of labor utilization, including effects resulting from 
CAFE/CO2 rulemakings. Changes to other policies such as 
trade policies and tariff policies are likely substantially to alter 
underlying assumptions presented in the analysis for the rulemaking, 
and these changes could dwarf any differences between policy 
alternatives presented. In this section the agencies discuss the 
assumptions made in the NPRM analysis, summarize comments received on 
that work, and respond to these comments.
(a) Labor Utilization Baseline (Including Multiplier Effect) and Data 
Description
    In prior CAFE/CO2 rulemakings, the agencies considered 
an analysis of employment impacts in some form in setting both CAFE and 
tailpipe CO2 emissions standards; NHTSA conducted an 
employment analysis in part to determine whether the standards the 
agency set were economically practicable, that is, whether the 
standards were ``within the financial capability of the industry, but 
not so stringent as to'' lead to ``adverse economic consequences, such 
as a significant loss of jobs or unreasonable elimination of consumer 
choice.'' \1942\ EPA similarly conducted an employment analysis under 
the authority granted to the agency under the Clean Air Act.\1943\ Both 
agencies recognized the uncertainties inherent in estimating employment 
impacts; in fact, both agencies dedicated a substantial amount of 
discussion to uncertainty in employment analyses in the 2012 final rule 
for MYs 2017 and beyond.\1944\ Notwithstanding these uncertainties, by 
imposing costs on new light duty vehicles, CAFE and CO2 
standards can have an impact on the demand for labor. Providing the 
best analysis practicable better informs stakeholders and the public 
about the standards' impact than would omitting any estimates of 
potential labor impacts.
---------------------------------------------------------------------------

    \1942\ 67 FR 77015, 77021 (Dec. 16, 2002).
    \1943\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-24 
(D.C. Cir. 1998) (ordinarily permissible for EPA to consider factors 
not specifically enumerated in the Act).
    \1944\ See 77 FR 62624, 62952, 63102 (Oct. 15, 2012).
---------------------------------------------------------------------------

    The NPRM quantified many of the effects that were previously 
qualitatively identified, but not considered. For instance, in the PRIA 
for the 2017-2025 rule EPA identified ``demand effects,'' ``cost 
effects,'' and ``factor shift effects'' as important considerations for 
labor, but the analysis did not attempt to quantify each of these 
effects.\1945\
---------------------------------------------------------------------------

    \1945\ U.S. EPA, ``Regulatory Impact Analysis: Final Rulemaking 
for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards 
and Corporate Average Fuel Economy Standards,'' at 8-24 to 8-32 
(Aug. 2012).
---------------------------------------------------------------------------

    The NPRM analysis considered direct labor effects on the automotive 
sector. The NPRM evaluated how labor utilization in different facets of 
the automobile manufacturing industry may be affected by the rule, 
including (1) dealership labor related to new light-duty vehicle unit 
sales; (2) assembly labor for vehicles, for engines and for 
transmissions related to new vehicle unit sales; and (3) labor related 
to mandated additional fuel savings technologies, accounting for new 
vehicle unit sales. Importantly, this analysis did not consider whether 
price reductions and regulatory savings associated with different 
standards would, because price reductions would allow consumers to save 
or spend that money on other things of value, increase the consumption 
of other vehicle technologies or, more generally, generate growth in 
other sectors of the overall economy. This means that the analysis is 
inherently and artificially narrow in its focus, and does not represent 
an attempt to quantify the overall labor or economic effects of this 
rulemaking. All labor effects were estimated and reported at a national 
level, in person-years, assuming 2,000 hours of labor per person-
year.\1946\
---------------------------------------------------------------------------

    \1946\ The agencies recognize a few local production facilities 
may contribute meaningfully to local economies, but the analysis 
reported only on national effects.
---------------------------------------------------------------------------

    The NPRM analysis estimated labor effects from the forecasted CAFE 
model technology costs and from review of automotive labor for the MY 
2016 fleet. For each vehicle in the CAFE model analysis, the locations 
for vehicle assembly, engine assembly, and transmission assembly and 
estimated labor in MY 2016 were recorded. The percent of U.S. content 
for each vehicle was also recorded.\1947\ The analysis also

[[Page 24738]]

took into account the portion of parts that are made in the U.S. by 
holding constant the percent of U.S. content for each vehicle as 
manufacturers add fuel-savings technologies. The analysis further 
assumes that the U.S. labor added would be proportional to U.S. 
content, which means that the analysis assumes that U.S. labor inputs 
would remain constant over time, but this does not reflect a prediction 
that U.S. labor inputs actually will remain constant.\1948\ From this 
foundation, the analysis forecasted automotive labor effects as the 
CAFE model added fuel economy technology and adjusted future sales for 
each vehicle.
---------------------------------------------------------------------------

    \1947\ NHTSA provides reports under 49 CFR part 583, ``American 
Automobile Labeling Act Reports'' with information NHTSA received 
from vehicle manufacturers about the U.S./Canadian content (by 
percentage value) of the equipment (parts) used to assemble 
passenger motor vehicles. See https://www.nhtsa.gov/part-583-american-automobile-labeling-act-reports.
    \1948\ This is a key assumption that should be revisited as 
trade deals and tax or tariff policies materially change.
---------------------------------------------------------------------------

    The NPRM analysis also accounted for sales projections in response 
to the different regulatory alternatives; the labor analysis considers 
changes in new vehicle prices and new vehicle sales (for further 
discussion of the sales model, see Section VI.D.1.b(2)). As vehicle 
prices rise, the analysis expected consumers to purchase fewer vehicles 
than they would have at lower prices.\1949\ As manufacturers sell fewer 
vehicles, the manufacturers may need less labor to produce the vehicles 
and dealers may need less labor to sell the vehicles. However, as 
manufacturers add equipment to each new vehicle, the industry will 
require labor resources to develop, sell, and produce additional fuel-
saving technologies. The analysis also accounted for the possibility 
that new standards could shift the relative shares of passenger cars 
and light trucks in the overall fleet (see Section VI.D.1.b(2)); 
insofar as different vehicles involved different amounts of labor, this 
shifting impacts the quantity of estimated labor. The labor analysis 
took into account the anticipated reduction in vehicle sales, shifts in 
the mix of passenger cars and light trucks, and addition of fuel-
savings technologies that result from the regulation--and, 
subsequently, the anticipated increase in sales and reduction of fuel-
savings technologies that are expected to result from a reduction in 
stringency.
---------------------------------------------------------------------------

    \1949\ Many commenters contend that higher prices for more 
efficient goods will have no effect on unit sales and hence 
necessary production resources and employment. The sales aspect of 
labor utilization is addressed in the sales section. NHTSA-2018-
0067-12000-35, Center for Biological Diversity, et al.
---------------------------------------------------------------------------

    For the NPRM analysis, the agencies assumed that some observations 
about the production of MY 2016 vehicles would carry forward, unchanged 
into the future. For instance, assembly plants would remain the same as 
MY 2016 for all products now, and in the future. The analysis assumed 
the percent of U.S. content would remain constant, even as 
manufacturers updated vehicles and introduced new fuel-saving 
technologies. The analysis further assumed that assembly labor hours 
per unit would remain at estimated MY 2016 levels for vehicles, 
engines, and transmissions, and the factor between direct assembly 
labor and parts production labors would remain the same. When 
considering shifts from one technology to another, the analysis assumed 
revenue per employee at suppliers and original equipment manufacturers 
would remain in line with MY 2016 levels, even as manufacturers added 
fuel-saving technologies and realized cost reductions from learning.
    The NPRM analysis focused on automotive labor because adjacent 
employment factors and consumer spending factors for other goods and 
services are uncertain and difficult to predict. The analysis did not 
consider how direct labor changes may affect the macro economy and 
possibly change employment in adjacent industries. For instance, the 
analysis did not consider possible labor changes in vehicle maintenance 
and repair, nor did it consider changes in labor at retail gas 
stations. The analysis did not consider possible labor changes due to 
raw material production, such as production of aluminum, steel, copper, 
and lithium, nor did the agencies consider possible labor impacts due 
to changes in production of oil and gas, ethanol, and electricity. The 
analysis did not analyze potential labor effects arising from 
consumption of other products that would not have occurred but for 
improved fuel economy, nor did the analysis assess the effects arising 
from reduced consumption of other products that results from more 
expensive fuel savings technologies at the time of purchase. The 
effects of increased usage of car-sharing, ride-sharing, and automated 
vehicles were not analyzed. The analysis did not estimate how changes 
in labor from any of these industries could affect gross domestic 
product and possibly affect other industries as a result.
    Many commenters voiced concerns that the NPRM analysis only 
included automotive direct employment, and did not explicitly consider 
other important factors, and that these factors would be better 
addressed with a macroeconomic model. For instance, the International 
Council on Clean Transportation contended that the dollars saved at the 
pump as a result of fuel saving technologies would be spent elsewhere 
in the economy, creating jobs.\1950\ The Association of Global 
Automakers also referenced macroeconomic studies that project long-term 
job gains due to savings at the pump, but also highlight short-term 
setbacks for jobs as money spent to purchase additional fuel saving 
technologies on new vehicles is not spent in other job creating sectors 
of the U.S. economy, which were not considered in an analysis that only 
addresses direct automotive employment.\1951\ The Union of Concerned 
Scientists and Environmental Defense Fund argued that the modeling of 
short-term job losses in the macroeconomic models is incorrect, and 
that purchasing a new vehicle, especially if financed, should increase 
disposable income, because monthly savings at the pump outpace the 
monthly financed cost of the fuel saving equipment, but also that 
consumers will not choose this equipment unless a stringent standard is 
chosen.\1952\ The Institute for Policy Integrity commented that an 
analysis looking only at direct employment is incomplete, and 
encouraged the agencies to include long-term and economy-wide effects 
in scope on employment discussions.\1953\
---------------------------------------------------------------------------

    \1950\ NHTSA-2018-0067-11741-145, ICCT.
    \1951\ NHTSA-2018-0067-12032-30, Association of Global 
Automakers.
    \1952\ NHTSA-2018-0067-12039-38, Union of Concerned Scientists; 
NHTSA-2018-0067-12397-4, Environmental Defense Fund, et al.
    \1953\ NHTSA-2018-0067-12213-66, Institute for Policy Integrity.
---------------------------------------------------------------------------

    The agencies have not quantified employment effects outside of 
automotive sector direct employment for this final rule. The agencies 
agree with commenters that the reductions in production costs of new 
vehicles will free up resources for other productive pursuits. Some 
producers may shift resources away from the development and production 
of fuel saving technologies and into the development and production of 
other vehicle attributes. In this case, there would be a transfer of 
labor resources within a firm. Other producers may instead pass along 
the reduction in production costs to consumers in the form of price 
reductions or avoided price increases, allowing those consumers to 
allocate those new funds between expenditure in other consumption 
categories or savings. The increased expenditure in other consumption 
categories would more efficiently create new employment in sectors 
expanding to cover new market-based (as opposed to regulatory-

[[Page 24739]]

based) demand. Increased savings also creates additional investment in 
new productive capital, which will generate employment opportunities in 
the future. However, the extent and nature of these effects are all 
highly uncertain, and the agencies have therefore not quantified the 
effect of the rule on economy-wide employment in the final rule 
analysis.
    Many commenters expressed concern that America would cede 
leadership in development and production of fuel saving technologies, 
and fuel-saving technology investment would be gutted if augural 
standards were not kept in place. For instance, the Mayor of the City 
of Chillicothe, and Mayors of other Ohio cities, pointed out that many 
light duty vehicles are built in Ohio and neighboring geographies, and 
that workers designing and producing fuel economy equipment make an 
average annual salary of $61,500, expressing concern that if standards 
are lowered, some of these jobs may no longer be necessary.\1954\ The 
BlueGreen Alliance pointed out that over the last twenty years, 
manufacturers have invested billions of dollars into fuel saving 
technologies, and that multinational companies may shift jobs to other 
countries if the standards do not require continued, strong, additional 
investment in even more fuel saving technologies.\1955\
---------------------------------------------------------------------------

    \1954\ NHTSA-2018-0067-12318-2, Mayors of the City of 
Chillicothe and other Ohio cities.
    \1955\ NHTSA-2018-0067-12009-6, BlueGreen Alliance.
---------------------------------------------------------------------------

    The agencies recognize that development of fuel saving technologies 
can be capital intensive. However, high fuel economy standards do not, 
per se, guarantee multinational companies will invest in American 
research and development or production. For example, the larger percent 
U.S. content in the MY 2017 light truck vs. the MY 2017 passenger car 
new vehicle fleet may be tied to the so-called ``Chicken Tax,'' a long-
established tariff on the import of light duty trucks.\1956\ On 
average, a light truck in the MY 2017 fleet contained 47.8 percent U.S. 
content, while a passenger car contained 36.0 percent U.S. content. To 
the extent that other policies encourage multi-national corporations to 
build and invest in U.S. production facilities, these organizations 
will need access to capital to do so. Notably, as part of the sales 
module, as fuel economy of the fleet improves, the agencies assume 
customers increasingly choose light trucks, meaning that a shift 
towards light-trucks is already considered in the CAFE model under the 
augural standards.
---------------------------------------------------------------------------

    \1956\ On average, a light truck in the MY 2017 fleet contained 
47.8 percent U.S. content, while a passenger car contained 36.0 
percent U.S. content.
[GRAPHIC] [TIFF OMITTED] TR30AP20.412

    Finally, no assumptions were made about part-time-level of 
employment in the broader economy and the availability of human 
resources to fill positions. When the economy is at full employment, a 
fuel economy regulation is unlikely to have much impact on net overall 
U.S. employment; instead, labor would primarily be shifted from one 
sector to another. These shifts in employment impose an opportunity 
cost on society, as regulation diverts workers from other market-based 
activities in the economy. In this situation, any effects on net 
employment are likely to be transitory as workers change jobs (e.g., 
some workers may need to be retrained or require time to search for new 
jobs, while short-term labor shortages in some sectors or regions could 
result in firms bidding up wages to attract workers). On the other 
hand, if a regulation comes into effect during a period of less-than-
full employment, a change in labor demand due to regulation would 
affect net overall U.S. employment because the labor market is not in 
equilibrium. Schmalensee and Stavins point out that net positive 
employment effects are possible in the near term when the economy is at 
less than full employment due to the potential hiring of idle labor 
resources by the regulated sector to meet new requirements (e.g., to 
install new equipment) and new economic activity

[[Page 24740]]

in sectors related to the regulated sector longer run.\1957\ However, 
the net effect on employment in the long run is more difficult to 
predict and will depend on the way in which the related industries 
respond to regulatory requirements. For that reason, this analysis does 
not include multiplier effects but instead focuses on labor impacts in 
the most directly affected industries, which would face the most 
concentrated labor impacts.
---------------------------------------------------------------------------

    \1957\ Schmalensee, Richard, and Robert N. Stavins. ``A Guide to 
Economic and Policy Analysis of EPA's Transport Rule.'' White paper 
commissioned by Excelon Corporation, March 2011 (Docket EPA-HQ-OAR-
2010-0799-0676).
---------------------------------------------------------------------------

(b) Estimating Labor for Fuel Economy Technologies, Vehicle Components, 
Final Assembly, and Retailers
    The following sections discuss the approaches to estimating factors 
related to dealership labor, final assembly labor and parts production, 
and fuel economy technology labor.
(i) Dealership Labor
    The NPRM analysis evaluated dealership labor related to new light-
duty vehicle sales, and estimated the labor hours per new vehicle sold 
at dealerships, including labor from sales, finance, insurance, and 
management. The effect of new car sales on the maintenance, repair, and 
parts department labor is expected to be limited, as this need is based 
on the vehicle miles traveled of the total fleet. To estimate the labor 
hours at dealerships per new vehicle sold, the agencies referenced the 
National Automobile Dealers Association 2016 Annual Report, which 
provides franchise dealer employment by department and function.\1958\ 
The analysis estimated that slightly less than 20 percent of dealership 
employees' work relates to new car sales (versus approximately 80 
percent in service, parts, and used car sales), and that on average 
dealership employees working on new vehicle sales labor for 27.8 hours 
per new vehicle sold. The analysis presented today retains assumptions 
about dealership labor hours per vehicle sold.
---------------------------------------------------------------------------

    \1958\ NADA Data 2016: Annual Financial Profile of America's 
Franchised New-Car Dealerships, National Automobile Dealers 
Association, https://www.nada.org/2016NADAdata/ (last visited 
December 20, 2019).
---------------------------------------------------------------------------

(ii) Final Assembly Labor and Parts Production
    As new vehicle sales increase or decrease, the amount of labor 
required to assemble parts and vehicles changes accordingly. The NPRM 
evaluated how the quantity of assembly labor and parts production labor 
for MY 2016 vehicles would increase or decrease in the future as new 
vehicle unit sales increased or decreased. Specific assembly locations 
for final vehicle assembly, engine assembly, and transmission assembly 
for each MY 2016 vehicle were identified. In some cases, manufacturers 
assembled products in more than one location, and the analysis 
identified such products and considered parallel production in the 
labor analysis.
    The analysis estimated average direct assembly labor per vehicle 
(30 hours), per engine (four hours), and per transmission (five hours) 
based on a sample of U.S. assembly plant employment and production 
statistics and other publicly available information. The analysis used 
the assembly locations and averages for labor per unit to estimate U.S. 
assembly labor hours for each vehicle. U.S. assembly labor hours per 
vehicle ranged from as high as 39 hours if the manufacturer assembled 
the vehicle, engine, and transmission at U.S. plants, to as low as zero 
hours if the manufacturer imported the vehicle, engine, and 
transmission.
    The analysis also considered labor for parts production. The 
agencies surveyed motor vehicle and equipment manufacturing labor 
statistics from the U.S. Census Bureau, the Bureau of Labor Statistics, 
and other publicly available sources. The agencies found that the 
historical average ratio of vehicle assembly manufacturing employment 
to employment for total motor vehicle and equipment manufacturing for 
new vehicles was roughly constant over the period from 2001 through 
2013, at a ratio of 5.26.\1959\ Observations from 2001-2013 included 
many combinations of technologies and technology trends, and many 
economic conditions, yet the ratio remained about the same over time. 
Accordingly, the analysis scaled up estimated U.S. assembly labor hours 
by a factor of 5.26 to consider U.S. parts production labor in addition 
to assembly labor for each vehicle. The estimates for vehicle assembly 
labor and parts production labor for each vehicle scaled up or down as 
unit sales scaled up or down over time in the CAFE model.
---------------------------------------------------------------------------

    \1959\ NAICS Code 3361, 3363.
---------------------------------------------------------------------------

    The analysis presented today retains assumptions about coefficients 
for final assembly labor and parts production, and updates production 
and final assembly locations for the MY 2017 fleet. As discussed in 
Section VI.D.1.b(2), today's analysis also applies updated methods for 
estimating the extent to which changes in CAFE and CO2 
standards might lead to changes in quantities of new vehicles sold each 
year. These estimated changes in sales lead to changes in estimated 
changes in domestic employment.
(iii) Fuel Economy Technology Labor
    As manufacturers spend additional dollars on fuel-saving 
technologies, parts suppliers and manufacturers require labor to bring 
those technologies to market. Manufacturers may add, shift, or replace 
employees in ways that are difficult for the agencies to predict; 
however, it is expected that the revenue per labor hour at original 
equipment manufacturers (OEMs) and suppliers will remain about the same 
as in MY 2016 even as manufacturers include additional fuel-saving 
technology. To estimate the average revenue per labor hour at OEMs and 
suppliers, the analysis looked at financial reports from publicly 
traded automotive businesses.\1960\ Based on recent figures, it was 
estimated that OEMs would add one labor year per each $633,066 
increment in revenue and that suppliers would add one labor year per 
$247,648 in revenue.\1961\ These global estimates are applied to all 
revenues, and U.S. content is applied as a later adjustment. In today's 
analysis, the agencies assume these ratios would remain constant for 
all technologies rather than that the increased labor costs would be 
shifted toward foreign countries. There are some reasons to believe 
that this may be a conservative assumption. For instance, domestic 
manufacturers may react to increased labor costs by searching for 
lower-cost labor in other countries.
---------------------------------------------------------------------------

    \1960\ The analysis considered suppliers that won the Automotive 
News ``PACE Award'' from 2013-2017, covering more than 40 suppliers, 
more than 30 of which are publicly traded companies. Automotive News 
gives ``PACE Awards'' to innovative manufacturers, with most recent 
winners earning awards for new fuel-savings technologies.
    \1961\ The analysis assumed incremental OEM revenue as the 
retail price equivalent for technologies, adjusting for changes in 
sales volume. The analysis assumed incremental supplier revenue as 
the technology cost for technologies before retail price equivalent 
mark-up, adjusting for changes in sales volume.
---------------------------------------------------------------------------

    The analysis presented today retains assumptions about coefficients 
for fuel economy technology labor, and updates the percent of U.S. 
content for the MY 2017 fleet.
(iv) Labor Calculations
    The agencies estimated the total labor effect as the sum of three 
components: changes to dealership hours, final assembly and parts 
production, and labor for fuel-economy technologies (at OEMs and 
suppliers) that are due to the final rule. The CAFE model calculated

[[Page 24741]]

additional labor hours for each vehicle, based on current vehicle 
manufacturing locations and simulation outputs for additional 
technologies, and sales changes. The analysis applied some constants to 
all vehicles.\1962\ Other constants were vehicle specific, for all 
years considered in the analysis.\1963\ Still, other constants were 
year-specific for a vehicle.\1964\ While a multiplier effect of all 
U.S. automotive related labor on non-auto related U.S. jobs was not 
considered for the final rule's analysis, the analysis did incorporate 
a ``global multiplier'' that can be used to scale up or scale down the 
total labor hours. This parameter exists in the parameters file, and 
for the final rule's analysis the analysis set the value at 1.00. The 
results of this analysis can be found in Table VI-201 below.
---------------------------------------------------------------------------

    \1962\ The analysis applied the same assumptions to all 
manufacturers for annual labor hours per employee, dealership hours 
per unit sold, OEM revenue per employee, supplier revenue per 
employee, and factor for the jobs multiplier.
    \1963\ The analysis made vehicle-specific assumptions about 
percent of U.S. content and U.S. assembly employment hours.
    \1964\ The analysis estimated technology cost for each vehicle, 
for each year based on the technology content applied in the CAFE 
model, year-by-year.
[GRAPHIC] [TIFF OMITTED] TR30AP20.413

    Results of this analysis can be found in Section VII. Considering 
that, all else equal, increases in new vehicle sales lead to increases 
in domestic employment while decreases in technology outlays lead to 
decreases in domestic employment, the agencies estimate that less 
stringent standards could slightly reduce domestic employment. It is 
important to note, however, that the reduction in person-years 
described in this table merely reflects the fact that, when compared to 
the standards set in 2012, fewer jobs will be specifically created to 
meet regulatory requirements that, for other reasons, are not 
economically practicable. It is also important to note that avoided 
outlays for technology can be invested by manufacturers into other 
areas, or passed on to consumers. Moreover, consumers can either take 
those cost savings in the form of a reduced vehicle price, or used 
toward the purchase of specific automotive features that they desire 
(potentially including a more-efficient vehicle), which would increase 
employment among suppliers and manufacturers.
2. Simulating Safety Impacts of Regulatory Alternatives
    The primary objectives of CAFE and CO2 standards are to 
achieve maximum feasible fuel economy and reduce CO2 
emissions, respectively, from the light-duty vehicle fleet. In setting 
standards to achieve these intended effects, the potential of the 
standards to affect vehicle safety is also considered. As a safety 
agency, NHTSA has long considered the potential for adverse safety 
consequences when establishing CAFE standards, and under the CAA, EPA 
considers factors related to public health and human welfare, including 
safety, in regulating emissions of air pollutants from mobile sources.
    Safety trade-offs associated with increases in fuel economy 
standards have occurred in the past--particularly before CAFE standards 
became attribute-based--because manufacturers chose to comply with 
stricter standards by building smaller and lighter vehicles. In cases 
where fuel economy improvements were achieved through reductions in 
vehicle size and mass, the smaller, lighter vehicles did not protect 
their occupants as effectively in crashes as larger, heavier vehicles, 
on average. Although the agencies now use attribute-based standards, in 
part to reduce the incentive to downsize vehicles to comply with CAFE 
and CO2 standards, the agencies must continue to be mindful 
of the possibility of safety-related trade-offs.
    Although prior analyses acknowledged that CAFE and CO2 
standards could influence factors that affect safety other than vehicle 
mass, those impacts were not estimated quantitatively.\1965\ Instead, 
the agencies focused exclusively on the safety impacts of changes in 
vehicle mass. In the proposal, the safety analysis was expanded to 
include a broader and more comprehensive measure of safety impacts. The 
final rule retains this comprehensive approach and analyzes the safety 
impact of three factors:
---------------------------------------------------------------------------

    \1965\ The agencies included a quantification of rebound-
associated safety impacts in its Draft TAR analysis, but because the 
scrappage model is new for this rulemaking, did not include safety 
impacts associated with the effect of standards on new vehicle 
prices and thus on fleet turnover. The fact that the scrappage model 
did not exist prior to this rulemaking does not mean that the 
effects that it aims to show were not important considerations, 
simply that the agencies were unable to account for them 
quantitatively prior to the current rulemaking.
---------------------------------------------------------------------------

    (1)Changes in Vehicle Mass. Similar to previous analyses, the 
agencies calculate the safety impact of changes in vehicle mass made to 
reduce fuel consumption and comply with the standards. The agencies' 
statistical analysis of historical crash data indicates reducing mass 
in heavier vehicles generally improves safety, while reducing mass in 
lighter vehicles generally reduces safety. NHTSA's crash simulation 
modeling of vehicle design

[[Page 24742]]

concepts for reducing mass revealed similar effects.
    (2)Impacts of Vehicle Prices. Vehicles have become safer over time 
through a combination of new safety regulations and voluntary safety 
improvements. The agencies expect this trend to continue as emerging 
technologies, such as advanced driver assistance systems, are 
incorporated into new vehicles. Safety improvements will likely 
continue regardless of changes to CAFE standards. However, the pace of 
such improvements may be modified if manufacturers choose to delay or 
forgo investments in safety technology because of the demands that 
complying with stricter CAFE and CO2 standards impose on 
scarce research, development, and manufacturing resources.
    As discussed in Section VI.D.1.b), technologies added to comply 
with fuel economy standards have an impact on vehicle prices, and, by 
extension, on the affordability of newer, safer vehicles, and therefore 
on the rates at which newer vehicles are acquired and older, less safe 
vehicles are retired from use. The delays in fleet turnover caused by 
the effect of new vehicle prices on sales and scrappage rates affect 
safety, by slowing the penetration of new safety technologies into the 
fleet.
    The standards also influence the composition of the light-duty 
fleet. As the safety provided by light trucks, SUVs and passenger cars 
responds differently to technology that manufacturers employ to meet 
the standards--particularly mass reduction--fleets with different 
compositions of body styles will have varying numbers of fatalities, so 
changing the share of each type of light-duty vehicle in the projected 
future fleet impacts safety outcomes.
    (3)Increased driving because of better fuel economy.The ``rebound 
effect'' predicts consumers will drive more when the cost of driving 
declines. More stringent standards reduce vehicle operating costs, and 
in response, some consumers may choose to drive more. Additional 
driving increases exposure to risks associated with motor vehicle 
travel, and this added exposure translates into higher fatalities and 
injuries.
    We measure the impact of these factors as differences in fatalities 
across the alternatives. Fatalities are calculated by deriving a fleet-
wide fatality rate (fatalities per vehicle mile of travel) 
incorporating the different factors and multiplying it by the 
alternative's expected VMT. Fatalities are converted into a societal 
cost by multiplying fatalities with the DOT-recommended value of a 
statistical life (VSL). As with the NPRM, traffic injuries and property 
damage are not modeled directly; \1966\ rather, traffic injuries and 
property damage continue to be estimated using adjustment factors that 
reflect the observed relationship between societal costs of fatalities 
and costs of injuries and property damage.
---------------------------------------------------------------------------

    \1966\ The agencies noted in the NPRM that traffic injuries and 
property damage are not directly modeled because of insufficient 
data. See PRIA at 43108.
---------------------------------------------------------------------------

    All three factors influence predicted fatalities, but only two of 
them--changes in vehicle mass and in the composition of the light-duty 
fleet in response to changes in vehicle prices--impose increased risks 
on drivers and passengers that are not compensated for by accompanying 
benefits. In contrast, increased driving associated with the rebound 
effect is a consumer choice that reveals the benefit of additional 
travel. Consumers who choose to drive more have apparently concluded 
that the utility of additional driving exceeds the additional costs for 
doing so--including the crash risk that they perceive additional 
driving involves. As discussed in Section VI.D.2.c), the agencies 
account for the benefits of rebound driving by offsetting a portion of 
the added safety costs.
    Some commenters argued that the agencies should be measuring the 
change in the fatality rate rather than the change in the number of 
fatalities. For example, EDF argued that changes in fatalities was a 
measurement of VMT and number of passengers rather than safety, and 
that ``NHTSA's job is to decrease the fatality rate per mile, not to 
decrease the number of miles people drive.'' \1967\ EDF also commented 
that the agencies were required to report the ``fatality rate data for 
the overall safety impacts.'' The agencies disagree with EDF. The 
agencies are responsible for measuring the impacts of fuel economy and 
CO2 standards, including changes to VMT. While other NHTSA 
safety rules have minimal impacts upon aggregate VMT, CAFE standards 
have a large impact on VMT and VMT-related costs, including fatalities.
---------------------------------------------------------------------------

    \1967\ EDF, Appendix A, NHTSA-2018-0067-12108, at 7-9.
---------------------------------------------------------------------------

    Although NHTSA often uses changes in fatality rates as a metric to 
evaluate the impact of regulations on safety, these rates are just a 
tool utilized to derive the relevant safety impact--namely the 
estimated change in fatalities. Furthermore, as part of the cost-
benefit analysis required by Executive Order 12866 and specified in OMB 
Circular A-4, the agencies must quantify and value safety impacts to 
compare them to the costs of the regulation. The fundamental metric for 
valuing loss of life is the VSL. To apply this metric, the agencies 
must first produce estimates of any change in the number of fatalities 
that results from the regulatory action. Fatalities prevented, as well 
as other safety impacts such as non-fatal injuries prevented and 
property damage crashes avoided, are appropriate measures of rules that 
affect motor vehicle safety.
(a) Impact of Weight Reduction on Safety
    Vehicle mass reduction can be one of the more cost-effective means 
of increasing fuel economy and reducing CO2 emissions to 
meet standards--particularly for makes and models not already built 
with much high strength steel or aluminum closures or low mass 
components. Manufacturers have stated that they will continue to reduce 
vehicle mass to meet more stringent standards, and therefore, this 
expectation is incorporated into the modeling analysis supporting the 
standards. Safety trade-offs associated with mass-reduction have 
occurred in the past, particularly before CAFE standards were 
attribute-based; past safety trade-offs may have occurred because 
manufacturers chose at the time, in response to CAFE standards, to 
build smaller and lighter vehicles. In cases where fuel economy 
improvements were achieved through reductions in vehicle size and mass, 
the smaller, lighter vehicles did not fare as well in crashes as 
larger, heavier vehicles, on average. Although the agencies now use 
attribute-based standards, in part to reduce or eliminate the incentive 
to downsize vehicles to comply with CAFE and CO2 
standards,\1968\ the agencies must be mindful of the possibility of 
related safety trade-offs.
---------------------------------------------------------------------------

    \1968\ CAFE and CO2 standards are ``footprint-
based,'' with footprint being defined as a measure of a vehicle's 
size, roughly equal to the wheelbase times the average of the front 
and rear track widths. Footprint-based standards create a 
disincentive for manufacturers to produce smaller-footprint 
vehicles. This is because, as footprint decreases, the corresponding 
fuel economy/CO2 emission target becomes more stringent. 
We also believe that the shape of the footprint curves themselves is 
such that the curves should neither encourage manufacturers to 
increase nor decrease the footprint of their fleets.
---------------------------------------------------------------------------

    Historically, as shown in FARS data analyzed by the agencies, mass 
reduction concentrated among the heaviest vehicles (chiefly, the 
largest LTVs, CUVs and minivans) is estimated to reduce overall 
fatalities, while mass reduction concentrated among the lightest 
vehicles (chiefly, smaller passenger cars) is estimated to increase

[[Page 24743]]

overall fatalities. Mass reduction in heavier vehicles is more 
beneficial to the occupants of lighter vehicles than it is harmful to 
the occupants of the heavier vehicles. Mass reduction in lighter 
vehicles is more harmful to the occupants of lighter vehicles than it 
is beneficial to the occupants of the heavier vehicles. In response to 
questions of whether designs and materials of more recent model year 
vehicles may have weakened the historical statistical relationships 
between mass, size, and safety, the agencies updated our public 
database for statistical analysis consisting of crash data. The 
analysis considered the full range of real-world crash types.
    The methodology used for the statistical analysis of historical 
crash data has evolved over many years. The methodology used for the 
NPRM and unchanged for the final rule reflects learnings and 
refinements from: NHTSA studies in 2003, 2010, 2011, 2012, and 2016; 
independent peer review of 23 studies by the University of Michigan 
Transportation Research Institute;\1969\ two public workshops hosted by 
NHTSA;\1970\ interagency collaboration among NHTSA, DOE and EPA; and 
comments to CAFE and CO2 rulemakings in 2010, 2012, the 2016 
Draft TAR, and the 2018 NPRM. As explained in greater detail below, the 
methodology used for the statistical analysis of historical crash data 
for the NPRM and final rule is the best and most up to date available.
---------------------------------------------------------------------------

    \1969\ Green, Paul E., Kostyniuk, Lidia P., Gordon, Timothy J., 
and Reed, Matthew P., Independent Review of Statistical Analyses of 
Relationship between Vehicle Curb Weight, Track Width, Wheelbase and 
Fatality Rates, UMTRI-2011-12, University of Michigan of 
Transportation Research Institute (2011). Available at http://www.umtri.umich.edu/our-results/publications/independent-review-statistical-analyses-relationship-between-vehicle-curb.
    \1970\ The workshops were held on February 25, 2011 and May 13-
14, 2013. Video, transcripts, and presentations are available on the 
NHTSA website (recommended search terms include ``workshop,'' 
``mass,'' ``safety,'' and the dates of the workshops).
---------------------------------------------------------------------------

    Additionally, to assess whether future vehicle designs may impact 
the relationship of vehicle mass reduction on safety, NHTSA sponsored a 
fleet crash simulation study using future mass reduction vehicle design 
concepts (see Fleet Simulation Study below). The results of the 
simulation research showed that future mass reduction techniques 
continue to exhibit impacts on safety and were consistent with the 
statistical analysis of FARS crash data. The agencies considered the 
findings of the study and concluded it was reasonable and appropriate 
to continue to consider the impact of mass reduction on safety for 
future vehicles because the data indicate the relationship between mass 
and safety will continue in the future.
    For the rulemaking analysis, the CAFE model tracks the amount of 
mass reduction applied to each vehicle model, and then applies 
estimated changes in societal fatality risk per 100 pounds of mass 
reduction determined through the statistical analysis of FARS crash 
data. This process allows the CAFE model to tally changes in fatalities 
attributed to mass reduction across all of the analyzed future model 
years. In turn, the CAFE model is able to provide an overall impact of 
the final standards and alternatives on fatalities attributed to mass 
reduction.
    A number of comments were received on technical aspects of the 
mass-safety analysis in the NPRM. The agencies carefully considered all 
comments. Where warranted, the agencies conducted additional analyses 
to determine whether commenters' suggestions would improve the 
analysis. The agencies found that the methodology employed by the 
proposal, which was developed over many years, subject to extensive 
review and feedback, remains the most rigorous methodology. The 
agencies found the alternative approaches raised in comments would 
provide less likely estimates, were statistically problematic, or, in 
some cases, advocated discarding or ignoring the most likely estimates 
altogether. The agencies' assessments of comments are discussed in 
detail in the subsections below.
    Overall, consistent with prior analyses, the data show that mass 
reduction concentrated in heavier vehicles is generally beneficial to 
overall safety, and mass reduction concentrated in lighter vehicles is 
harmful.
(1) Crash Data
    The agencies use real-world crash data as the basis for projecting 
the future safety implications for regulatory changes. To support the 
2012 rulemaking, NHTSA created a common, updated database for 
statistical analysis consisting of crash data. The initial iteration 
contained crash data for model years 2000-2007 vehicles in calendar 
years 2002-2008. NHTSA made the preliminary version of the new 
database, which was the basis for NHTSA's 2011 preliminary report 
(hereinafter 2011 Kahane report),\1971\ available to the public in May 
2011, and an updated version in April 2012 (used in NHTSA's 2012 final 
report, hereinafter 2012 Kahane report), \1972\ enabling other 
researchers to analyze the same data and, hopefully, minimize 
discrepancies in results caused by reporting inconsistencies across 
databases.\1973\ NHTSA updated the crash and exposure databases for the 
2016 Draft TAR analysis.
---------------------------------------------------------------------------

    \1971\ Kahane, C, J. Relationships Between Fatality Risk, Mass, 
and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--Final 
Report, National Highway Traffic Safety Administration (Aug. 2012). 
Available at https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811665.
    \1972\ Kahane, C, J. Relationships Between Fatality Risk, Mass, 
and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--
Preliminary Report. Docket No. NHTSA-2010-0152-0023. Washington, DC: 
National Highway Traffic Safety Administration.
    \1973\ See 75 FR 25324, 25395-96 (May 7, 2010).
---------------------------------------------------------------------------

    For the proposed rule and unchanged for today's final rule, the 
crash and exposure databases were updated again. The databases are the 
most up-to-date possible (MY 2004-2011 vehicles in CY 2006-2012), given 
the processing time for crash data and the need for enough crash cases 
to permit statistically meaningful analyses. As in previous analyses, 
NHTSA has made the new databases available to the public on its 
website.\1974\
---------------------------------------------------------------------------

    \1974\ ftp://ftp.nhtsa.dot.gov/CAFE/2018_mass_size_safety/.
---------------------------------------------------------------------------

(2) Methodology
    The relationship between a vehicle's mass, size, and fatality risk 
is complex, and it varies in different types of crashes. The agencies 
have been examining this relationship for more than two decades. The 
basic analytical method used to analyze the impacts of weight reduction 
on safety for the proposal, and unchanged for this final rulemaking, is 
the same as in 2016 Puckett and Kindelberger report.\1975\ NHTSA 
released the 2016 Puckett and Kindelberger report as a preliminary 
report on the relationship between fatality risk, mass, and footprint 
in June 2016 in advance of the Draft TAR. The 2016 Puckett and 
Kindelberger report covered the same scope as previous NHTSA 
reports,\1976\ offering a detailed description of the crash and 
exposure databases, modeling approach, and analytical results on 
relationships among vehicle size, mass, and fatalities that informed 
the Draft TAR. The

[[Page 24744]]

modeling approach described in the 2016 Puckett and Kindelberger report 
was developed with the collaborative input of NHTSA, EPA and DOE, and 
subject to extensive public review, scrutiny in two NHTSA-sponsored 
workshops, and a thorough peer review that compared it with the 
methodologies used in other studies.\1977\
---------------------------------------------------------------------------

    \1975\ Puckett, S.M. and Kindelberger, J.C. (2016, June). 
Relationships between Fatality Risk, Mass, and Footprint in Model 
Year 2003-2010 Passenger Cars and LTVs--Preliminary Report. (Docket 
No. NHTSA-2016-0068). Washington, DC: National Highway Traffic 
Safety Administration, available at https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/2016-prelim-relationship-fatalityrisk-mass-footprint-2003-10.pdf.
    \1976\ The 2016 Puckett and Kindelberger report is an extension 
of 2011 Kahane report and 2012 Kahane report.
    \1977\ Previous reports from which the 2016 Puckett and 
Kindelberger report was derived from, were also subject to extensive 
peer reviews. Farmer, Green, and Lie, who reviewed the 2010 Kahane 
report, also peer-reviewed the 2011 Kahane report. In preparing his 
2012 report (along with the 2016 Puckett and Kindelberger report and 
Draft TAR), Kahane also took into account Wenzel's assessment of the 
preliminary report and its peer reviews, DRI's analyses published 
early in 2012, and public comments such as the International Council 
on Clean Transportation's comments submitted on NHTSA and EPA's 2010 
notice of joint rulemaking. These comments prompted supplementary 
analyses, especially sensitivity tests, discussed at the end of this 
section.
---------------------------------------------------------------------------

    In computing the impact of changes in mass on safety, the agencies 
are faced with competing challenges. Research has consistently shown 
that mass reduction affects ``lighter'' and ``heavier'' vehicles 
differently across crash types. The 2016 Puckett and Kindelberger 
report found mass reduction concentrated amongst the heaviest vehicles 
is likely to have a beneficial effect on overall societal fatalities, 
while mass reduction concentrated among the lightest vehicles is likely 
to have a detrimental effect on fatalities.\1978\ To accurately capture 
the differing effect on lighter and heavier vehicles, the agencies must 
split vehicles into lighter and heavier vehicle classifications in the 
analysis.\1979\ However, this poses a challenge of creating 
statistically-meaningful results. There is limited relevant crash data 
to use for the analysis. Each partition of the data reduces the number 
of observations per vehicle classification and crash type, and thus 
reduces the statistical robustness of the results. The methodology 
employed by the agencies was designed to balance these competing forces 
as an optimal trade-off to accurately capture the impact of mass-
reduction across vehicle curb weights and crash types while preserving 
the potential to identify robust estimates.
---------------------------------------------------------------------------

    \1978\ The findings of the 2016 Puckett and Kindelberger report 
are consistent with the results of the 2012 Kahane report and Draft 
TAR.
    \1979\ If lighter and heavier vehicles are left undistinguished, 
the agencies analysis would be restricted to identifying a single 
effect of mass reduction for passenger cars and a single effect of 
mass reduction for truck-based LTVs. As discussed below, distinct 
effects have been estimated historically for lighter versus heavier 
vehicles for cars and LTVs, confirming the validity of 
distinguishing by curb weight where feasible.
---------------------------------------------------------------------------

    For the proposal and the final rule, the agencies employed the 
modeling technique developed in the 2016 Puckett and Kindelberger 
report to analyze the updated crash and exposure data by examining the 
cross sections of the societal fatality rate per billion vehicle miles 
of travel (VMT) by mass and footprint, while controlling for driver 
age, gender, and other factors, in separate logistic regressions for 
five vehicle groups and nine crash types. ``Societal'' fatality rates 
include fatalities to occupants of all the vehicles involved in the 
collisions, plus any pedestrians, cyclists, or occupants of other 
conveyances (e.g., motorcyclists). The agencies utilize the 
relationships between weight and safety from this analysis, expressed 
as percentage increases in fatalities per 100-pound weight reduction, 
to examine the weight impacts applied in this CAFE analysis. The 
effects of mass reduction on safety were estimated relative to 
(incremental to) the regulatory baseline (augural standards) in the 
CAFE analysis, across all vehicles for MYs 2018 and beyond.
    As in the 2012 Kahane report, 2016 Puckett and Kindelberger report, 
and the Draft TAR, the vehicles are grouped into three classes: 
Passenger cars (including both two-door and four-door cars); CUVs and 
minivans; and truck-based LTVs. The curb weight of passenger cars is 
formulated, as in the 2012 Kahane report, 2016 Puckett and Kindelberger 
report, and Draft TAR, as a two-piece linear variable to estimate one 
effect of mass reduction in the lighter cars and another effect in the 
heavier cars. The boundary between ``lighter'' and ``heavier'' cars is 
3,201 pounds (which is the median mass of MY 2004-2011 cars in fatal 
crashes in CY 2006-2012, up from 3,106 pounds for MY 2000-2007 cars in 
CY 2002-2008 in the 2012 NHTSA safety database, and up from 3,197 
pounds for MY 2003-2010 cars in CY 2005-2011 in the 2016 NHTSA safety 
database). Likewise, for truck-based LTVs, curb weight is a two-piece 
linear variable with the boundary at 5,014 pounds (again, the MY 2004-
2011 median, higher than the median of 4,594 pounds for MY 2000-2007 
LTVs in CY 2002-2008 and the median of 4,947 pounds for MY 2003-2010 
LTVs in CY 2005-2011). CUVs and minivans are grouped together in a 
single group covering all curb weights of those vehicles; as a result, 
curb weight is formulated as a simple linear variable for CUVs and 
minivans. Historically, CUVs and minivans have accounted for a 
relatively small share of new-vehicle sales over the range of the data, 
resulting in less crash data available than for cars or truck-based 
LTVs. In sum, vehicles are distributed into five groups by class and 
curb weights: Passenger cars < 3,201 pounds; passenger cars 3,201 
pounds or greater; truck-based LTVs < 5,014 pounds; truck-based LTVs 
5,014 pounds or greater; and all CUVs and minivans.
    There are nine types of crashes specified in the analysis for each 
vehicle group: three types of single-vehicle crashes, five types of 
two-vehicle crashes; and one classification of all other crashes. 
Single-vehicle crashes include first-event rollovers, collisions with 
fixed objects, and collisions with pedestrians, bicycles and 
motorcycles. Two-vehicle crashes include collisions with: heavy-duty 
vehicles; cars, CUVs, or minivans < 3,187 pounds (the median curb 
weight of other, non-case, cars, CUVs and minivans in fatal crashes in 
the database); cars, CUVs, or minivans >= 3,187 pounds; truck-based 
LTVs < 4,360 pounds (the median curb weight of other truck-based LTVs 
in fatal crashes in the database); and truck-based LTVs >= 4,360 
pounds. Grouping partner-vehicle CUVs and minivans with cars rather 
than LTVs is more appropriate because their front-end profile and 
rigidity more closely resemble a car than a typical truck-based LTV. An 
additional crash type includes all other fatal crash types (e.g., 
collisions involving more than two vehicles, animals, or trains). 
Splitting the vehicles from this crash type involved in crashes 
involving two light-duty vehicles into a lighter and a heavier group 
permits more accurate analyses of the mass effect in collisions of two 
vehicles.
    For a given vehicle class and weight range (if applicable), 
regression coefficients for mass (while holding footprint constant) in 
the nine types of crashes are averaged, weighted by the number of 
baseline fatalities that would have occurred for the subgroup MY 2008-
2011 vehicles in CY 2008-2012 if these vehicles had all been equipped 
with electronic stability control (ESC). The adjustment for ESC, a 
feature of the analysis added in 2012, takes into account results will 
be used to analyze effects of mass reduction in future vehicles, which 
will all be ESC-equipped, as required by NHTSA's safety regulations.
    The agencies received multiple comments on how they distribute 
vehicles into classifications. IPI, quoting a study by Tom Wenzel, 
commented that sorting vehicles into footprint deciles shows positive 
impacts from mass reduction for the majority of the

[[Page 24745]]

footprint deciles.\1980\ CARB commented that the agencies should have 
used the curb weight of all vehicles to calculate the thresholds for 
``lighter'' and ``heavier'' vehicle types rather than just the curb 
weights of vehicles involved in fatal crashes.\1981\ CARB also 
commented that pickup trucks and SUVs that are not subject to CAFE 
regulation (i.e., Class 2b and Class 3 vehicles, such as \3/4\-ton and 
one-ton pick-up trucks, vans and related SUVs) should not be included 
in the assessment of the impact of mass on safety and doing so raises 
the median weight of trucks.\1982\ CARB also commented that the median 
weights are static values representing the historical fleet, but the 
median weights and proportions of crash types involving given vehicle 
weight categories should change with median weight of the fleet modeled 
by the CAFE model.\1983\ Commenters generally believed that the 
agencies' approach ``results in inappropriate apportioning of cars and 
trucks into the corresponding lighter or heavier bins,'' which in turn 
causes the agencies to overestimate the fatalities associated with mass 
reduction.\1984\
---------------------------------------------------------------------------

    \1980\ IPI, Detailed Comments, Docket No. NHTSA-2018-0067-12213, 
at 127 (quoting Tom Wenzel, Assessment of NHTSA's Report 
``Relationships Between Fatality Risk, Mass, and Footprint in Model 
Year 2004-2011 Passenger Cars and LTVs,'' (LBNL Phase 1, 2018). 
Available at https://escholarship.org/uc/item/4726g6jq.
    \1981\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 276.
    \1982\ Tom Wenzel of Lawrence Berkeley National Laboratories, 
Comment, EPA-HQ-OAR-2018-0283-4118, at 1; see also CARB, Detailed 
Comments, Docket No. NHTSA-2018-0067-11873, at 259.
    \1983\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 260.
    \1984\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 276.
---------------------------------------------------------------------------

    Dividing vehicles into footprint deciles and excluding Class 2b and 
3 vehicles pose sample size and data coverage issues. If vehicles were 
grouped into footprint deciles, the sample sizes in each decile would 
be approximately one-fifth as large as the corresponding sample sizes 
in each of the agencies' four passenger car and LTV vehicle classes 
(and one-tenth as large as the sample size for CUVs and minivans). 
Smaller parameter estimates require correspondingly smaller standard 
errors (i.e., relatively precise estimates) to achieve statistical 
significance, but splitting the limited data into deciles yields larger 
standard errors, restricting the ability to identify statistically-
significant estimates. Likewise, by extending the footprint-curb 
weight-fatality data to include Class 2b and 3 trucks that are 
functionally and structurally similar to corresponding \1/2\-ton models 
that are subject to CAFE regulation,\1985\ the sample size and ranges 
of curb weights and footprint are improved. Sample size is a challenge 
for estimating relationships between curb weight and fatality risk for 
individual crash types in the main analysis; dividing the sample 
further or removing observations makes it exceedingly difficulty to 
identify meaningful estimates and the relationships that are present in 
the data.
---------------------------------------------------------------------------

    \1985\ Class 2b and 3 pickup trucks, vans and SUVs have physical 
characteristics and usage profiles that are substantially similar to 
their Class 2a counterparts. For example, the Class 2a version of 
the Ford F-150 has similar physical characteristics to and has a 
similar usage profile to the Class 2b Ford F150. Same for the Class 
2a Ford F150 relative to the Class 2b and 3 Ford F250, and for the 
GMC Yukon relative to the Yukon XL. The Class 2b and 3 pickup trucks 
in the sample generally have gross vehicle weight ratings of 10,000 
pounds or less, and thus are subject to the same Federal motor 
vehicle safety standards as their light-duty counterparts. Likewise, 
these vehicles generally have similar physical dimensions (e.g., 
ground clearance, width) as related light-duty vehicles. Key 
differentiating factors among these vehicles are height, payload, 
and towing capacity. There are likely to be unobserved differences 
in how these vehicles are driven relative to light-duty 
alternatives; however, the crash data include a census of fatal 
crashes involving case vehicles and the Class 2b and 3 vehicles 
included in the analysis, in turn representing the relative risk of 
differences in curb weight in crashes involving Class 2b and 3 
vehicles. Despite being regulated by different fuel economy and 
emissions regulations as they become heavier (i.e., once a given 
model crosses a mass threshold changes classes), the vehicles may 
continue to be used in similar ways over time; in turn, the safety 
implications of the presence of these vehicles may continue to be 
similar. In contrast, other types of heavy-duty vehicles, such as 
box trucks, buses, refuse trucks, fire trucks, and other heavy-duty 
commercial vehicles are substantially different from light duty 
vehicles in their physical characteristics and usage profiles, and 
it would not be appropriate to include them in the statistical 
analysis to determine the impact of mass on crash fatalities.
---------------------------------------------------------------------------

    Compounding the issue is the fact the analysis focuses on societal 
fatality risk (i.e., all fatalities, including crash partners and 
people outside of vehicles, such as pedestrians, cyclists, and 
motorcyclists) rather than merely in-vehicle fatality risk, which 
yields estimates that are smaller in magnitude (and thus more difficult 
to identify meaningful differences from zero) than estimates 
representing changes in in-vehicle fatality risk. That is, compared to 
an analysis of in-vehicle fatality risk (which would tend to yield 
relatively large estimated effects of mass reduction), the focus on 
societal fatalities tends to yield relatively small (net) effects of 
mass reduction on fatality risk.
    Including Class 2b and 3 vehicles in the analysis to determine the 
relationship of vehicle mass on safety has the added benefit of 
improving correlation constraints. Notably, curb weight increases 
faster than footprint for large light trucks and Class 2b and 3 pickup 
trucks and SUVs, in part because the widths of vehicles are constrained 
more tightly (i.e., due to lane widths) than their curb weights. 
Including data from Class 2b and 3 pick-up truck and SUV fatal crashes 
provides data over a wider range of vehicle weights, which improves the 
ability to estimate the mass-crash fatality relationship. The agencies 
believe the decision of whether to include Class 2b and 3 vehicles in 
the analysis should be made based on whether the additional data 
improves the estimate of the safety impact of mass reduction in light 
trucks, and that the fatality data should not be simplistically 
excluded because the vehicles are not regulated under the CAFE and 
CO2 emissions programs. Ultimately, the agencies find that: 
(1) The fundamental objective is to capture the strongest, meaningful 
signal regarding societal fatality risk as a function of the mass of 
light trucks; (2) that incorporating information on fatal incidents 
involving Class 2b and 3 trucks improves the quality of the signal the 
agencies can capture, and (3) including the vehicles provides the best 
estimate of the impacts of mass on societal fatalities.
    In assessing whether to calculate the median curb weight threshold 
from all vehicles involved in accidents or on the road, the agencies 
weighed changing the process used to establish the thresholds and the 
potential impact on the robustness of the statistical analysis. From a 
statistical perspective, using thresholds that allocate a similar 
number of fatal crash cases to both the lower vehicle weight group and 
the higher vehicle weight group for a given vehicle type will minimize 
the average standard errors of estimates for both groups, which 
provides the best estimates for each group. Because reducing average 
standard errors strengthens the statistical analysis, the agencies 
conclude using only the curb weight of vehicles involved in fatal 
crashes to calculate the median curb weight threshold produces the best 
estimate. This conclusion is the same that was reached previously when 
considering the same issue for the 2011 Kahane, 2012 Kahane, and 2016 
Puckett and Kindelberger analyses.
    On a related note, the regression models are estimated based on 
with respect to the total number of fatalities associated within each 
vehicle weight group classification (referred to as vehicle group 
below, for brevity). Shifting the threshold would change the estimated 
incremental impact of changes in curb weight in each vehicle

[[Page 24746]]

group, but the net effects would offset each other across vehicle 
groups, resulting in the same overall estimated effect of changes in 
vehicle mass on societal fatality risk. For example, if one restricted 
the ``lightest'' group for a vehicle type to include only the bottom 
ten percentiles of vehicle weight, one would expect to identify a very 
strong detrimental effect (or weakest beneficial effect) of mass 
reduction for that group. However, the estimated effect of mass 
reduction in that group has minimal implications for the fleet (i.e., 
because there are fewer vehicles in the group), and the corresponding 
estimated effect of mass reduction for other groups would also mute the 
impact (i.e., because there are many vehicles in the group that vary in 
mass to a much larger degree than in the ``lighter'' group). 
Ultimately, the mean effect of mass reduction across the lighter and 
heavier groups would be the same as when using the median as the 
threshold (or at least, similar, subject to limitations in statistical 
optimization), but with a different point of reference when comparing 
the groups. Thus, the agencies believe the selection of curb weight 
threshold has a minimal impact on the estimated effects of mass 
reduction across all vehicle types.
    Full consideration of CARB's comment on mass thresholds, and 
whether they should change as the median weight of the fleet modeled by 
the CAFE model changes, requires a deeper look at each of the crash 
types considered in the analysis. That is, the point estimates 
presented in Table VI-202 represent weighted averages across nine 
separate, mutually-exclusive and exhaustive crash models (analyzed 
separately for cars, LTVs, and CUVs and minivans). For example, an 
individual model for first-event rollovers yields estimates of the 
percentage change in societal fatality risk per 100-pound mass 
reduction for lighter and heavier (or, in the case of CUVs and 
minivans, all) vehicles in the target vehicle class. The final, overall 
point estimate for a given vehicle type is found by: (1) Multiplying 
the estimate associated with an individual crash type by the estimated 
share of societal fatalities involving the vehicle class (adjusting for 
two-vehicle collisions that span vehicle classes to avoid double-
counting); and (2) summing the values estimated in (1) across all crash 
types. In its comments, CARB noted that if the distribution of vehicles 
in terms of curb weight changes through lightweighting, the shares of 
(fatal) two-vehicle crashes involving a given pair of vehicles as 
defined by weight class (e.g., car below a given threshold colliding 
with a LTV above a given threshold) would change. In turn, the 
appropriate weighting across the crash types modeled in the analysis 
would likewise be different (involving an increasing share of vehicles 
below a given curb weight threshold). Due to these potential 
limitations, CARB questioned the stability of the summary point 
estimates relative to changes in the shares of fatalities within each 
crash type in the analysis.\1986\
---------------------------------------------------------------------------

    \1986\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 278-79.
---------------------------------------------------------------------------

    To evaluate CARB's concerns regarding future crash mixes and 
definitions of vehicle weight classes, the agencies performed an 
exploratory analysis examining the scope and impacts of potential model 
changes. In doing so, the agencies examined the degree of change in the 
median vehicle fleet weight in the NPRM analysis relative to the fixed 
mass threshold values, and also how sensitive the curb weight safety 
point estimates are to assumptions about the distribution of curb 
weights in future vehicle fleets. The agencies also considered the 
feasibility of changing the shares of fatalities by crash type as a 
function of forthcoming or developing vehicle safety technologies. This 
information would help inform adjustments to fatality rate impacts for 
each vehicle type, because the likelihood of observing individual fatal 
crash types could change in different ways across vehicle types in the 
analysis as the vehicle mix changes. However, the agencies identified 
no studies on the effectiveness of forthcoming or developing vehicle 
safety technologies that could inform projections of shares of 
fatalities across crash types, nor did the commenters reference any 
such studies. Likewise, commenters provided no data that would enable 
projections of these factors. Thus, for a given vehicle mix, the 
agencies have no information available to justify changing the shares 
of fatalities across crash types over time. Therefore, the agencies 
decided to keep the distribution of fatality shares constant for: 
First-event rollovers; fixed-object collisions; collisions with 
pedestrians, bicyclists, and motorcycles; collisions with heavy 
vehicles; collisions with one other light-duty vehicle (i.e., a 
constant share across the sum of these crashes, but not constant for 
any given type of crash partner); and all other crashes.
    The agencies had sufficient information to evaluate the effects of 
changes in the fatal crash mix for cases involving two light-duty 
vehicles. The agencies agreed that it was internally consistent to 
adjust fatality shares by crash type proportionally to the distribution 
of vehicle types and curb weight classes for a given focal MY. An 
important technical question associated with this approach is the level 
of disaggregation. The agencies considered an alternative in which the 
agencies would estimate and apply unique curb weight point estimates 
for each calendar year in the analysis for each regulatory alternative. 
This alternative would account for changes in the distribution of crash 
types associated with changes in both vehicle type shares (i.e., shifts 
from passenger cars to CUVs and LTVs) and vehicle mass shares (i.e., 
shifts from vehicles above the curb weight thresholds to vehicles below 
the thresholds). As in the status quo analysis of curb weight and 
fatality risk, the resulting point estimates would be weighted averages 
across the individual crash type models as presented in the NPRM, but 
re-weighted to reflect projected changes to the fleet.
    The agencies investigated this alternative and identified several 
concerns. A key functional constraint is that the curb weight safety 
point estimates are applied in the CAFE model as a lump-sum, lifetime 
effect to a given vehicle. This characteristic of the model limits the 
ability to apply calendar-year-specific effects of changes in curb 
weight and vehicle type distributions when evaluating safety impacts of 
changes in curb weights. The safety point estimates also represent net 
effects of changes in curb weights over the lifetime of a given vehicle 
in the CAFE model; any changes in the calculation of safety point 
estimates would need to preserve this characteristic. More broadly, the 
vehicle fleet is not static over a vehicle's lifetime (i.e., the 
distributions of curb weight and vehicle type change each year), so the 
effective probabilities of each crash type over a given vehicle's 
lifetime are a function of many calendar-year-level curb weight and 
vehicle type distributions. To capture any effects of changes in 
vehicle mass distributions over time within the current CAFE model 
structure, the agencies would need to enact a method that: (1) 
Identifies defensible changes in fatality risk associated with vehicle 
mass as the distribution of vehicle mass changes (e.g., accounting for 
changes in the likelihood of observing particular fatal crash types 
that reflect projected changes in the distribution of vehicle types and 
curb weights across vehicles); and (2) allocates calendar-year-specific 
impacts of curb weight on fatality risk to each vehicle in the fleet 
across the

[[Page 24747]]

analysis horizon. Identifying how best to achieve this would be 
complex, and would require the development of an alternative analytical 
approach that would be outside the scope of this rulemaking.
    With these concerns in mind, the agencies explored an alternative 
approach to test the sensitivity of the safety point estimates to 
distributions of vehicles by curb weight and vehicle type. The starting 
point for the alternative approach is maintaining the understanding 
that the nine crash type models that are present in the curb weight 
safety analysis represent the best statistical alternatives for 
evaluating the crash data in the database (i.e., optimal statistical 
precision conditional on the coverage of the data). Furthermore, the 
nine crash type models are defined in terms of physical relationships 
(i.e., crashes involving vehicles of particular curb weight ranges and 
vehicle types) that are invariant to changes in the distributions of 
vehicles for those same characteristics. That is, the estimated changes 
in societal fatality risk as curb weights change for a focal vehicle 
(i.e., of a particular type and weight range) that is involved in a 
particular type of crash apply equally to any scenario involving such 
vehicle, regardless of changes in the probability of observing such a 
scenario. For example, the agencies would expect the societal fatality 
risk for a crash involving a passenger car lighter than 3,201 pounds 
colliding with a LTV heavier than 4,360 pounds to be the same 
regardless of how many such collisions take place. Thus, the agencies 
would expect the net effect of a given change in curb weight for a 
given vehicle type in a given crash type to scale proportionally with 
the probability of such crashes occurring. Put simply, if there are 
half as many potential crash partners of a given type in a future year 
compared to a base year, the agencies would expect a given curb weight 
reduction to have half as large of a net effect on fatalities in the 
future year relative to the base year. In the extreme, curb weight 
changes would have no net effect on fatalities at all for a given crash 
type if such crashes had a zero percent probability of occurring (i.e., 
if there are no potential crash partner vehicles).
    Based on this maintained hypothesis, the agencies examined test 
curb weight safety point estimates under alternative scenarios, in 
which fatality shares by crash type were proportional to the 
distribution of vehicle types and curb weight classes across a range of 
outcomes reflecting different model years and policy alternatives 
represented in the NPRM. The sensitivities of the safety point 
estimates to changes in the distributions of vehicle curb weights and 
vehicle types were tested by adjusting fatality shares across the 
relevant crash types in the analysis (i.e., involving two light-duty 
vehicles) in a manner consistent with potential changes in the vehicle 
fleet, while holding the outputs of the individual crash type models 
the same as in the NPRM.
    For example, compare the safety point estimate for LTVs lighter 
than 5,014 pounds in the NPRM with an alternative point estimate for an 
extreme hypothetical future year where 80 percent of the LTV fleet is 
lighter than the median curb weight for crash partners (4,360 pounds):
[GRAPHIC] [TIFF OMITTED] TR30AP20.414


[[Page 24748]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.415

    The estimated net societal effect of a 100-pound mass reduction is 
equal to: (1) The sum of the estimated net effects across all crash 
types, divided by (2) the baseline estimate of annual fatalities 
involving the vehicle class (adjusted to avoid double-counting) for the 
most recent four MYs in the database (MYs 2008-2011), or 1,782 
fatalities per year. In the NPRM, the estimated net societal effect of 
a 100-pound mass reduction for lighter LTVs was a 5.5 fatality 
increase, or a 0.31 percent increase relative to a baseline of 1,782 
fatalities. Changing the share of crash fatalities involving heavier 
LTVs to be consistent with a fleet with only 20 percent of LTVs above 
the curb weight threshold yields: (1) An increase in incremental 
fatalities in crashes involving lighter LTVs (from 0.5 fatality to 0.7 
fatality); and (2) a decrease in incremental fatalities in crashes 
involving heavier LTVs (from 1.5 fatalities to 0.7 fatality); for a 
total net increment of 4.9 fatalities compared to the NPRM's estimate 
of 5.5 fatalities. Thus, the agencies estimate that, in a future year 
where the fleet differs from the baseline by having an extreme case of 
80 percent of LTVs below the crash-partner curb weight threshold, the 
net societal effect of a 100-pound mass reduction in LTVs lighter than 
5,014 pounds would be 4.9 divided by 1,782, or 0.28 percent, versus 
0.31 percent in the baseline.
    This simple example confirms that the estimates do indeed change as 
the distribution of curb weights changes. In this case, the change is 
intuitive: As the LTV fleet becomes lighter, mass reduction among LTVs 
below 5,014 pounds becomes less detrimental to society. However, the 
incremental effect is estimated to be quite small: Shifting from an 
even mix of LTVs above and below the threshold to an extreme 20 percent 
to 80 percent split only changes the estimated net societal effect by 
0.03 percent in absolute terms. Thus, the model results for lighter 
LTVs appear relatively insensitive to the LTV curb weight distribution. 
Indeed, in the limit, where all LTVs are below the crash-partner curb 
weight threshold (and thus there are no fatality impacts for crashes 
involving heavier LTVs), the estimated net societal effect of a 100-
pound mass reduction for LTVs below 5,014 pounds (i.e., all LTVs in 
this case) is 0.25 percent, a difference of 0.06 percent in absolute 
terms compared to the baseline. This result is driven by the dominating 
effects of crash types involving either: (1) No crash partner (e.g., 
first-event rollovers); (2) one crash partner from a group not 
associated with a given change in a curb weight distribution (e.g., 
heavy vehicles, bicyclists, passenger cars); or (3) multiple crash 
partners (an element of ``all other crashes''). That is, even extreme 
changes in the distribution of curb weights for a given vehicle type 
will not change the role that vehicle mass plays in crashes for a focal 
vehicle when that vehicle does not collide with another vehicle from 
the distribution in question. In the above example involving lighter 
LTVs, 90 percent of fatalities involve incidents that do not include a 
single LTV crash partner, and 66 percent of fatalities involve 
incidents that do not include a single light-duty crash partner.
    Continuing with this example scenario, the point estimate for LTVs 
heavier than 5,014 pounds becomes larger in magnitude (i.e., more 
societally beneficial mass reduction) to a similar degree as the 
reduction in magnitude for lighter LTVs when moving to an extreme 20 
percent to 80 percent split of crash partner LTVs above (versus below 
in the case above) the curb weight threshold:

[[Page 24749]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.416

    In the NPRM and this analysis, the estimated net societal effect of 
a 100-pound mass reduction for lighter LTVs was a 20.0 fatality 
decrease, or a 0.61 percent decrease relative to a baseline of 3,304 
fatalities. Changing the share of crash fatalities involving heavier 
LTVs to be consistent with a fleet with only 20 percent of LTVs above 
the curb weight threshold yields: (1) A larger reduction in fatalities 
in crashes involving lighter LTVs per 100-pound mass reduction (from 
4.0 fatalities to 6.1 fatalities); and (2) a decrease in incremental 
fatalities in crashes involving heavier LTVs (from 1.6 fatalities to 
0.7 fatality); for a total net change of -22.9 fatalities compared to a 
baseline of -20.0 fatalities. Thus, the agencies estimate that, in a 
future year where the fleet differs from the baseline by having 80 
percent of LTVs below the crash-partner curb weight threshold, the net 
societal effect of a 100-pound mass reduction in LTVs 5,014 pounds or 
heavier would be -22.9 divided by 3,304, or -0.69 percent, versus -0.61 
percent in the baseline. Consistent with the test results for lighter 
LTVs, the model results for heavier LTVs appear relatively insensitive 
to the LTV curb weight distribution. In the limit, where all LTVs 
(except for one remaining heavier LTV in consideration) are below the 
crash-partner curb weight threshold (and thus there are no effective 
fatality impacts for crashes involving heavier LTVs), the estimated net 
societal effect of a 100-pound mass reduction for the remaining LTV 
above 5,014 pounds is -0.76 percent, a difference of 0.15 percent in 
absolute terms compared to the baseline.
    Expanding the analysis to account for changes in the relative sales 
shares of each vehicle type dampens the net effects further. As the 
fleet share of passenger cars decreases, the net effects of mass 
reduction among LTVs become less societally beneficial. That is, as 
there are fewer relatively vulnerable passenger cars in the fleet, 
there become fewer opportunities to reduce fatalities in collisions 
between LTVs and passenger cars through mass reduction. In some 
scenarios considered in the exploratory analysis, the effects of sales 
shifts from passenger cars to LTVs at least fully offset the estimated 
improvements in net fatalities associated with mass reduction 
summarized above as the LTV fleet becomes lighter.
    Ultimately, the exploratory analysis using extreme example cases 
confirmed that the baseline safety point estimates are very reasonable 
for the feasible ranges of mixes of vehicle types and curb weights 
across the model years in the CAFE model analysis. The sensitivities of 
the point estimates are relatively low across relative shares of 
lighter versus heavier LTVs (especially relative to the uncertainty in 
the baseline estimates), and similarly low and offsetting across 
decreasing fleet shares for passenger cars. Because shifts in mass in 
the rulemaking analysis would have insignificant impacts on the safety 
estimated values and therefore rulemaking decision making, the agencies 
conclude no changes are warranted for this final rule analysis.
Mass Safety Results
    Table VI-204 presents the estimated percent increase in U.S. 
societal fatality risk per 10 billion VMT for each 100-pound reduction 
in vehicle mass, while holding footprint constant, for each of the five 
vehicle classes:

[[Page 24750]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.417

    Techniques developed in the 2011 (preliminary) and 2012 (final) 
Kahane reports have been retained to test statistical significance and 
to estimate 95-percent confidence bounds (sampling error) for mass 
effects and to estimate the combined annual effect of removing 100 
pounds of mass from every vehicle (or of removing different amounts of 
mass from the various classes of vehicles), while holding footprint 
constant.
    None of the estimated effects have 95-percent confidence bounds 
that exclude zero, and thus are not statistically significant at the 
95-percent confidence level. The NPRM reported that two estimated 
effects are statistically significant at the 85-percent level. Societal 
fatality risk is estimated to: (1) Increase by 1.2 percent if mass is 
reduced by 100 pounds in the lighter cars; and (2) decrease by 0.61 
percent if mass is reduced by 100 pounds in the heavier truck-based 
LTVs. The estimated increases in societal fatality risk for mass 
reduction in the heavier cars and the lighter truck-based LTVs, and the 
estimated decrease in societal fatality risk for mass reduction in CUVs 
and minivans are not significant, even at the 85-percent confidence 
level. Although 85-percent statistical significance is not a 
traditional metric of meaningful differences to zero, this result 
confirms that the estimated effects for vehicles with curb weights most 
dissimilar to the median vehicle are the most likely to be 
significantly different to zero.
    The agencies judge the central value estimates are the best and 
most up-to-date estimates available; the estimates offer a stronger 
statistical representation of relationships among vehicle curb weight, 
footprint and fatality risk than an assumption of no correlation 
whatsoever. The agencies appropriately present the statistical 
uncertainty. For example, the central values for the highest vehicle 
weight group (LTVs 5,014 pounds or heavier) and the lowest vehicle 
weight group (passenger cars lighter than 3,201 pounds) (which, based 
on fundamental physics, are expected to have the greatest impact of 
mass reduction on safety) are economically significant,\1987\ and are 
in line with the prior analyses used in past NHTSA CAFE and EPA 
CO2 rulemakings. As shown in Table VI-205, the estimated 
coefficients have trended to lower numerical values in successive 
studies, but remain positive for lighter cars and negative for heavier 
LTVs. The 85-percent confidence level was reported only to show the 
scope of uncertainty at the first rounded (to five percent) threshold 
where the coefficient estimates were significantly different to zero 
for the two vehicle groups at the extremes of the curb weight 
distribution. No preference was suggested for an 85-percent confidence 
bound. Rather, the agencies found value in reporting confidence 
intervals for all five coefficients at the threshold where the 
estimates for the two extremes of the curb weight distribution were 
significantly different to zero. The agencies determined it was better 
to include the estimates, despite the slightly lower confidence level, 
than knowingly omitting economically significant results.
---------------------------------------------------------------------------

    \1987\ The agencies use ``economically significant results'' to 
mean values that have an important, practical implication, but may 
be derived from estimates that do not meet traditional levels of 
statistical significance. For example, if the projected economic 
benefit of a project equaled $100 billion, the agencies would 
consider the impact economically significant, even if the estimates 
used to derive the impact were not statistically significant at the 
95-percent confidence level. Conversely, if the projected economic 
benefit of a project equaled $1, the agencies would not consider the 
impact economically significant, even if the estimates used to 
derive the impact were statistically significant at the 99.99-
percent confidence level. In the case above, we considered the 
results associated with the lightest and heaviest vehicle types to 
be economically significant because the associated safety costs were 
large and the estimates had magnitudes meaningfully different from 
zero and were statistical significant at the 85-percent confidence 
level.
---------------------------------------------------------------------------

    The regression results are constructed to project the effect of 
changes in mass, independent of all other factors, including footprint. 
With each additional change from the current environment (e.g., the 
scale of mass change, presence and prevalence of safety features, 
demographic characteristics), the results may become less 
representative. That is, although safety features and demographic 
factors are accounted for separately, the estimated effects of mass are 
identified under the specific mix of vehicles and drivers in the data. 
The agencies note that the analysis accounts for safety features that 
are optional but available across all MYs in the sample (most notably 
electronic stability control, which was not yet mandatory for all model 
years in the sample), and calibrates historical safety data to account 
for future fleets with full ESC penetration to reflect the mandate.
    The agencies considered the near multicollinearity of mass and 
footprint to be a major issue in the 2010 Kahane report and voiced 
concern about inaccurately estimated regression coefficients. High 
correlations between mass and footprint and variance inflation factors 
(VIF) have persisted from MY 1991-1999 to MY 2004-2011; large footprint 
vehicles continued to be, on the average, heavier than small footprint 
vehicles to the same extent as in the previous decade.
    Nevertheless, multicollinearity appears to have become less of a 
problem in the 2012 Kahane, 2016 Puckett and Kindelberger/Draft TAR, 
and current analyses. Ultimately, only three of the 27 core models of 
fatality risk by vehicle type in the current analysis indicate the 
potential presence of effects of multicollinearity, with estimated 
effects of mass and footprint

[[Page 24751]]

reduction greater than two percent per 100-pound mass reduction and 
one-square-foot footprint reduction, respectively; these three models 
include passenger cars and CUVs in first-event rollovers, and CUVs in 
collisions with LTVs greater than 4,360 pounds. This result is 
consistent with the 2016 Puckett and Kindelberger report, which also 
found only three cases out of 27 models with estimated effects of mass 
and footprint reduction greater than two percent per 100-pound mass 
reduction and one-square-foot footprint reduction.
    For comparison, Table VI-205 shows the fatality coefficients from 
the 2012 Kahane report (MY 2000-2007 vehicles in CY 2002-2008) and the 
2016 Puckett and Kindelberger report and Draft TAR (MY 2003-2010 
vehicles in CY 2005-2011).
[GRAPHIC] [TIFF OMITTED] TR30AP20.418

    The new results are directionally the same as in 2012; in the 2016 
analysis, the estimate for lighter LTVs was of opposite sign (but small 
magnitude). Consistent with the 2012 Kahane and 2016 Puckett and 
Kindelberger reports, mass reductions in lighter cars are estimated to 
lead to increases in fatalities, and mass reductions in heavier LTVs 
are estimated to lead to decreases in fatalities.
---------------------------------------------------------------------------

    \1988\ Median curb weights in the 2012 Kahane report: 3,106 
pounds for cars, 4,594 pounds for truck-based LTVs. Median curb 
weights in the 2016 Puckett and Kindelberger report: 3,197 pounds 
for cars, 4,947 pounds for truck-based LTVs.
---------------------------------------------------------------------------

    The estimated mass effect for heavier truck-based LTVs is stronger 
in this analysis and in the 2016 Puckett and Kindelberger report than 
in the 2012 Kahane report; both estimates are statistically significant 
at the 85-percent confidence level, unlike the corresponding estimate 
in the 2012 Kahane report. The estimated mass effect for lighter truck-
based LTVs is insignificant and positive in this analysis and the 2012 
Kahane report, while the corresponding estimate in the 2016 Puckett and 
Kindelberger report was insignificant and negative.
    Multiple commenters, including the South Coast Air Quality 
Management District and States and Cities, challenged the practical 
value of using estimates with statistical significance at the 85-
percent level, arguing that below 95 (or 90) percent are insufficiently 
reliable.\1989\ For example, CARB stated, ``[d]ue to the lack of 
statistical significance, NHTSA should not be attributing any increase 
in fatalities due to mass reduction'' and argues that the ``effect of 
mass reduction on fatality risk should be set to zero since the 
estimates are not statistically different to zero.'' \1990\
---------------------------------------------------------------------------

    \1989\ See South Coast Air Quality Management District, Detailed 
Comments, Docket No. NHTSA-2018-0067-11813, at 6 (internal citation 
omitted); States and Cities, Detailed Comments, Docket No. NHTSA-
2018-0067-11735, at 95.
    \1990\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 269.
---------------------------------------------------------------------------

    The agencies believe the updated analysis that was presented in the 
NPRM represents the most up to date and best estimate of the impacts of 
mass reduction on crash fatalities; and, that it is appropriate for the 
analysis to use the best and most likely estimates for safety, even if 
the estimates are not statistically significant at the 95-percent 
confidence level. Significance at the 85-percent confidence level is 
important evidence that the relevant point estimates are meaningfully 
different to zero (e.g., approximately five to six times more likely to 
be non-zero than zero). The agencies believe it would be misleading to 
ignore these data or to use values of zero for the rulemaking analysis, 
as doing so would not properly inform decision makers on the safety 
impacts of the regulatory alternatives and final standards. Similar to 
past analyses, the NPRM and this final rule analysis use the best 
available estimates. The agencies feel it is inappropriate to ignore 
likely impacts of the standards simply because the best available 
estimates have confidence levels below 95 percent; uniform estimates of 
zero are statistically weaker than the estimates identified in the 
analysis, and thus are not the best available. Because the point 
estimates are derived from the best-fitting estimates for each crash 
type (all of which are non-zero), the confidence bounds around an 
overall estimate of zero would necessarily be larger than the 
corresponding confidence bounds around the point estimates presented 
here.
    The sensitivity analysis in Section VII.C Sensitivity Analysis 
provides an evaluation of extreme cases in which all of the estimated 
net fatality rate impacts of mass reduction are either at their fifth- 
or 95th-percentile values. The range of net impacts in the sensitivity 
analysis not only covers the relatively more likely case that 
uncertain, yet

[[Page 24752]]

generally offsetting, effects are distinct from the central estimates 
considered here (e.g., in a plausible case where mass reduction in the 
heaviest LTVs is less beneficial than indicated by the central 
estimates, it would also be relatively likely that mass reduction in 
the lightest passenger cars would be less harmful, yielding a similar 
net impact), but also covers the relatively unlikely case that all of 
the estimates are uncertain in the same direction.
    At a broader level, multiple commenters asserted that the role of 
safety-related estimates should be restricted because of what they 
claim is a weak historical relationship between fuel economy and 
vehicle safety. For example, the Green Energy Institute at Lewis & 
Clark Law School commented, ``[o]ver the past 40 years, per-capita 
vehicle fatalities decreased by 50%, while average fuel economy 
doubled.'' \1991\ However, this statistic is misleading because it does 
not account for vehicle safety factors and changes in driving behavior 
external to fuel economy (e.g., FMVSS and other safe design advances, 
reductions in drunk driving, increases in seat belt use). That is, 
fatality rates have decreased due to a range of factors that are 
unrelated to fuel economy efforts. The methodology in the 2012 Kahane 
report, the 2016 Puckett and Kindelberger, the Draft TAR, the 2018 NPRM 
analysis and today's final rule analysis addresses these other changes 
in order to isolate the impacts of mass reduction alone. The role of 
the safety analysis outlined in this document is to isolate incremental 
effects on safety outcomes that are related to changes in fuel economy.
---------------------------------------------------------------------------

    \1991\ Green Energy Institute at Lewis & Clark Law School, 
Docket No. NHTSA-2018-0067-12213, at 3.
---------------------------------------------------------------------------

    Multiple commenters disagreed with the results in Table VI-204, 
maintaining that mass reduction need not reduce societal safety. EDF 
cited a Michigan Manufacturing Technology Center (MMTC) review as 
supporting that widespread lightweighting would decrease crash severity 
through reduced kinetic energy in multiple-vehicle crashes. Similarly, 
the Aluminum Association commented, ``[v]ehicle size, not weight, has 
been shown to be the leading safety determinant.'' \1992\ Other 
commenters cited Anderson and Auffhammer (2014), which finds that the 
safety effects of mass reduction in one vehicle are offset by the 
safety effects in the crash partner vehicle.\1993\ The South Coast Air 
Quality Management District asserted that NHTSA and EPA appear to argue 
``that fuel-efficient vehicles are lighter than other vehicles, and 
therefore, less safe.'' The North Carolina Department of Environmental 
Quality asserted that a takeaway from the preferred alternative is that 
larger vehicles are safer than smaller vehicles. The agencies' 
conclusion is that, at the societal level, it is the distribution of 
changes in vehicle mass that matter (i.e., mitigating mass reduction in 
the lightest vehicles is societally beneficial, while mitigating mass 
reduction in the heaviest vehicles is societally harmful).
---------------------------------------------------------------------------

    \1992\ The Aluminum Association, Detailed Comments, Docket No. 
NHTSA-2018-0067-12213, at 3.
    \1993\ Anderson, M.L. and M. Auffhammer (2014). ``Pounds that 
Kill,'' Review of Economic Studies, Vol. 81, No. 2, pp. 535-71.
---------------------------------------------------------------------------

    The 2012 Kahane report, the 2016 Puckett and Kindelberger, the 
Draft TAR, the 2018 NPRM analysis and today's final rule analysis all 
have shown that both mass and vehicle size impact societal safety. 
Across recent rulemakings, the analyses have confirmed a protective 
effect of vehicle size (i.e., societal fatality risk decreases as 
footprint increases). As mentioned previously, the agencies believe 
vehicle footprint-based standards help to discourage vehicle 
manufacturers from downsizing their vehicles, and therefore assume 
changes in CAFE and CO2 standards will not impact vehicle 
size and size-related safety impacts. On the other hand, mass reduction 
is a cost-effective technology for increasing fuel economy and reducing 
CO2 emissions. Therefore, the agencies do include the 
assessment of safety impacts related to mass reduction. As discussed 
throughout this mass-safety subsection, the agencies present 
comprehensive consideration of the various studies and workshops on the 
impact of vehicle mass on safety, and conclude there is in fact a 
relationship. The fleet simulation study, discussed in the next 
subsection, further supports the existence of this relationship and 
that this relationship will continue to exist in future vehicle 
designs.
    The principal difference between heavier vehicles, especially 
truck-based LTVs, and lighter vehicles, especially passenger cars, is 
that mass reduction has a different effect in collisions with another 
car, LTV, or other object such as a lamp post. When two vehicles of 
unequal mass collide, the change in velocity (delta-V) is greater in 
the lighter vehicle. Through conservation of momentum, the degree to 
which the delta-V in the lighter vehicle is greater than in the heavier 
vehicle is proportional to the ratio of mass in the heavier vehicle to 
mass in the lighter vehicle:
[GRAPHIC] [TIFF OMITTED] TR30AP20.419

[GRAPHIC] [TIFF OMITTED] TR30AP20.420

    Because fatality risk is a positive function of delta-V, the 
fatality risk in the lighter vehicle in two-vehicle collisions is also 
higher. Vehicle design can reduce the magnitude of delta-V to some 
degree (e.g., changing the stiffness

[[Page 24753]]

of a vehicle's structure could dampen delta-V for both crash partners). 
These considerations drive the overall result: mass reduction is 
associated with an increase in fatality risk in lighter cars, a 
decrease in fatality risk in heavier LTVs, CUVs, and minivans, and has 
smaller effects in the intermediate groups. Mass reduction may also be 
harmful in a crash with a movable object such as a small tree, which 
may break if hit by a high mass vehicle resulting in a lower delta-V 
than may occur if hit by a lower mass vehicle which does not break the 
tree and therefore has a higher delta-V. However, in some types of 
crashes not involving collisions between cars and LTVs, especially 
first-event rollovers and impacts with fixed objects, mass reduction 
may not be harmful and may even be beneficial.
    Ultimately, delta-V is a direct function of relative vehicle mass 
for given vehicle structures. Removing some mass from the heavier 
vehicle involved in an accident with a lighter vehicle reduces the 
delta-V in the lighter vehicle, where fatality risk is higher, 
resulting in a large benefit to the passengers of the lighter vehicle. 
This is partially offset by a small increase in the delta-V in the 
heavy vehicle; however, the fatality risk is lower in the heavier 
vehicle and remains relatively low despite the increase in delta-V. In 
sum, the change in mass and delta-V from mass reduction in heavier 
vehicles results in a net societal benefit.
    Multiple commenters claimed that the agencies' analysis does not 
allow for the likely outcome that mass reduction would be concentrated 
among relatively heavy vehicles.\1994\ For example, Global Automakers 
commented that the agencies should not include weight reduction in 
their safety analysis because ``very few vehicles [have] implemented 
lightweight material substitution strategies.'' \1995\
---------------------------------------------------------------------------

    \1994\ See also, e.g., South Coast Air Quality Management 
District, Detailed Comments, Docket No. NHTSA-2018-0067-11813, at 6.
    \1995\ Association of Global Automakers, Attachment A, Docket 
No. NHTSA-2018-0067-12032, at A-32.
---------------------------------------------------------------------------

    Neither CAFE standards nor this analysis mandate mass reduction, or 
mandate mass reduction occur in any specific manner. However, mass 
reduction is a highly cost effective technology for improving fuel 
economy and CO2 emissions. The steel, aluminum, plastics, 
composite, and other material industries are developing new materials 
and manufacturing equipment and facilities to produce those materials. 
In addition, suppliers and manufacturers are optimizing designs to 
maintain or improve functional performance with lower mass. 
Manufacturers have stated that they will continue to reduce vehicle 
mass to meet more stringent standards, and therefore, this expectation 
is incorporated into the modeling analysis supporting the standards to: 
(1) Determine capabilities of manufacturers; and (2) to predict costs 
and fuel consumption effects of CAFE standards. The CAFE and 
CO2 rulemakings in 2012, the Draft TAR and EPA Preliminary 
Determination, imposed an artificial constraint on vehicle mass 
reduction to achieve a desired safety-neutral outcome. For the current 
rulemaking, this artificial constraint is eliminated so the analysis 
reflects manufacturers applying the most cost effective technologies to 
achieve compliance with the regulatory alternatives and the final 
standards; this approach allows mass reduction to be applied across the 
fleet. This is consistent with industry trends.\1996\ To the extent 
that mass reduction is only cost-effective for the heaviest vehicles, 
the CAFE model would create the outcome predicted by commenters. In 
reality, however, mass reduction is a cost-effective means of improving 
fuel economy and does take place across vehicles of all sizes and 
weights. Accordingly, the model reflects that manufacturers may reduce 
vehicle mass--regardless of vehicle class--when doing so is cost 
effective.
---------------------------------------------------------------------------

    \1996\ The baseline MY 2016 (for the NPRM) and MY 2017 (for this 
final rule analysis) vehicle fleet data show manufacturers have in 
fact implemented mass reduction technology across vehicle types and 
sizes- including smaller and lighter vehicles.
---------------------------------------------------------------------------

    The National Tribal Air Association claimed the 2015 NAS study 
found ``evidence suggest[ing] that the [2012] standards will lead the 
nation's light-duty vehicle fleet to become lighter but not less 
safe.'' \1997\ The agencies note the NAS quote is one phrase from the 
press release that accompanied the NHTSA sponsored 2015 NAS 
study,\1998\ and the agencies do not believe the phrase in isolation 
reflects the findings of the NAS Committee, which are discussed in over 
3 pages of the report.\1999\ The 2015 NAS report supported the 
analytical methodology used for the 2012 NHTSA CAFE and EPA 
CO2 rulemaking and found it reasonable. As discussed in the 
subsections further above, a nearly identical methodology was used for 
the NPRM analysis and for this final rule.
---------------------------------------------------------------------------

    \1997\ National Tribal Air Association, Detailed Comments, 
Docket No. NHTSA-2018-0067-11948, at 2.
    \1998\ NAS (2015). Press Release. ``Analysis Used by Federal 
Agencies to Set Fuel Economy and Greenhouse Gas Standards for U.S. 
Cars Was Generally of High Quality; Some Technologies and Issues 
Should Be Re-examined.'' June 18, 2015. Available at http://www8.nationalacademies.org/onpinews/newsitem.aspx?RecordID=21744.
    \1999\ Key excerpts from the report include: ``[o]ccupants of 
smaller vehicles are at a greater risk of fatality in crashes, 
particularly in a crash with a vehicle of greater mass;'' and 
``[t]he 2012 studies (by NHTSA, Lawrence Berkeley National 
Laboratories, and Dynamic Research, Inc.) indicate that mass 
reduction while holding footprint constant is associated with a 
small increase in risk for lighter-than-average cars only; the 
estimated effect on other vehicle types is not statistically 
significant.'' National Research Council (2015). Cost, 
Effectiveness, and Deployment of Fuel Economy Technologies for 
Light-Duty Vehicles, available at https://doi.org/10.17226/21744. 
pp. 224-28.
---------------------------------------------------------------------------

    The agencies received several comments about the relationship 
between mass and crash avoidance. The NRDC commented that the analysis 
should account for the expected result that mass reduction makes it 
easier to avoid crashes.\2000\ Conversely, IPI quoted a finding by LNL 
that ``found that mass reductions may increase the number of accidents 
but that each crash results in fewer fatalities.'' \2001\
---------------------------------------------------------------------------

    \2000\ NRDC, Detailed Comments, Docket No. NHTSA-2018-0067-
11973.
    \2001\ IPI, Detailed Comments, Docket No. NHTSA-2018-0067-12213, 
at 129.
---------------------------------------------------------------------------

    The phenomenon touched upon by IPI and NRDC has been identified in 
past rulemakings as well, and highlights that the relationship between 
mass reduction and societal fatality risk include two partially-
offsetting components (i.e., increased exposure to crashes is offset 
partially by decreased risk in some vehicles conditional on a crash 
occurring). The agencies note that this relationship, while not 
reported separately, is in fact embedded within the analysis detailed 
in this document, as the extent to which some vehicles are more 
maneuverable and faster-braking, the crash data reflect those 
characteristics through lower observed fatality rates. However, when 
considering the purposes of estimating effects of mass reduction on 
fatalities, it is immaterial what share of the effect is comprised of 
crash avoidance factors and crashworthiness factors, the ultimate 
effect is present within the data evaluated in the analysis. The mass-
safety impacts estimated by the statistical analysis of crash data are 
based on the safety technologies and mass levels present among the 
vehicle fleets for the calendar and model years in the data. As 
discussed below in this section, the analysis separately accounts for 
the effects of future safety technologies.
(4) Sensitivity Analysis
    Table VI-206 shows the principal findings and includes sampling-
error

[[Page 24754]]

confidence bounds for the five parameters used in the CAFE model. The 
confidence bounds represent the statistical uncertainty that is a 
consequence of having less than a census of data. NHTSA's 2011, 2012, 
and 2016 reports acknowledged another source of uncertainty: The 
central (baseline) statistical model can be varied by choosing 
different control variables or redefining the vehicle classes or crash 
types, which for example, could produce different point estimates.
    Beginning with the 2012 Kahane report, NHTSA has provided results 
of 11 plausible alternative models that serve as sensitivity tests of 
the baseline model. Each alternative model was tested or proposed by: 
Farmer (IIHS) or Green (UMTRI) in their peer reviews; Van Auken (DRI) 
in his public comments; or Wenzel in his parallel research for DOE. The 
2012 Kahane and 2016 Puckett and Kindelberger reports provide further 
discussion of the models and the rationales behind them.
    Alternative models use NHTSA's databases and regression-analysis 
approach but differ from the central model in one or more explanatory 
variables, assumptions, or data restrictions. The agencies applied the 
11 techniques to the latest databases to generate alternative CAFE 
model coefficients. The range of estimates produced by the sensitivity 
tests offers insight to the uncertainty inherent in the formulation of 
the models, subject to the caveat that these 11 tests are, of course, 
not an exhaustive list of conceivable alternatives.
    The central and alternative results follow, ordered from the lowest 
to the highest estimated increase in societal risk per 100-pound 
reduction for cars weighing less than 3,201 pounds:
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.421


[[Page 24755]]


BILLING CODE 4910-59-C
    The sensitivity tests illustrate both the fragility and the 
robustness of central estimates. On the one hand, the variation among 
the coefficients is quite large relative to the central estimate: In 
the preceding example of cars < 3,201 pounds, the estimated 
coefficients range from almost zero to almost double the central 
estimate. This result underscores the key relationship that the 
societal effect of mass reduction is small. In other words, varying how 
to model some of these other vehicle, driver, and crash factors, which 
is exactly what sensitivity tests do, can appreciably change the 
estimate of the societal effect of mass reduction.
    On the other hand, variations are not particularly large in 
absolute terms. The ranges of alternative estimates are generally in 
line with the sampling-error confidence bounds for the central 
estimates. Generally, in alternative models as in the central model, 
mass reduction tends to be relatively more harmful in the lighter 
vehicles and more beneficial in the heavier vehicles, just as they are 
in the central analysis. In all models, the point estimate of the 
coefficient is positive for the lightest vehicle class, cars < 3,201 
pounds. In 10 out of 11 models, the point estimate is negative for CUVs 
and minivans, and in nine out of 11 models the point estimate is 
negative for LTVs >= 5,014 pounds. The agencies believe the central 
case uses the most rigorous methodology, as discussed further above, 
and provides the best estimates of the impacts of mass reduction on 
safety.
    Tom Wenzel commented confirming a preference for the alternative 
model with footprint separated into track width and wheelbase, and with 
the induced exposure data limited to stopped vehicle cases.\2002\ 
Wenzel asserts that splitting footprint into its components reduces 
multicollinearity with curb weight, and that limiting induced exposure 
cases to stopped vehicles mitigates bias against driver-vehicle pairs 
that are less likely to be involved in crashes. Based on this feedback 
and the intuitiveness of the approach, the agencies further considered 
the alternative model with footprint split into track width and 
wheelbase. Consistent with previous analyses and assessments, there are 
problems with splitting footprint into its components within the mass-
size-safety models because of strong correlations among curb weight, 
track width and wheelbase. For all vehicle classes in the analysis, 
curb weight is correlated either nearly as high or higher with track 
width as with footprint. Track width and wheelbase are also highly 
correlated with one another (ranging from around 0.64 to 0.80, with the 
exceptions of smaller correlations for large pickups and minivans). 
Viewed from another angle, wheelbase is almost perfectly correlated 
with footprint (with correlations ranging from around 0.95 to 0.97).
---------------------------------------------------------------------------

    \2002\ Wenzel, T., Lawrence Berkeley National Laboratories, 
Docket No. EPA-HQ-OAR-2018-0283-4118.
---------------------------------------------------------------------------

    Considered in concert, the track width and wheelbase model not only 
essentially incorporates the full correlation issues from the baseline 
model (curb weight highly correlated with another independent 
variable), but also adds a further correlation issue (the variable that 
is highly correlated with curb weight is also highly correlated with a 
separate independent variable). The agencies examined supplementary 
means of confirming the relative methodological merit of the footprint-
based model and the track-width-wheelbase-based alternative. The 
supplementary analysis centered on the condition index, which 
quantifies the invertibility of the matrix of independent variables in 
a given model through its measure, the condition number.\2003\ A model 
with a low condition number has relatively low correlations among its 
independent variables, and thus its invertibility and the corresponding 
model outputs are robust to variations in model input values. A model 
with a high condition number has relatively high correlations among its 
independent variables, and thus its invertibility and model outputs are 
not robust to variations in model input values. That is, a model with a 
high condition number is likely to be subject to the problems 
associated with multicollinearity. Although there is no strict 
threshold condition number value to indicate multicollinearity, higher 
values indicate greater likelihood that the independent variables are 
correlated to a problematic degree.
---------------------------------------------------------------------------

    \2003\ See Belsley, D.A., Kuh, E., and Welsch, R.E. (1980). 
``The Condition Number.'' Regression Diagnostics: Identifying 
Influential Data and Sources of Collinearity. New York: John Wiley & 
Sons; Freund, R.J. and Littell, R.C. (2000). SAS System for 
Regression, Third Edition. Cary, NC: SAS Institute, Inc.; and 
Hallahan, C. (1995). ``Understanding the Multicollinearity 
Diagnostics in SAS/Insight and Proc Reg.'' SAS Conference 
Proceedings, Washington, DC, October 8-10, 1995.
---------------------------------------------------------------------------

    The condition index offers an alternative means of capturing the 
same forces as the variance inflation factor (VIF), which the agencies 
have used historically (including in this rulemaking) as a diagnostic 
of multicollinearity. However, the condition index offers some 
advantages relative to the VIF. Notably, the condition index applies 
regardless of the econometric form of the model (i.e., the 
decomposition of the independent variables is the same regardless of 
how the variables are applied in the model). This is distinct from the 
VIF, which is limited to a linear diagnostic of the data that may not 
map well to non-linear econometric models, including the logistic 
regression models that form the core of the curb weight-fatality risk 
analysis. The condition index estimates the incremental effects of 
individual variables, which is helpful in an analysis of which 
independent variables are the most problematic. Conversely, the 
diagnostic values from the VIF are not necessarily sensitive to 
incremental correlated variables, as the VIF value (1/(1-R\2\) does not 
necessarily change much once correlations are relatively high (i.e., 
when R\2\ is already high, the inclusion of one or more highly 
correlated variables may not change R\2\, and in turn, the VIF, by 
much.
    An incremental comparison of VIF estimates for the data confirmed 
the potential weakness of the VIF in this case. For the CUV-minivan 
model data, the VIF decreases from 9.4 to 6.7 when: (1) Substituting 
either track width or footprint for footprint that has an identical 
correlation with curb weight as footprint; and (2) adding the other 
component of footprint. This result is counterintuitive (i.e., the 
simpler model should necessarily have fewer issues of 
multicollinearity), and may be an artifact of differences in model fit 
(e.g., a higher R\2\ in the simpler model could indicate better model 
fit rather than anything problematic in terms of correlation 
structure). This result led the agencies to question how well the VIF 
identifies relative impacts of multicollinearity across related models, 
especially in non-linear applications.
    The calculated condition numbers for the curb weight-footprint 
models and their corresponding curb weight-wheelbase-track width 
alternatives were consistent with expectations regarding 
multicollinearity, however. The condition numbers for the curb weight-
wheelbase-track width models are approximately two to three times 
higher than the condition numbers for the curb weight-footprint models. 
This indicates that the level of imprecision in model estimates using 
track width and wheelbase would be expected to be between approximately 
two to three times higher than in the baseline models using footprint. 
Unlike the VIF, the condition index supports a hypothesis that 
multicollinearity would not be mitigated in an alternative with 
disaggregated variables that are highly

[[Page 24756]]

correlated with both the variable of interest and the variable they are 
replacing. Considering these results, the agencies that using footprint 
to represent vehicle size in the safety models provides a more reliable 
estimate of safety impacts than splitting footprint into track width 
and wheelbase.
    The agencies also considered the use of stopped-vehicle data as an 
alternative. The primary problem with this approach is that the 
agencies do not observe as large of a share of cases on roads with 
higher travel speeds (e.g., interstate highways) when including only 
stopped vehicles; this relationship influences the extent to which the 
induced exposure data reflect the distributions of driver attributes 
and contextual effects across national VMT. Based on this assessment, 
the agencies believe the methodology used for the analysis in the 
proposal provides a more reliable and representative estimate of safety 
impacts, and thus is not changing the methodology for today's final 
rule.
    In a related comment, Wenzel proposes that future analyses should 
directly account for differences in curb weight between vehicles in 
two-vehicle crashes. The agencies believe that would require the 
development of a model that directly accounts for the relative weights 
of vehicles in two-vehicle crashes, and that such a model would require 
peer review. Key alternatives to test would vary in terms of the 
functional form of the mass disparity between two crash partners (e.g., 
a relative mass ratio consistent with the delta-V calculation presented 
above, linear mass difference, non-linear mass difference). The 
agencies will consider initiating work to explore such a model in the 
future.
    DRI requested the agencies clarify whether the analysis accounts 
for all road users (i.e., including pedestrians, bicyclists, 
motorcyclists, and other crash partners), while the Pennsylvania 
Department of Environmental Protection commented, ``[i]t is inadequate 
for the agencies' analysis for this Proposed Rule to only focus on 
frontal crashes while omitting near-frontal collisions, side-impact 
collisions, rear-end collisions, rollover accidents, impacts with 
stationary objects and accidents involving pedestrians.'' \2004\ The 
agencies confirm that the analysis presented in this section continues 
to apply the methodology developed by Kahane, which incorporates all 
road users, without double-counting, to identify societal fatality rate 
impacts. Because every fatal crash (across crash types) is included in 
the analysis, not just frontal crashes, the agencies find this comment 
lacks a basis. The agencies believe the commenter's confusion may stem 
from the use of front-to-back crashes to generate estimates of the 
proportions of all driving for each vehicle model associated with 
particular characteristics of drivers (e.g., age, gender) and crashes 
(e.g., urban/rural, day/night). These crashes represent the best 
available trade-off among sample size, representativeness of overall 
vehicle and driver exposure, and mitigating bias in a sample that is 
intended to be effectively random (i.e., the probability of being 
struck from behind by an at-fault driver is assumed to be a function of 
characteristics of other drivers and travel demand, but not of the 
struck driver or the struck vehicle).
---------------------------------------------------------------------------

    \2004\ Pennsylvania Department of Environmental Protection, 
Detailed Comments, Docket No. NHTSA-2018-0067-11956, at 9.
---------------------------------------------------------------------------

(5) Fleet Simulation Study
    Commenters to recent CAFE rulemakings, including some vehicle 
manufacturers, have suggested designs and materials of more recent 
model year vehicles may have weakened the historical statistical 
relationships between mass, size, and safety. NHTSA and EPA agreed that 
the statistical analysis would be improved by using an updated crash 
and exposure database reflecting more recent safety technologies, 
vehicle designs and materials, and reflecting changes in the vehicle 
fleet. As mentioned above, a new crash and exposure database was 
created with the intention of capturing modern vehicle engineering and 
has been employed for assessing safety effects for CAFE rules since 
2012.
    The agencies have traditionally relied solely on real-world crash 
data as the basis for projecting the future safety implications for 
regulatory changes. The agencies are required to consider relevant data 
in setting standards.\2005\ Every fleet regulated by the agencies' 
standards differs from the fleet used to establish said standard, and 
as such, the light-duty vehicle fleet in the MY 2021-2026 timeframe 
will be different from the MY 2004-2011 fleet analyzed above. This is 
not a new or unique phenomenon, but instead is an inherent challenge in 
regulating an industry reliant on continual innovation. This is the 
agencies' sixth evaluation of effects of mass reduction and/or 
downsizing,\2006\ comprising databases ranging from MYs 1985 to 2011. 
Despite continual claims that modern lightweight engineering will 
render current data obsolete, results of the six studies, while not 
identical, have been generally consistent in showing a small, negative 
impact related to mass reduction. The agencies strongly believe that 
real-world crash data remains the best, relevant data to measure the 
effect of mass reduction on safety.
---------------------------------------------------------------------------

    \2005\ See Center for Biological Diversity v. NHTSA, 538 F.3d 
1172, 1203 (9th Cir. 2008).
    \2006\ As outlined throughout this section, NHTSA's six related 
studies include the new analysis supporting this rulemaking, and: 
Kahane, C.J. Vehicle Weight, Fatality Risk and Crash Compatibility 
of Model Year 1991-99 Passenger Cars and Light Trucks, National 
Highway Traffic Safety Administration (Oct. 2003), available at 
https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/809662; 
Kahane, C.J. Relationships Between Fatality Risk, Mass, and 
Footprint in Model Year 1991-1999 and Other Passenger Cars and LTVs 
(Mar. 24, 2010), in Final Regulatory Impact Analysis: Corporate 
Average Fuel Economy for MY 2012-MY 2016 Passenger Cars and Light 
Trucks, National Highway Traffic Safety Administration (Mar. 2010) 
at 464-542; Kahane, C.J. Relationships Between Fatality Risk, Mass, 
and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--
Preliminary Report, National Highway Traffic Safety Administration 
(Nov. 2011), available at Docket ID NHTSA-2010-0152-0023); Kahane, 
C.J. Relationships Between Fatality Risk, Mass, and Footprint in 
Model Year 2000-2007 Passenger Cars and LTVs: Final Report, NHTSA 
Technical Report. Washington, DC: NHTSA, Report No. DOT-HS-811-665; 
and Puckett, S.M., & Kindelberger, J.C. Relationships between 
Fatality Risk, Mass, and Footprint in Model Year 2003-2010 Passenger 
Cars and LTVs--Preliminary Report, National Highway Traffic Safety 
Administration (June 2016), available at https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/2016-prelim-relationship-fatalityrisk-mass-footprint-2003-10.pdf.
---------------------------------------------------------------------------

    However, because lightweight vehicle designs introduce fundamental 
changes to the structure of the vehicle, there remains a persistent 
question of whether historical safety trends will apply. To address 
this concern and to verify that real-world crash data remain an 
appropriate source of data for projecting mass-safety relationships in 
the future fleet, in 2014, NHTSA sponsored research to develop an 
approach to utilize experimental lightweight vehicle designs to 
evaluate safety in a broader range of real-world representative 
crashes.\2007\ NHTSA contracted with George Washington University to 
perform a fleet simulation model to study the impact and relationship 
of light-weighted vehicle design with injuries and fatalities.\2008\ 
The study involved simulating crashes on eight test vehicles, five of 
which were equipped with lightweight materials

[[Page 24757]]

and advanced designs not yet incorporated into the U.S. fleet. The 
study assessed a range of frontal crashes, including crashes with fixed 
objects and other vehicles, across wide range of vehicle speeds, and 
with mid-size male and mid-size female dummies.\2009\ In all, more than 
440 vehicle crashes with 1,520 dummy passengers were simulated for a 
range of crash speeds and crash configurations. Results from the fleet 
simulation study showed the trend of increased societal injury risk for 
light-weighted vehicle designs occurs for both single vehicle and two-
vehicle crashes. Results are listed in Table VI-207.\2010\
---------------------------------------------------------------------------

    \2007\ See also 83 FR at 43133 (Aug 24, 2018).
    \2008\ Samaha, R.R., Prasad, P., Marzougui, D., Cui, C., Digges, 
K., Summers, S., Patel S., Zhao, L., & Barsan-Anelli, A. (2014, 
August). Methodology for evaluating fleet protection of new vehicle 
designs--Application to lightweight vehicle designs. Report No. DOT 
HS 812 051A, Washington, DC--National Highway Traffic Safety 
Administration.
    \2009\ Regulatory and consumer information crash safety tests 
are performed at high speeds, and the dummy occupant is generally a 
mid-size male. In the real world, crashes occur at various impact 
velocities and configurations; with various impact partners (e.g., 
rigid obstacles, lighter or heavier vehicles); and involve occupants 
of various sizes and ages.
    \2010\ This fleet simulation study does not provide information 
that can be used to modify coefficients derived for the NPRM 
regression analysis because of the restricted types of crashes and 
vehicle designs. Additionally, the fleet simulation study assumed 
restraint equipment to be as in the baseline model, in which 
restraints/airbags are not redesigned to be optimal with light-
weighting.
[GRAPHIC] [TIFF OMITTED] TR30AP20.422

[GRAPHIC] [TIFF OMITTED] TR30AP20.423

    The change in the safety risk from the fleet simulation study was 
directionally consistent with results for passenger cars from the 2012 
Kahane report,\2011\ the 2016 Puckett and Kindelberger report, and the 
analysis used for the proposal and today's final rule. As noted, fleet 
simulations were performed in frontal crash mode and did not consider 
other crash modes such as rollover crashes.\2012\ The fleet simulation 
analysis confirmed that real-world crash data were still a reliable 
source for analyzing mass safety impacts.
---------------------------------------------------------------------------

    \2011\ The 2012 Kahane study considered only fatalities, 
whereas, the fleet simulation study considered severe (AIS 3+) 
injuries and fatalities (DOT HS 811 665).
    \2012\ The risk assessment for CUV in the regression model 
combined CUVs and minivans in all crash modes and included belted 
and unbelted occupants.
---------------------------------------------------------------------------

    Despite the results of the fleet simulation analysis, which was 
republished in the proposal, the agencies received additional comments 
questioning the assumption that relationships among vehicle mass, size, 
and fatality risk will continue in the future. For example, the 
Alliance for Vehicle Efficiency asserted that using lighter frame 
materials has no impact on safety, noting that any mass reduction 
strategies are applied to components that are unrelated to crash safety 
and crash ratings have not declined for vehicles over the past five 
years.\2013\ CARB commented that the agencies did not account for new 
vehicle improvements and claimed the data used for the analysis was 
``not a good indicator of the safety performance of future purpose-
designed lightweighted vehicles.'' \2014\ Consumers Union offered a 
similar appraisal, indicating that the MYs in the sample are ``unlikely 
to capture the current and future mass/fatality relationship of modern 
vehicles.'' \2015\ While the Aluminum Association commented vehicle 
size, not mass, is the only physical feature that impacts safety.\2016\ 
The American Chemistry Council, Hyundai, and Tesla commented that it is 
feasible to utilize

[[Page 24758]]

design improvements and technologies to offset the incremental risk for 
vehicle occupants associated with mass reduction.\2017\ EDF said the 
mass-safety analysis did not agree with conclusions from a study by the 
Michigan Manufacturing Technology Center.\2018\ Comments from States 
and Cities, American Honda, ICCT, and NRDC shared these 
sentiments.\2019\
---------------------------------------------------------------------------

    \2013\ Alliance for Vehicle Efficiency, Detailed Comments, 
Docket No. NHTSA-2018-0067-11696, at 11.
    \2014\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 270.
    \2015\ Consumers Union, Detailed Comments, Docket No. NHTSA-
2018-0067-12068, at 18.
    \2016\ Aluminum Association, Detailed Comments, Docket No. 
NHTSA-2018-0067-11952, at 3.
    \2017\ American Chemistry Council, Detailed Comments, Docket No. 
EPA-HQ-OAR-2018-0283-1415, at 2-8; Hyundai-Kia America Technical 
Center, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-4411, at 
13; Tesla, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-4186, 
at 21-23.
    \2018\ Michigan Manufacturing Technology Center study ``Vehicle 
Lightweighting: A Review of the Safety of Reduced Weight Passenger 
Cars and Light Duty Trucks,'' October 2018, available at https://advocacy.consumerreports.org/wp-content/uploads/2018/10/CU-MMTC-Safety-Study-10-24-2018.pdf.
    \2019\ States and Cities, Detailed Comments, Docket No. NHTSA-
2018-0067-11735 at 81 and 95; American Honda, Detailed Comments, 
Docket No. NHTSA-2018-0067-11818, at 15; ICCT, Detailed Comments, 
Docket No. NHTSA-2018-0067-11741, at II-10-11. National Resources 
Defense Council, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-
4410, at 11-14.
---------------------------------------------------------------------------

    These comments and the MMTC study ignored the results of the fleet 
simulation study and seem premised on the notion that a vehicles' 
performance on NHTSA FMVSS, NHTSA voluntary NCAP, and IIHS voluntary 
safety tests is the only measure for assessing societal safety impacts 
for mass reduction. The regulatory and consumer information tests are 
representative of real-world, single-vehicle crash configurations. 
However, the tests are performed at constant speeds, and the dummy 
occupant is generally a mid-size male. In the real world, crashes occur 
at various impact velocities and configurations; with various impact 
partners (e.g., rigid obstacles, lighter or heavier vehicles); and 
involve occupants of various sizes and ages. The fleet simulation 
study, summarized above, assessed additional types of frontal crashes, 
including crashes with fixed objects and other vehicles at a wide range 
of vehicle speeds, and with mid-size male and mid-size female dummies. 
The fleet simulation study was more comprehensive and focused on the 
need to assess overall societal safety impacts. The fleet simulation 
study found that vehicle mass does impact safety with future 
lightweight vehicle designs that perform well on regulatory and 
consumer information tests.
    The agencies received one comment regarding the fleet simulation 
analysis. CARB commented that the analysis tested too few vehicles and 
crash types, should have optimized restraints in the lightweighted 
models to simulate future safety improvements instead of using modern 
restraints, and lacked credibility because the results of the fleet 
simulation analysis did not reproduce the same results of other 
studies.\2020\ CARB's comments demonstrate a general misunderstanding 
of the fleet simulation analysis; the analysis was not intended to 
serve as a prediction of how the future vehicle fleet will perform, but 
rather was an exploration of whether expected lightweighting techniques 
would alter the dynamic between mass reduction and safety. The analysis 
was not an attempt to model every potential vehicle construction or 
crash scenario. Attempting to simulate every future crash would be 
impractical and ineffective. The combination of vehicles and crash 
simulations were purposely selected to provide the strongest insight 
into the effective of lightweighting techniques. For passenger cars and 
light trucks, frontal crashes account for 58 percent of fatal crashes; 
\2021\ it is appropriate to focus research on understanding the effects 
of mass reduction where the largest issue exists. For the study, the 
use of generic restraint systems as the foundations for the models was 
intentional so that the models would be more representative of a 
vehicle class rather than a specific vehicle. The models of the 
restraint systems represented designs currently in production at time 
of the study in terms of pretensioners, load limiters and air bag 
inflators. It is worth noting that in general, driver air bags are 
similar in most vehicles. And finally, the analysis was not an attempt 
to reproduce the 2012 Kahane report or any other study. The fact that 
the fleet simulation analysis showed mass-reduction to be detrimental 
in more types of vehicles than in the FARS data only further highlights 
the need to consider how today's standards may impact mass-safety. 
While in the future there may be resources and opportunity to expand 
the fleet simulation approach to other crash scenarios and, if they 
become available, to include additional vehicle mass reduction 
concepts, the lack of potential future data does not justify ignoring 
the data that currently exist.
---------------------------------------------------------------------------

    \2020\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
11873, at 272-73.
    \2021\ Samaha, R.R., Prasad, P., Marzougui, D., Cui, C., Digges, 
K., Summers, S., Patel S., Zhao, L., & Barsan-Anelli, A. (2014, 
August). Methodology for evaluating fleet protection of new vehicle 
designs--Application to lightweight vehicle designs. Report No. DOT 
HS 812 051A, Washington, DC--National Highway Traffic Safety 
Administration.
---------------------------------------------------------------------------

    From a higher perspective, the comments, and in particular CARB's 
comment, identify the problem with abandoning real-world crash data: 
There is no alternate methodology or data that can account for the full 
diversity of crash scenarios that occur in the real world. Real-world 
crash data is the only data type that can achieve that. Therefore, the 
agencies have determined that, while simulations can prove helpful to 
understanding potential effects of key crash scenarios and as a check 
on the agencies' preferred analysis, real-world data still is still the 
best, most relevant data available for assessing safety.
(6) Summary of Mass Safety Impacts
    Table VI-208 through Table VI-213 show results of NHTSA's vehicle 
mass-size-safety analysis over the cumulative lifetime of MY 1977-2029 
vehicles, for both the CAFE and CO2 programs, based on the 
MY 2017 baseline fleet, accounting for the projected safety baselines. 
Results are driven extensively by the degree to which mass is reduced 
in relatively light passenger cars and in relatively heavy vehicles 
because their coefficients in the logistic regression analysis have the 
most significant values. The agencies assume any impact on fatalities 
will occur over the lifetime of the vehicle, and the chance of a 
fatality occurring in any particular year is directly related to the 
weighted vehicle miles traveled in that year.
BILLING CODE 4910-59-P

[[Page 24759]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.424


[[Page 24760]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.425


[[Page 24761]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.426


[[Page 24762]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.427


[[Page 24763]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.428


[[Page 24764]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.429

    As shown in the tables above, all of the alternatives are estimated 
to lead to a decrease in the number of mass-related fatalities over the 
cumulative lifetime of MY 1977-2029 vehicles. The effects of mass 
changes on fatalities

[[Page 24765]]

range from a combined decrease (relative to the augural standards, the 
baseline) of 143 fatalities for Alternative #7 to a combined decrease 
of 288 fatalities for Alternatives #1 and #2. The difference in results 
by alternative depends upon how much weight reduction is used in that 
alternative and the types and sizes of vehicles to which the weight 
reduction applies. The decreases in fatalities are driven by impacts 
within passenger cars (decreases of between 167 and 380 fatalities) and 
are offset by impacts within light trucks (increases of between 9 and 
92 fatalities).
    Changes in vehicle mass are estimated to decrease social safety 
costs over the lifetime of the nine model years by between $2.5 billion 
(for Alternative #7) and $5.1 billion (for Alternatives #1 and #2) 
relative to the augural standards at a three-percent discount rate and 
by between $1.5 billion and $3.1 billion at a seven-percent discount 
rate. The estimated decreases in social safety costs are driven by 
estimated decreases in costs associated with passenger cars, ranging 
from $3.0 billion (for Alternative #7) to $6.7 billion (for 
Alternatives #1 and #2) relative to the augural standards at a three-
percent discount rate and by between $1.8 billion and $4.0 billion at a 
seven-percent discount rate. The estimated decreases in costs 
associated with passenger cars are offset partially by estimated 
increases in costs associated with light trucks, ranging from $0.1 
billion (for Alternative #5) to $1.6 billion (for Alternatives #1 and 
#2) relative to the Augural standards at a three-percent discount rate 
and by between $0.1 billion and $0.9 billion at a seven-percent 
discount rate.
    In this analysis, the profile of mass reduction across vehicle 
models leads to a small, but beneficial effect on fatalities as fuel 
economy standards are tightened. Table VI-212 through Table VI-219 
present average annual estimated safety effects of vehicle mass 
changes, for CYs 2036-2045. The CY-level values offer a complementary 
view of the impacts of fuel economy standards on mass-related 
fatalities relative to model-year-level results. Effects by CY over the 
interval selected (2036-2045) enable a summary view of (a flow of) 
annual fatality impacts during a period where vehicles subjected to the 
standards have not only fully entered the fleet, but also interact with 
both older and newer vehicles. Conversely, the MY-level values offer a 
summary view of (a stock of) the impacts of fuel economy standards for 
the lifetime of a given MY:

[[Page 24766]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.430


[[Page 24767]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.431


[[Page 24768]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.432


[[Page 24769]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.433


[[Page 24770]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.434


[[Page 24771]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.435

    For all light-duty vehicles, mass changes are estimated to lead to 
an average annual decrease in fatalities in all alternatives evaluated 
for CYs 2035-2045. The effects of mass changes on fatalities range from 
a combined

[[Page 24772]]

decrease (relative to the augural standards) of 20 fatality per year 
for Alternative #7 to a combined decrease of 37 fatalities per year for 
Alternative #4. The difference in the results by alternative depends 
upon how much weight reduction is used in that alternative and the 
types and sizes of vehicles to which the weight reduction applies. The 
decreases in fatalities are generally driven by impacts within 
passenger cars (decreases of between 22 and 50 fatalities per year 
relative to the augural standards) and are offset by impacts within 
light trucks (increases of between 2 and 12 fatalities per year).
    Changes in vehicle mass are estimated to decrease average annual 
social safety costs in CY 2035-2045 by between $0.3 billion (for 
Alternative #7) and $0.6 billion (for Alternative #4) at a three-
percent discount rate relative to the augural standards (decrease of 
between $0.1 and $0.2 billion at a seven-percent discount rate). 
Average annual social safety costs associated with passenger cars in CY 
2035-2045 are estimated to decrease by between $0.3 billion and $0.7 
billion at a three-percent discount rate (decrease of between $0.1 
billion and $0.3 billion at a seven-percent discount rate), but this 
effect is partially offset by a corresponding increase in costs 
associated with light trucks (increase of $0.2 billion or less across 
alternatives at three-percent and seven-percent discount rates).
    To help illuminate effects at the model year level, Table VI-220 
presents the lifetime fatality impacts associated with vehicle mass 
changes for passenger cars, light trucks, and all light-duty vehicles 
by model year under the preferred alternative, relative to the augural 
standards for the CAFE Program. Table VI-221 presents an analogous 
table for the CO2 Program.

[[Page 24773]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.436

    Under the preferred alternative, passenger car fatalities 
associated with mass changes are estimated to decrease relative to the 
augural standards steadily from MYs 2018-19 (decrease of 5 fatalities) 
through MY 2028 (decrease of

[[Page 24774]]

53 fatalities). Conversely, light truck fatalities associated with mass 
changes under the preferred alternative are estimated to increase 
relative to the augural standards from MY 2019 (increase of 2 
fatalities) through MY 2029 (increase of 9 fatalities).
    Table VI-222 and Table VI-223 present estimates of monetized 
lifetime social safety costs associated with mass changes by model year 
at three-percent and seven-percent discount rates, respectively for the 
CAFE Program. Table VI-224 and Table VI-225 show comparable tables from 
the perspective of the CO2 Program.

[[Page 24775]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.437


[[Page 24776]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.438

    Lifetime social safety costs associated with mass change in 
passenger cars are estimated to decrease by between $0.1 billion (for 
MYs 2020-22) and $0.3 billion (for MYs 2026-29) at a three-percent 
discount rate. At a seven-

[[Page 24777]]

percent discount rate, lifetime social safety costs associated with 
mass change in passenger cars are estimated to decrease by between $0.1 
billion and $0.2 billion from MY 2021 through MY 2029. Lifetime social 
safety costs associated with mass change in light trucks are estimated 
to increase by $0.1 billion or less for all MYs at three-percent and 
seven-percent discount rates.
BILLING CODE 4910-59-P

[[Page 24778]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.439


[[Page 24779]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.440


[[Page 24780]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.441


[[Page 24781]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.442


[[Page 24782]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.443


[[Page 24783]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.444

    As shown in the tables above, all of the alternatives are estimated 
to lead to a decrease in the number of mass-related fatalities over the 
cumulative lifetime of MY 1977-2029 vehicles. The effects of mass 
changes on fatalities

[[Page 24784]]

range from a combined decrease (relative to the augural standards, the 
baseline) of 126 fatalities for Alternative #7 to a combined decrease 
of 253 fatalities for Alternatives #1 and #2. The difference in results 
by alternative depends upon how much weight reduction is used in that 
alternative and the types and sizes of vehicles to which the weight 
reduction applies. The decreases in fatalities are driven by impacts 
within passenger cars (decreases of between 146 and 33 fatalities) and 
are offset by impacts within light trucks (increases of between 8 and 
81 fatalities).
    Changes in vehicle mass are estimated to decrease social safety 
costs over the lifetime of the nine model years by between $2.2 billion 
(for Alternative #7) and $4.5 billion (for Alternatives #1 and #2) 
relative to the augural standards at a three-percent discount rate and 
by between $1.3 billion and $2.7 billion at a seven-percent discount 
rate. The estimated decreases in social safety costs are driven by 
estimated decreases in costs associated with passenger cars, ranging 
from $2.6 billion (for Alternative #7) to $5.9 billion (for 
Alternatives #1 and #2) relative to the Augural standards at a three-
percent discount rate and by between $1.6 billion and $3.5 billion at a 
seven-percent discount rate. The estimated decreases in costs 
associated with passenger cars are offset partially by estimated 
increases in costs associated with light trucks, ranging from $0.1 
billion (for Alternative #5) to $1.4 billion (for Alternatives #1 and 
#2) relative to the Augural standards at a three-percent discount rate 
and by between $0.1 billion and $0.8 billion at a seven-percent 
discount rate.
    In this analysis, the profile of mass reduction across vehicle 
models leads to a small, but beneficial effect on fatalities as fuel 
economy standards are tightened. Table VI-232 through Table VI-237 
present average annual estimated safety effects of vehicle mass 
changes, for CYs 2035-2045:
BILLING CODE 4910-59-P

[[Page 24785]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.445


[[Page 24786]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.446


[[Page 24787]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.447


[[Page 24788]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.448


[[Page 24789]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.449


[[Page 24790]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.450

BILLING CODE 4910-59-C
    For all light-duty vehicles, mass changes are estimated to lead to 
an average annual decrease in fatalities in all alternatives evaluated 
for CYs 2035-

[[Page 24791]]

2045. The effects of mass changes on fatalities range from a combined 
decrease (relative to the augural standards) of 17 fatality per year 
for Alternative #7 to a combined decrease of 34 fatalities per year for 
Alternative #4. The difference in the results by alternative depends 
upon how much weight reduction is used in that alternative and the 
types and sizes of vehicles to which the weight reduction applies. The 
decreases in fatalities are generally driven by impacts within 
passenger cars (decreases of between 19 and 44 fatalities per year 
relative to the augural standards) and are offset by impacts within 
light trucks (increases of between 2 and 11 fatalities per year).
    Changes in vehicle mass are estimated to decrease average annual 
social safety costs in CY 2035-2045 by between $0.2 billion (for 
Alternative #7) and $0.5 billion (for Alternative #4) at a three-
percent discount rate relative to the augural standards (decrease of 
between $0.1 and $0.2 billion at a seven-percent discount rate). 
Average annual social safety costs associated with passenger cars in CY 
2035-2045 are estimated to decrease by between $0.3 billion and $0.6 
billion at a three-percent discount rate (decrease of between $0.1 
billion and $0.3 billion at a seven-percent discount rate), but this 
effect is partially offset by a corresponding increase in costs 
associated with light trucks (increase of $0.1 billion or less across 
alternatives at three-percent and seven-percent discount rates).
    To help illuminate effects at the model year level, Table VI-238 
presents the lifetime fatality impacts associated with vehicle mass 
changes for passenger cars, light trucks, and all light-duty vehicles 
by model year under the preferred alternative, relative to the Augural 
standards for the CAFE Program. Table VI-239 presents an analogous 
table for the CO2 Program.

[[Page 24792]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.451

    Under the preferred alternative, passenger car fatalities 
associated with mass changes are estimated to decrease relative to the 
augural standards steadily from MYs 2018-19 (decrease of 4 fatalities) 
through MYs 2028-29

[[Page 24793]]

(decrease of 46 fatalities). Conversely, light truck fatalities 
associated with mass changes under the preferred alternative are 
estimated to increase relative to the augural standards from MY 2019 
(increase of 1 fatality) through MY 2029 (increase of 8 fatalities).
    Table VI-240 and Table VI-241 present estimates of monetized 
lifetime social safety costs associated with mass changes by model year 
at three-percent and seven-percent discount rates, respectively for the 
CAFE Program. Table VI-242 and Table VI-243 show comparable tables from 
the perspective of the CO2 Program.

[[Page 24794]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.452


[[Page 24795]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.453

BILLING CODE 4910-59-C
    Lifetime social safety costs associated with mass change in 
passenger cars are estimated to decrease by between $0.1 billion (for 
MYs 2020-23) and $0.3

[[Page 24796]]

billion (for MYs 2026-29) at a three-percent discount rate. At a seven-
percent discount rate, lifetime social safety costs associated with 
mass change in passenger cars are estimated to decrease by between $0.1 
billion and $0.2 billion from MY 2022 through MY 2029. Lifetime social 
safety costs associated with mass change in light trucks are estimated 
to increase by less than $0.1 billion for all MYs at three-percent and 
seven-percent discount rates.
b) Impact of Vehicle Prices on Fatalities
    The sales and scrappage responses discussed above have important 
safety consequences and influence safety outcomes through the same 
basic mechanism, fleet turnover. In the case of the scrappage response, 
delaying fleet turnover keeps drivers in older vehicles which are less 
safe than newer vehicles.\2022\ Similarly, the sales response slows the 
rate at which newer vehicles, and their associated safety improvements, 
enter the on-road population. The sales response also influences the 
mix of vehicles on the road--with more stringent CAFE standards leading 
to a higher share of light trucks sold in the new vehicle market, 
assuming all else is equal. Light trucks have higher rates of fatal 
crashes when interacting with passenger cars and, as earlier sections 
discussed, different directional responses to mass reduction technology 
based on the existing mass and body style of the vehicle.\2023\
---------------------------------------------------------------------------

    \2022\ See Passenger Vehicle Occupant Injury Severity by Vehicle 
Age and Model Year in Fatal Crashes, Traffic Safety Facts Research 
Note, DOT-HS-812-528, National Highway Traffic Safety 
Administration, April, 2018, and The Relationship Between Passenger 
Vehicle Occupant Injury Outcomes and Vehicle Age or Model Year in 
Police-Reported Crashes, Traffic Safety Facts Research Note, DOT-HS- 
(812-937), National Highway Traffic Safety Administration, March, 
2020.
    \2023\ See Section 6. Analytical Approach as Applied to 
Regulatory Alternatives] for a full explanation of the sales and 
scrappage effects and how they are modeled.
---------------------------------------------------------------------------

    With an integrated fleet model now part of the analytical framework 
for CAFE analysis, any effects on fleet turnover (either from delayed 
vehicle retirement or deferred sales of new vehicles) will affect the 
distribution of both ages and model years present in the on-road fleet. 
Because each of these vintages carries with it inherent rates of fatal 
crashes, and newer vintages are generally safer than older ones, 
changing that distribution will change the total number of on-road 
fatalities under each regulatory alternative. Similarly, the dynamic 
fleet share model captures the changes in the fleet's composition of 
cars and trucks. As cars and trucks have different fatality rates, 
differences in fleet composition across the alternatives will affect 
fatalities.
    At the highest level, the agencies calculate the impact of the 
sales and scrappage effects by multiplying the VMT of a vehicle by the 
fatality risk of that vehicle. For this analysis, calculating VMT is 
rather simple: the agencies use the distribution of miles calculated in 
Section VI.D.1.b)(5)(b). The trickier aspect of the analysis is 
creating fatality rate coefficients. The fatality risk measures the 
likelihood that a vehicle will be involved in fatal accident per mile 
driven. As explained below, the agencies' methodology changed from the 
proposal to this final rule in response to comments, but the basic 
analytical framework remains the same. The agencies calculate the 
fatality risk of a vehicle based on the vehicle's model year, age, and 
style, while controlling for factors which are independent of the 
intrinsic nature of the vehicle, such as behavioral characteristics.
(1) How the Agencies Modeled Impacts of Vehicle Scrappage and Sales on 
Fatalities in the NPRM
    In the proposal, the sales-scrappage safety model comprised two 
components.\2024\ First, the agencies estimated an empirical 
relationship among vehicle age, model year or vintage, and fatalities 
using the FARS database of fatal crashes, vehicle registration data 
from Polk to represent the on-road vehicle population, and the mileage 
accumulation schedules discussed in Section VI.D.1.b)(5) Vehicles Miles 
Traveled to estimate total vehicle use.\2025\ These data were used to 
construct per-mile fatality rates that varied by vehicle vintage, and 
also accounted for the influence of vehicle age. To accomplish this, 
the agencies used FARS data at a lower level of resolution; rather than 
looking at each crash and the specific factors that contributed to its 
occurrence, the agencies looked at the total number of fatal crashes 
involving light-duty vehicles over time with a focus on the influence 
of vehicle age and vehicle vintage. The model used in the proposal 
incorporated a weighted quartic polynomial regression (with each 
observation weighted by the number of registered vehicles it 
represented) on vehicle age, and included fixed effects for each model 
year present in the dataset. The model reproduced the observed 
fatalities of a given model year, at each age, reasonably well with 
more recent model years estimated with smaller errors. These estimates 
were used to account for the inherent safety risks of the legacy fleet 
and the influence of age on a vehicle's fatality rate.
---------------------------------------------------------------------------

    \2024\ The derivation of the NPRM analysis is discussed in 
detail in Section 7 of the FRIA.
    \2025\ The analysis supporting the CAFE rule for MYs 2017 and 
beyond did not account for differences in exposure or inherent 
safety risk as vehicles aged throughout their useful lives. However, 
the relationship between vehicle age and fatality risk is an 
important one. In a 2013 Research Note, NHTSA's National Center for 
Statistics and Analysis (NCSA) concluded a driver of a vehicle that 
is 4-7 years old is 10% more likely to be killed in a crash than the 
driver of a vehicle 0-3 years old, accounting for the other factors 
related to the crash. This trend continued for older vehicles more 
generally, with a driver of a vehicle 18 years or older being 71% 
more likely to be killed in a crash than a driver in a new vehicle. 
``How Vehicle Age and Model Year Relate to Driver Injury Severity in 
Fatal Crashes,'' DOT HS 811 825, NHTSA NCSA, August 2013. While 
there are more registered vehicles that are 0-3 years old than there 
are 20 years or older (nearly three times as many) because most of 
the vehicles in earlier vintages are retired sooner, the average age 
of vehicles in the United States is 11.6 years old and has risen 
significantly in the past decade.
---------------------------------------------------------------------------

    In the proposal, the agencies noted that factors other than the 
advent of new safety technologies have affected the historical trend in 
fatality and injury rates and are likely to continue to do so in the 
future. These include changes in driver behavior, including seat belt 
use, driving under the influence of alcohol or drugs, and driver 
distraction, particularly from the use of hand-held electronic devices 
such as smartphones, all of which affect either the frequency with 
which drivers are involved in crashes or the severity of accidents. 
They also include changes in the demographic composition of driving, 
since drivers of different ages, gender, income levels, and educational 
attainment have differing accident-involvement rates, as well as in the 
geographic distribution of motor vehicle travel, since road and driving 
conditions (visibility, etc.) tend to be poorer in rural areas than in 
urban locations, thus leading to more frequent and more severe crashes. 
Other factors affecting safety trends include infrastructure 
investments and road maintenance practices that improve road design and 
travel conditions, thus reducing the frequency and severity of crashes, 
improvements in accident response and emergency medical care, and 
cyclical variation in economic activity, which affects the demographic 
composition of drivers on the road.
    Seat belts have historically been the single most effective safety 
technology, preventing roughly half of all fatalities in the event of a 
potentially fatal crash, and accounting for over half the lives 
cumulatively saved by all FMVSS-related safety technologies since 
1960.\2026\ While belts have been in passenger vehicles since the 
1960s, few

[[Page 24797]]

drivers or passengers initially used them. Over the past 3 decades, 
seat belt usage rates have steadily climbed from under 60 percent in 
the early 1990s to roughly 90 percent in 2018 and has been the single 
most significant factor in reducing fatality rates over time. 
Additional changes in seat belt use are possible but challenging to 
achieve, since the last drivers to buckle up are typically the most 
likely to be risk takers and are often the most resistant to changing 
their habits. Moreover, with usage rates already at 90 percent, there 
is less potential for continued improvement.
---------------------------------------------------------------------------

    \2026\ Kahane, C.J., Lives Saved by Vehicle Safety Technologies 
and Associated Federal Motor Vehicle Safety Standards, 1960 to 
2012--Passenger Cars and LTVs, National Highway Traffic Safety 
Administration, Paper Number 15-0291. https://www-esv.nhtsa.dot.gov/Proceedings/24/files/24ESV-000291.PDF.
---------------------------------------------------------------------------

    Overall, the agencies believe improvement in seat belt use is 
unlikely to have the impact going forward that it has in the past. 
Technological fixes are possible for seat belt use and impaired 
driving, but would likely require the promulgation of new regulation, 
and therefore cannot be assumed. Similarly, individual States could 
take steps to address impaired driving, speeding, driver distraction, 
seat belt use and roadway infrastructure improvements, but the pace and 
impact of such improvements is speculative. The agencies also note that 
improvements in roadway infrastructure and human factors such as belt 
and alcohol use potentially affect both old and new vehicles alike. If 
improvements in these non-vehicle factors are equally spread across 
vehicles of all MY age groups, the differences in their fatality rates 
would not change. In other words, these types of improvements might 
shift the entire MY fatality rate curve down rather than change its 
slope.
    Nonetheless, the agencies stated that it was reasonable to expect 
some continuation in the generalized trend from non-vehicle technology 
factors such as these. In the analysis supporting the NPRM, our 
statistical model controlled for non-vehicle safety factors by 
accounting for the well-documented fact that older vehicles tend to be 
owned and driven by drivers whose demographic characteristics, 
behavior, and geographic location tends are associated with more 
frequent or severe crashes.
    Second, the agencies created estimates of future fatality rates. 
The agencies noted that predicting future safety trends has an inherent 
degree of uncertainty, which was amplified due to the dearth of 
academic and empirical research available at the time of the proposal. 
Although the agencies expected further safety improvements because of 
advanced driver assistance systems, such as automatic braking and 
eventually fully automated vehicles, the pace of development and extent 
of consumer acceptance of these improvements was uncertain. Thus, 
instead of attempting to model the impact of future safety features 
directly, the agencies relied on two different trend models to predict 
future safety trends. The first model relied on the results from a 
previous NCSA study that measured the effect of known safety 
regulations on fatality rates by performing statistical evaluations of 
the effectiveness of motor vehicle safety technologies based on real 
world performance in the on-road vehicle fleet to determine the 
effectiveness of each safety technology.\2027\ The agencies used this 
information to forecast future fatality rates. The second model 
employed was simpler. The agencies used actual, aggregate fatality 
rates measured from 2000 through 2016 and modeled the fatality rate 
trend based on these historical data.
---------------------------------------------------------------------------

    \2027\ Blincoe, L. and Shankar, U., ``The Impact of Safety 
Standards and Behavioral Trends on Motor Vehicle Fatality Rates,'' 
National Highway Traffic Safety Administration, DOT HS 810 777, 
Washington, DC, January, 2007.
---------------------------------------------------------------------------

    The agencies noted that both models had significant limitations and 
predicted significantly different safety trends. The NCSA study focused 
on projections to reflect known technology adaptation requirements, but 
it was conducted prior to the 2008 recession, which disrupted the 
economy and changed travel patterns throughout the country, and 
predated the emergence of newer technologies in the 2010s. The NCSA 
anticipated continued improvement well beyond 2020. By contrast, the 
historical fatality rate model reflected shifts in safety not captured 
by the NCSA model, but gave arguably implausible results after 2020 
because of an observed upward shift in fatalities between 2014 and 
2015. It essentially represented a scenario in which economic, market, 
or behavioral factors minimize or offset much of the potential impact 
of future safety technology. To reconcile the two projections of safety 
improvements beyond 2015, the agencies averaged the NCSA and historical 
fatality rate models, accepting each as an illustration of different 
and conflicting possible future scenarios.
    The agencies received a number of comments on the provisional model 
used in the NPRM, which focused mainly on its omission of variables 
that change over time and can affect the safety of all vehicles in use, 
regardless of their original model year or current age. As indicated 
previously, these include changes in seat belt use, driving under the 
influence of alcohol or drugs, use of hand-held electronic devices, 
driver demographics, the geographic distribution of vehicle use, road 
design and maintenance, emergency response and medical care, and 
overall economic activity.
    For example, CARB asserted that the NPRM modeling overestimated 
fatality rates for older vehicles because it did not ``control for 
factors that can have a significant influence on fatality risk, such as 
crash circumstances and driver characteristics.'' Elsewhere, CARB 
highlighted the omission of calendar year effects from the NPRM 
analysis, adding ``the agencies only model fatality rate as a function 
of model year, but fatality rate should be a function of both model 
year and calendar year [. . .] [which] would account for systematic 
safety improvements to the entire on-road fleet.'' \2028\ CARB also 
argued that analysis should account for safety differences between body 
styles, noting that passenger cars and other LTVs ``have historically 
had different safety regulations.'' \2029\ Passenger cars and LTVs are 
not always regulated at exactly the same pace and in some 
circumstances, LTV regulations have differed from passenger car 
regulations. However, with a few exceptions, both types of passenger 
vehicles are equipped with safety technologies that address the same 
basic safety hazards. Historically, these involve regulations that 
preserve passenger compartment integrity and protect passengers in the 
event of a crash. These include technologies such as air bags, seat 
belts, stronger roof structures, side door beams, and fuel tank 
integrity. Further, going forward, the agencies expect that both 
vehicle types will eventually all be equipped with the same advanced 
crash avoidance safety technologies that are currently being developed. 
Whatever differences there are have influenced the fatality rates and 
since this rulemaking uses combined average fatality rates (for PCs and 
LTVs) for the model, the results should closely mirror the results from 
an analysis that calculates the two vehicle types separately and then 
adds them together.
---------------------------------------------------------------------------

    \2028\ CARB, Detailed Comments, NHTSA-2018-0067-11873 at 263.
    \2029\ CARB, Auken Fatality Report, NHTSA-2018-0067-11881, at 
25.
---------------------------------------------------------------------------

    Similarly, States and Cities noted the potential importance of 
factors that can affect trends in vehicle safety over time, pointing 
out that ``increased seat belt use over time, improvements in roadway 
design and life-saving emergency response and treatment, and crash 
compatibility with other vehicles improve the overall safety of 
vehicles currently on the road'' and therefore

[[Page 24798]]

concluded that ``the CAFE model's assumption that the fatality rate of 
a 1985 model year vehicle is 23.8 per billion vehicle miles traveled 
for any calendar year is incorrect. That error increases the risk of 
fatalities determined by the NPRM for scrappage by around 25 percent.'' 
\2030\ Consumers Union echoed this argument and suggested driver 
characteristics and behavior may ``more strongly influence fatality 
risk than a vehicle's model year.'' \2031\
---------------------------------------------------------------------------

    \2030\ States and Cities, Detailed Comments, NHTSA-2018-0067-
11735, at 101 (internal citation omitted).
    \2031\ Consumers Union, et al., NHTSA-2018-0067-11731, 
Attachment 11, at 14.
---------------------------------------------------------------------------

    IPI speculated that omitting the effect of variables that change 
over time in ways that could affect fleet-wide safety may have caused 
the agencies' analysis to over-emphasize the role of safety 
improvements to new vehicles. Specifically, IPI observed that ``the 
agencies could not adequately control for driver behavior trends. And a 
decrease in fatalities could look like it was caused by vehicle 
improvements over time rather than societal changes.'' \2032\
---------------------------------------------------------------------------

    \2032\ IPI, NHTSA-2018-0067-12213, at 71.
---------------------------------------------------------------------------

    The agencies also received a few comments on their modeling 
choices. For example, CARB commented that the agencies equation for the 
legacy fleet was ``either incorrect or [had] limited domain-of-validity 
because it can potentially predict negative fatality rates'' and 
because it was missing an intercept term.\2033\ CARB suggested a 
logarithmic function would fix the problem. The agencies note that the 
polynomial specification of the safety model the agencies developed for 
the legacy fleet was extremely unlikely to predict negative fatality 
rates in light of the estimated values of its coefficients, and that 
its fixed-effects specification in effect included separate intercept 
terms for each model year, with that for the earliest model year 
serving as the ``reference case'' and thus performing the normal role 
of the constant term.
---------------------------------------------------------------------------

    \2033\ CARB, Auken Fatality Report, NHTSA-2018-0067-11881, at 
25.
---------------------------------------------------------------------------

    In electing to offset rebound-related safety consequences for the 
NPRM, the agencies distinguished the rebound effect from mass and fleet 
turnover impacts by describing the former as a voluntary consumer 
choice and the latter as imposed by the standards on consumers.\2034\ 
The agencies acknowledged in the NPRM that a reasonable argument might 
be made that consumers' decisions to purchase newer and safer cars or 
light trucks and to keep older models in service are also voluntary 
consumer choices, in which case changes in their decisions in response 
to newly-adopted CAFE and CO2 standards might be accompanied 
by offsetting gains or losses in benefits. The agencies dismissed this 
argument in the NPRM by noting that new vehicle prices act as a barrier 
to entry for some consumers, hence--at least ``marginal'' shoppers--
purchasing a more expensive vehicle is not a choice; and, without the 
ability to determine how many potential purchasers are `priced out' of 
the new vehicle market, it would be inappropriate to offset sales and 
scrappage safety impacts.\2035\ The agencies sought comment on this 
assumption.
---------------------------------------------------------------------------

    \2034\ See 83 FR at 43107.
    \2035\ The agencies further augmented the discussion by 
explaining that less stringent standards encouraged new vehicle 
purchases through lower vehicle prices while simultaneously 
discouraging additional driving due to higher operating costs. See 
id.
---------------------------------------------------------------------------

    The agencies did not receive any suggestions for distinguishing 
between consumers who voluntarily delayed purchases and those who were 
forced to delay a purchase due to high vehicle prices. Thus, the 
problem of deciphering the motives behind delayed purchases still 
lingers. However, the agencies did receive several comments advocating 
that the agencies offset fatalities attributable to sales and scrappage 
as they do for the rebound effect. For example, NCAT commented that 
``consumer purchases are voluntary and this effect should not be 
attributed to the standards.'' \2036\ The environmental group coalition 
commented that miles driven in older vehicles are ``a consumer choice, 
not something the standards compel.'' \2037\ In comparing the decision 
to retain and drive older vehicles to the decision to drive new 
vehicles more, i.e. the rebound effect, EDF concluded, ``to treat these 
identical choices in 180 degree different manners is of course 
manifestly arbitrary.'' \2038\
---------------------------------------------------------------------------

    \2036\ NCAT, Comments, NHTSA-2018-0067-11969, at 32-33.
    \2037\ Environmental Group Coalition, Appendix A, NHTSA-2018-
0067-12000, at 40-41.
    \2038\ EDF, Appendix B, NHTSA-2018-0067-12108, at 58.
---------------------------------------------------------------------------

    On a rudimentary level, the agencies agree with commenters that 
purchasing decisions are a consumer choice. While reducing the 
stringency of the standards should make new vehicles more affordable, 
nothing in today's rule requires consumers to purchase a new vehicle; 
likewise, the analysis does not assume every older vehicle will be 
replaced immediately. There is no strict requirement that the agencies 
must offset consumer choices. In fact, such a viewpoint would be 
untenable. Nothing in today's rule compels private parties to do 
anything. If the agencies assumed all freely chosen or voluntary 
actions, such as driving or manufacturing automobiles, were not 
attributable to the rule, then each regulatory scenario would have the 
same net benefit--zero. As such, the agencies explanation in the 
proposal of freely chosen and voluntary was likely imprecise and led 
commenters to an overly broad conclusion. Deciding which behavioral 
responses are unambiguously attributable to a regulation and should 
thus be quantified, and distinguishing them from responses that would 
be anticipated to occur in its absence is inherently part of the 
rulemaking process, and inevitably requires agencies considering new 
regulations to apply careful judgment in making those distinctions.
    To that end, the agencies felt it was appropriate to offset 
rebound-related safety costs because of the benefit rebound miles 
confer to society. As described in more detail in Section 1.b)(6), 
additional driving that occurs as a consequence of the fuel economy 
rebound effect is undertaken voluntarily, and the agencies can infer 
from the fact that it is freely chosen that the mobility benefits it 
provides necessarily exceed the additional operating costs and 
increased exposure to safety risks it entails. Since reducing the 
standards has the ancillary effect of reducing rebound miles, the 
agencies concluded that including safety costs associated with rebound 
driving would cause the agencies to underestimate the lost value of 
rebound driving; therefore, it was appropriate to offset rebound safety 
costs to account for the lost benefits.\2039\ Thus, the significance of 
the terms freely chosen and voluntary was to signal that consumers' 
actions were motivated in part by benefits that may not have been not 
explicitly identified or accounted for, rather than to act as a 
prohibitive characteristic.
---------------------------------------------------------------------------

    \2039\ Arguably rebound fatalities and non-fatal injuries should 
be included in today's analysis as a cost without an offset. While a 
perfectly rational driver would fully and accurately internalize the 
costs associated with driving on a per-mile basis and would only 
drive if the expected benefits at least offset the expected costs, 
it is difficult to ascertain how much of the risk a real person 
internalizes. If not for the reduced standards, fatalities would 
increase due to rebound driving.
---------------------------------------------------------------------------

    When considering commenters' suggestion to offset fleet turnover 
fatalities (as well as injury and ancillary costs), the agencies 
attempted to identify specific benefits whose loss would be logically 
attributable to the changes in standards this rule adopts, and were not 
accounted for elsewhere in

[[Page 24799]]

their analysis. The agencies considered whether accelerated turnover of 
the car and light truck fleet could cause mobility losses analogous to 
those resulting from the rebound effect, but determined that on 
balance, increasing the pace at which new vehicles replace older models 
that are retired from use provides additional mobility and other 
benefits.\2040\ In addition, the agencies considered whether consumers 
experience some previously unidentified loss in welfare when they 
purchase new vehicles, particularly when they do so to replace an older 
model. As explained in in Section 1.b)(6) and 1.b)(8), the agencies 
instead concluded that purchasers instead experience gains in welfare 
as a result, but that the resulting benefits are already accounted for 
elsewhere in their analysis.
---------------------------------------------------------------------------

    \2040\ This occurs because newer vehicles are not only more 
fuel-efficient on average than the older models they replace, but 
also provide more reliable, comfortable, and otherwise higher-
quality transportation service, so they tend to be driven more than 
those they replace.
---------------------------------------------------------------------------

    Finally, the agencies contemplated whether--as commenters 
contended-- owners of older vehicles derive some heretofore 
unaccounted-for benefit from continuing to use them, which might be 
reduced when the rule encourages more rapid retirement of older models. 
Applying the same logic used to explain additional driving in response 
to the rebound effect, an older vehicle will continue to be maintained 
in working condition and driven when the benefits provided to the owner 
is sufficient to offset the costs of maintenance and operation, 
including the economic costs associated with additional exposure to 
safety risks. Therefore, there is a benefit to driving an older 
vehicle. But the relevant question is not whether a benefit exists but 
how this rule might affect those benefits. With the very limited 
exception of classic cars, it is unlikely that the benefit of driving 
an older vehicle confers a greater benefit than driving a newer 
vehicle.\2041\ Normally, when a vehicle is scrapped, it is replaced 
with a newer vehicle. Hence mobility is not lost, but rather 
transferred between vehicles--and with it, the associated 
benefits.\2042\ In the limited instances where a retired vehicle is not 
replaced with a newer vehicle, that action is freely taken and the 
agencies can infer from that decision that the benefit derived from 
scrapping the vehicle outweighed any possible loss, including lost 
mobility. Offsetting the reduction in scrappage safety costs--realized 
because of the standards--without a complementary benefit would be 
directionally inconsistent.\2043\
---------------------------------------------------------------------------

    \2041\ If the benefit of driving an older vehicle was higher 
than the benefit of driving a newer vehicle, we would anticipate 
consumers to forgo replacing older vehicles with newer vehicles.
    \2042\ Since driving newer vehicles, including newer used 
vehicles, likely confers greater benefits than would-be scrapped 
vehicle, the agencies are likely underestimating the value of 
increased scrappage.
    \2043\ A similar argument could be made that consumers 
`internalize' additional fuel costs, and therefore pre-tax fuel 
savings should also be offset. However, this would also ignore that 
benefits are remaining constant while the costs to obtain those 
benefits is increasing.
---------------------------------------------------------------------------

    The agencies reaffirm that off-setting safety costs attributable to 
the sales and scrappage effects is inappropriate. Commenters' arguments 
relied exclusively on the premise that driving older vehicles is freely 
chosen and thus must have associated benefits, without considering the 
impact of accelerating their retirement on the rule's overall net 
safety and mobility benefits. Furthermore, the agencies remain 
concerned that potential buyers may be ``frozen out'' of the new 
vehicle market by prohibitively high prices; in which case enabling 
access to newer, safer vehicles provides measurable safety benefits 
that should be considered by the analysis.
    However, in an abundance of caution, the agencies performed a 
sensitivity analysis that applies the same safety offset to sales/
scrappage safety impacts that was applied to the rebound effect safety 
impacts. The results are provided in Table VI-244 below. As might be 
expected, this adjustment reduces net benefits in all scenarios, but 
does not substantially shift the relative scope among alternatives.
BILLING CODE 4910-59-P

[[Page 24800]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.454

BILLING CODE 4910-59-C
    Again, the agencies feel that this offset is inappropriate. The 
sensitivity case disregards many of the tangible gains in safety 
expected from increased sales and

[[Page 24801]]

scrappage. Furthermore, the agencies note that--even if they replaced 
the central analysis' assumptions with this sensitivity case--the 
anticipated changes in net benefits would not be enough to change their 
decision.
(2) Revised Sales-Scrappage Safety Model
    In response to the comments, the agencies have taken several steps 
to revise the sales-scrappage safety model. First, the agencies 
developed a revised statistical model to explain historical 
improvements in the lifetime safety performance of each successive new 
vintage of cars and light trucks, and used the results of this improved 
model to project the future trend in the overall fatality rates. While 
the revised historical trend model itself is more complex than the one 
utilized in the proposal, the overall procedure is simpler; the 
agencies have collapsed the two piecemeal components discussed above 
into one model and eliminated the need to `reconcile' differences 
between competing future projections. Next, the agencies applied 
detailed empirical estimates of the market uptake and improving 
effectiveness of crash avoidance technologies to estimate their effect 
on the fleet-wide fatality rate, including explicitly incorporating 
both the direct effect of those technologies on the crash involvement 
rates of new vehicles equipped with them, as well as the ``spillover'' 
effect of those technologies on improving the safety of occupants of 
vehicles that are not equipped with these technologies.
(a) Crash Avoidance
    In the NPRM, the agencies took a very generalized approach to 
estimating the pace of future safety trends. For reasons discussed 
above, the agencies noted that there was uncertainty regarding actual 
trends in fatality rates. This issue was addressed by numerous 
commenters who took opposing positions. Among them, IPI stated that 
``[t]he agencies have not provided an adequate explanation for why past 
safety trends are likely to continue until the mid-2020s.'' IPI further 
noted that ``crash avoidance technology may not be adopted as easily or 
readily as crash mitigation technologies have been.'' \2044\ In 
response, the agencies note that the trend the agencies adopted for the 
NPRM was not a direct continuation of past trends. Rather, it was a 
simple average of several possible models the agencies had examined, 
accepting each as an illustration of different and conflicting possible 
future scenarios.
---------------------------------------------------------------------------

    \2044\ IPI, Appendix, NHTSA-2018-0067-12213, at 98.
---------------------------------------------------------------------------

    By contrast, States and Cities asserted that fatality rates may be 
lower in the future than the agencies estimated, noting that the NPRM 
analysis did not ``account for safety benefits that new safety 
technologies in future vehicles will have on the agencies predicted 
outcome.'' \2045\ While the agencies agree that the NPRM analysis did 
not analyze individual safety benefits of new technologies, the trends 
included in the NPRM were intended, in part, as a proxy estimate of the 
impact of these technologies. As discussed in the NPRM, these 
technologies were cited as a justification for assuming a continued 
downward trend in the fatality rate through roughly 2035.
---------------------------------------------------------------------------

    \2045\ States and Cities, Detailed Comments, NHTSA-2018-0067-
11735, at 80.
---------------------------------------------------------------------------

    Nonetheless, the agencies believe that further analysis of these 
potential trends can now be ascertained for several explicit 
technologies. In response to comments suggesting that the agencies 
account more directly for new safety technologies, the agencies 
augmented the sales-scrappage safety analysis for the final rule with 
recent research into the effectiveness of specific advanced crash 
avoidance safety technologies (also known as ADAS or advanced driver 
assistance systems) that are expected to drive future safety 
improvement to estimate the impacts of crash avoidance technologies. 
The analysis analyzes six crash avoidance technologies that are 
currently being produced and commercially deployed in the new vehicle 
fleet. These include Frontal Collision Warning (FCW), Automatic 
Emergency Braking (AEB), Lane Departure Warning (LDW), Lane Keep Assist 
(LKA), Blind Spot Detection (BSD), and Lane Change Alert (LCA).\2046\ 
These are the principal technologies that are being developed and 
adopted in new vehicle fleets and will likely drive vehicle-based 
safety improvements for the coming decade. These technologies are being 
installed in more and more new vehicles; in fact, 12 manufacturers 
recently reported that they voluntarily installed AEB systems in more 
than 75 percent of their new vehicles sold in the year ending August 
31, 2019.\2047\ The agencies note that the terminology and the detailed 
characteristics of these systems may differ across manufacturers, but 
the basic system functions are common across all.
---------------------------------------------------------------------------

    \2046\ A full description of these technologies and several 
other technologies referenced below may be found in the 
corresponding FRIA safety impacts discussion.
    \2047\ NHTSA Announces Update to Historic AEB Commitment by 20 
Automakers, NHTSA press release December 17, 2019. https://www.nhtsa.gov/press-releases/nhtsa-announces-update-historic-aeb-commitment-20-automakers.
---------------------------------------------------------------------------

    These six technologies address three basic crash scenarios through 
warnings to the driver or alternately, through dynamic vehicle control:
    1. Forward collisions, typically involving a crash into the rear of 
a stopped vehicle;
    2. Lane departure crashes, typically involving inadvertent drifting 
across or into another traffic lane; and
    3. Blind spot crashes, typically involving intentional lane changes 
into unseen vehicles driving in or approaching the driver's blind spot.
    Unlike traditional safety features where the bulk of the safety 
improvements were attributable to improved protection when a crash 
occurs (crash worthiness), the impact of advanced crash avoidance 
technologies (ADAS or advanced driver assistance systems) will have on 
fatality and injury rates is a direct function of their effectiveness 
in preventing or reducing the severity of the crashes they are designed 
to mitigate. This effectiveness is typically measured using real world 
data comparing vehicles with these technologies to similar vehicles 
without them. While these technologies are actively being deployed in 
new vehicles, their penetration in the larger on-road vehicle fleet has 
been at a low, but growing level. This limits the precision of 
statistical regression analyses, at least until the technologies become 
more common in the on-road fleet.
    Our approach in the final rule is to derive effectiveness rates for 
these advanced crash-avoidance technologies from safety technology 
literature. The agencies then apply these effectiveness rates to 
specific crash target populations for which the crash avoidance 
technology is designed to mitigate and adjusted to reflect the current 
pace of adoption of the technology, including the public commitment by 
manufactures to install these technologies. The products of these 
factors, combined across all 6 advanced technologies, produce a 
fatality rate reduction percentage that is applied to the fatality rate 
trend model discussed below, which projects both vehicle and non-
vehicle safety trends. The combined model produces a projection of 
impacts of changes in vehicle safety technology as well as behavioral 
and infrastructural trends.

[[Page 24802]]

(i) Technology Effectiveness Rates
(a) Forward Crash Collision Technologies
    For forward collisions, manufacturers are currently equipping 
vehicles with FCW, which warns drivers of impending collisions, as well 
as AEB, which incorporates the sensor systems from FCW together with 
dynamic brake support (DBS) and crash imminent braking (CIB) to help 
avoid crashes or mitigate their severity. Manufacturers have committed 
voluntarily to install some form of AEB on all light vehicles by the 
2023 model year (September 2022).\2048\
---------------------------------------------------------------------------

    \2048\ See https://www.nhtsa.gov/press-releases/nhtsa-iihs-announcement-aeb.
---------------------------------------------------------------------------

    Table VI-245 summarizes studies which have measured effectiveness 
for various forms of FCW and AEB over the past 13 years. Most studies 
focused on crash reduction rather than injury reduction. This is a 
function of limited injury data in the on-road fleet, especially during 
the early years of deployment of these technologies. In addition, it 
reflects engineering limitations in the technologies themselves. 
Initial designs of AEB systems were basically incapable of detecting 
stationary objects at speeds higher than 30 mph, making them 
potentially ineffective in higher speed crashes that are more likely to 
result in fatalities or serious injury. For example, Wiacek et al. (2-
15) conducted a review of rear-end crashes involving a fatal occupant 
in the 2003-2012 NASS-CDS data-bases to determine the factors that 
contribute to fatal rear-end crashes.\2049\ They found that the speed 
of the striking vehicle was the primary factor in 71 percent of the 
cases they examined. The average Delta-V of the striking vehicle in 
these cases was 46 km/h (28.5 mph), implying pre-crash travel speeds in 
excess of this speed. While Table VI-245 includes studies going back to 
2005, the agencies focus our discussion on more recent studies 
conducted after 2012 in order to reflect more current safety systems 
and vehicle designs.2050 2051 2052 2053 2054 2055 2056 2057
---------------------------------------------------------------------------

    \2049\ Wiacek, C., Bean, J., Sharma, D., Real World Analysis of 
Fatal Rear-End Crashes, National Highway Traffic Safety 
Administration, 24th Enhanced Safety of Vehicles Conference, 150270, 
2015.
    \2050\ Sugimoto, Y., and Sauer, C., (2005). Effectiveness 
Estimation Method for Advanced Driver Assistance System and its 
Application to Collision Mitigation Brake systems, paper number 05-
148, 19th International Technical Conference on the Enhanced safety 
of Vehicles (ESV), Washington DC, June 6-9, 2005.
    \2051\ Page, Y., Foret-Bruno, J., & Cuny, S. (2005). Are 
expected and observed effectiveness of emergency brake assist in 
preventing road injury accidents consistent?, 19th ESV Conference, 
Washington DC.
    \2052\ Najm, W.G., Stearns, M.D., Howarth, H., Koopman, J. & 
Hitz, J., (2006). Evaluation of an Automotive Rear-End Collision 
Avoidance System (technical report DOT HS 810 569), Cambridge, MA: 
John A. Volpe National Transportation System Center, U.S. Department 
of Transportation.
    \2053\ Breuer, JJ., Faulhaber, A., Frank, P. and Gleissner, S. 
(2007). Real world Safety Benefits of Brake Assistance Systems, 
Proceedings of the 20th International Technical Conference of the 
Enhanced Safety of Vehicles (ESV) in Lyon, France June 18-21, 2007.
    \2054\ Keuhn, M., Hummel, T., and Bende J., Benefit estimation 
of advanced driver assistance systems for cars derived from real-
world accidents, Paper No. 09-0317, 21st International Technical 
Conference on the Enhanced Safety of Vehicles (ESV)--International 
Congress Centre, Stuttgart, Germany, June 15-18, 2009.
    \2055\ Grover, C., Knight, I., Okoro, F., Simmons I., Couper, 
G., Massie, P., and Smith, B. (2008). Automated Emergency Brake 
Systems: Technical requirements, Costs and Benefits, PPR227, TRL 
Limited, DG Enterprise, European Commission, April 2008.
    \2056\ Kusano, K.G., and Gabler, H.C. (2015). Comparison of 
Expected Crash Injury and Injury Reduction from Production Forward 
Collision and Lane Departure Warning Systems, Traffic Injury 
Prevention 2015; Suppl. 2: S109-14.
    \2057\ HLDI (2011). Volvo's City Safety prevents low-speed 
crashes and cuts insurance costs, Status Report, Vol. 46, No. 6, 
July 19,2011.
[GRAPHIC] [TIFF OMITTED] TR30AP20.455


[[Page 24803]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.456

    Doecke et al. (2012) created 
2058 2059 2060 2061 2062 2063 simulations of 103 real world 
crashes and applied AEB system models with differing specifications to 
determine the change in impact speed that various AEB interventions 
might produce. Their modeling found significant rear-end crash speed 
reductions with various AEB performance assumptions. In addition, they 
estimated a 29 percent reduction in rear-end crashes and that 25 
percent of crashes over 10 km/h were reduced to 10 km/h or less.
---------------------------------------------------------------------------

    \2058\ Docke, S.D., Anderson, R.W.G., Mackenzie, J.R.R., Ponte, 
G. (2012). The potential of autonomous emergency braking systems to 
mitigate passenger vehicle crashes. Australian Road Safety Research 
Policing and Education Conference, October 4-6, 2012, Wellington, 
New Zealand.
    \2059\ Chauvel, C., Page, Y., Files, B.N., and Lahausse, J. 
(2013). Automatic emergency braking for pedestrians effective target 
population and expected safety benefits, Paper No. 13-0008, 23rd 
International Technical Conference on the Enhanced Safety of 
Vehicles (ESV), Seoul, Republic of Korea, May 27-30, 2013.
    \2060\ Fildes B., Keall M., Bos A., Lie A., Page, Y., Pastor, 
C., Pennisi, L., Rizzi, M., Thomas, P., and Tingvall, C. 
Effectiveness of Low Speed Autonomous Emergency Braking in Real-
World Rear-End Crashes. Accident Analysis and Prevention, AAP-D-14-
00692R2.
    \2061\ Cicchino, J.B. (2017). Effectiveness of forward collision 
warning and autonomous emergency braking systems in reducing front-
to-rear crash rates. Accident Analysis and Prevention, V. 99, Part 
A, February 2017, Pages 142-52.
    \2062\ Kusano, K.D., and Gabler H.C. (2012). Safety Benefits of 
Forward Collision Warning, Brake Assist, and Autonomous Braking 
Systems in Rear-End Collisions, Intelligent Transportation Systems, 
IEEE Transactions, Volume 13 (4).
    \2063\ Leslie, A, Kiefer, R., Meitzner, M, and Flannagan, C. 
(2019). Analysis of the Field Effectiveness of General Motors 
Production Active Safety and Advanced headlighting Systems. 
University of Michigan Transportation Research Institute, UMTRI-
2019-6, September, 2019.
---------------------------------------------------------------------------

    Cicchino (2016) analyzed the effectiveness of a variety of forward 
collision mitigation systems including both FCW and AEB systems. 
Cicchino used a Poisson regression to compare rates of police-reported 
crashes per insured vehicle year between vehicles with these systems 
and the same models that did not elect to install them. The analysis 
was based on crashes occurring during 2010 to 2014 in 22 States and 
controlled for other factors that affected crash risk. Cicchino found 
that FCW reduced all rear-end striking crashes by 27 percent and rear-
end striking injury crashes by 20 percent, and that AEB functional at 
high-speeds reduced these crashes by 50 and 56 percent, respectively. 
She also found that low speed AEB without driver warning reduced all 
crashes by 43 percent and injury crashes by 45 percent. She also found 
that even low-speed AEB could impact crashes at higher speed limits. 
Reductions were found of 53 percent, 59 percent, and 58 percent for all 
rear-end striking crash rates, rear-end striking injury crash rates, 
and rear-end third party injury crash rates, respectively, at speed 
limits of 40-45 mph. For speed limits of 35 mph or less, reductions of 
40 percent, 40 percent, and 43 percent were found. For speed limits of 
50 mph or greater, reductions of 31 percent, 30 percent, and 28 
percent, were found. Further, Cicchino (2016) found significant 
reductions (30 percent) in rear-end injury crashes even in crashes on 
roadways where speed limits exceeded 50 mph.
    Kusano and Gabler (2012) examined the effectiveness of various 
levels of forward collision technologies including FCW and AEB based on 
simulations of 1,396 real world rear end crashes from 1993-2008 NASS 
CDS data-bases. The authors developed a probability-based framework to 
account for variable driver responses to the warning systems. Kusano 
and Gabler found FCW systems could reduce rear-end crashes by 3.2 
percent and driver injuries in rear-end crashes by 29 percent. They 
also found that full AEB systems with FCW, pre-crash brake assist, and 
autonomous pre-crash braking could reduce rear-end crashes by 7.7 
percent and reduce moderate to fatal driver injuries in rear-end 
crashes by 50 percent.
    Fildes et al. (2015) performed meta-analyses to evaluate the 
effectiveness of low-speed AEB technology in passenger vehicles based 
on real-world crash experience across six different predominantly 
European countries. Data from these countries was pooled into a 
standard analysis format and induced exposure methods were used to 
control for extraneous effects. The study found a 38 percent overall 
reduction in rear-end crashes for vehicles with AEB compared to similar 
vehicles without this technology. The study also found no statistical 
evidence for any difference in effectiveness between urban roads with 
speed limits less than or equal to 60 km/h, and rural roads with speed 
limits greater than 60 km/h. Fildes et al. (2015) found no statistical 
difference in the performance of AEBs on lower speed urban or higher 
speed rural roadways.
    Kusano and Gabler (2015) simulated rear-end crashes based on a 
sample of 1,042 crashes in the 2012 NASS-CDS. Modelling was based on 54 
model year 2010-2014 vehicles that were evaluated in NHTSA's New Car 
Assessment Program (NCAP). Kusano and Gabler found FCW systems could 
prevent 0-67 percent of rear-end crashes and 2-69 percent of serious to 
fatal driver injuries.
    Leslie et al. (2019) analyzed the relative crash performance of 
123,377 General Motors (GM) MY 2013 to 2017 vehicles linked to State 
police-reported crashes by Vehicle Identification numbers (VIN). GM 
provided VIN-linked safety content information for these vehicles to 
enable precise identification of safety technology content. The authors 
analyzed the effectiveness of a variety of crash avoidance technologies 
including both FCW and AEB separately. They estimated effectiveness 
comparing system-relevant crashes to baseline

[[Page 24804]]

(control group) crashes using a quasi-induced exposure method in which 
rear-end struck crashes are used as the control group. Leslie et al. 
found that FCW reduced rear-end striking crashes of all severities by 
21 percent, and that AEB (which includes FCW) reduced these crashes by 
46 percent.\2064\
---------------------------------------------------------------------------

    \2064\ The agencies note that UMTRI, the sponsoring organization 
for the Leslie et al. study, published a previous version of this 
same study utilizing the same methods in March of 2018 (Flannagan, 
C. and Leslie, A, Crash Avoidance Technology Evaluation Using real-
World crashes, University of Michigan Transportation research 
Institute, March 22, 2018). The agencies focused on the more recent 
2019 study because its sample size is significantly larger and it 
represents more recent model year vehicles. The revised (2019) study 
uses the same basic techniques but incorporated a larger data-base 
of system-relevant and control cases (123,377 cases in the 2019 
study vs. 35,401 in the 2018 study). Relative to the Flannagan and 
Leslie (2018) findings, the results of the 2019 study varied by 
technology. The revised study found effectiveness rates of 21% for 
FCW and 46% for AEB, compared to 16% and 45% in the 2018 study. The 
revised study found effectiveness rates of 10% for LDW and 20% for 
LKA, compared to 3% and 30% for these technologies in the 2018 
study. The revised study found effectiveness rates of 3% for BSD and 
26-37% for LCA systems, compared to 8% and 19-32% for these 
technologies in the 2018 study. Thus, some system effectiveness 
estimates increased while others decreased.
---------------------------------------------------------------------------

    For this analysis, the agencies based their projections on Leslie 
et al. because they are the most recent study, and thus reflect the 
most current versions of these systems in the largest number of 
vehicles, and also because they arguably have the most precise 
identification of the presence of the specific technologies in the 
vehicle fleet. Furthermore, Leslie et al. was the only study to report 
estimates for each of the six crash avoidance technologies analyzed for 
the final rule, hence providing a certain level of consistency amongst 
estimates. The agencies recognize that there is uncertainty in 
estimates of these technologies effectiveness, especially at this early 
stage of deployment. For this reason, the agencies examine a range of 
effectiveness rates to estimate boundary outcomes in a sensitivity 
analysis.
    Leslie et al. measured effectiveness against all categories of 
crashes, but did not specify effectiveness against crashes that result 
in fatalities or injuries. The agencies examined a range of 
effectiveness rates against fatal crashes using a central case based on 
boundary assumptions of no effectiveness and full effectiveness across 
all crash types. Our central case is thus a simple average of these two 
extremes. Sensitivity cases were based on the 95th percent confidence 
intervals calculated from this central case. Leslie et al. found 
effectiveness rates of 21 percent for FCW and 46 percent for AEB. Our 
central fatality effectiveness estimates will thus be 10.5 percent for 
FCW and 23 percent for AEB. The calculated 95th percentile confidence 
limits range is 8.11 to 12.58 percent effective for FCW and 20.85 to 
25.27 for AEB. The agencies note that our central estimate is 
conservative compared to averages of those studies that did 
specifically examine fatality impacts; that is, the analysis assumes 
reduced future fatalities less than most of, or the average of, those 
studies, and thus minimizes the estimate of lives saved under 
alternatives to the augural standards. Furthermore, the agencies note 
that the estimates against fatal crashes is higher in the recent 
studies in Table VI-245, which reflects the agencies' understanding 
that earlier iterations of AEB and FCW may have been less effective 
against crashes that result in fatalties than newer and improved 
versions.\2065\
---------------------------------------------------------------------------

    \2065\ As an example of improvements, the agencies note that the 
Mercedes system described in their 2015 owner's manual specified 
that for stationary objects the system would only work in crashes 
below 31 mph, but that in their manual for the 2019 model, the 
systems are specified to work in these crashes up to 50 mph.
---------------------------------------------------------------------------

(b) Lane Departure Crash Technologies
    For lane departure crashes, manufacturers are currently equipping 
vehicles with lane departure warning (LDW), which monitors lane 
markings on the road and alerts the driver when their vehicle is about 
to drift beyond a delineated edge line of their current travel lane, as 
well as lane keep assist (LKA), which provides gentle steering 
adjustments to help drivers avoid unintentional lane crossing. Table 
VI-246 summarizes studies which have measured effectiveness for LDW and 
LKA.
[GRAPHIC] [TIFF OMITTED] TR30AP20.457

    Cicchino (2018) examined crash involvement rates per insured 
vehicle 2066 2067 2068 2069 2070 year for

[[Page 24805]]

vehicles that offered LDW as an option and compared crash rates for 
those that had the option installed to those that did not. The study 
focused on single-vehicle, sideswipe, and head-on crashes as the 
relevant target population for LDW effectiveness rates. The study 
examined 5,433 relevant crashes of all severities found in 2009-2015 
police-reported data from 25 States. The study was limited to crashes 
on roadways with 40 mph or greater speed limits not covered in ice or 
snow since lower travel speeds would be more likely to fall outside of 
the LDW systems' minimum operational threshold. Cicchino found an 
overall reduction in relevant crashes of 11 percent for vehicles that 
were equipped with LDW. She also found a 21 percent reduction in injury 
crashes. The result for all crashes was statistically significant, 
while that for injury crashes approached significance (p<0.07). 
Cicchino did not separately analyze LKA systems.
---------------------------------------------------------------------------

    \2066\ Cicchino, J.B. (2018). Effects of lane departure warning 
on police-reported crash rates, Journal of Safety Research 66 
(2018), pp.61-70. National Safety Council and Elsevier Ltd., May, 
2018.
    \2067\ Sternlund, S., Strandroth, J., Rizzi, M., Lie, A., and 
Tingvall, C. (2017). ``The effectiveness of lane departure warning 
systems--A reduction in real-world passenger car injury crashes,'' 
Traffic Injury Prevention V. 18 Issue 2 (Jan 2017).
    \2068\ Leslie et al., supra note 2063.
    \2069\ Kusano & Gable, supra note 2056.
    \2070\ Kusano, K., Gorman, T.I., Sherony, R., and Gabler, H.C. 
Potential occupant injury reduction in the U.S. vehicle fleet for 
lane departure warning-equipped vehicles in single-vehicle crashes. 
Traffic Injury Prevention 2014 Suppl 1:S157-64.
---------------------------------------------------------------------------

    Sternlund et al. (2017) studied single vehicle and head-on injury 
crash involvements relevant to LDW and LKA in Volvos on Swedish 
roadways. They used rear-end crashes as a control and compared the 
ratio of these two crash groups in vehicles that had elected to install 
LDW or LCA to the ratio in vehicles that did not have this content. 
Studied crashes were limited to roadways with speeds of 70-120 kph and 
not covered with ice or snow. Sternlund et al. found that LDW/LKA 
systems reduced single vehicle and head-on injury crashes in their 
crash population by 53 percent, with a lower limit of 11 percent, which 
they determined corresponded to a reduction of 30 percent (lower limit 
of 6 percent) across all speed limits and road surface assumptions.
    Leslie et al. (2019) analyzed the relative crash performance of 
123,377 General Motors (GM) MY 2013 to 2017 vehicles linked to state 
police-reported crashes by Vehicle Identification numbers (VIN). GM 
provided VIN-linked safety content information for these vehicles to 
enable precise identification of safety technology content. The authors 
analyzed the effectiveness of a variety of crash avoidance technologies 
including both LDW and LKA separately. They estimated effectiveness 
comparing system-relevant crashes to baseline (control group) crashes 
using a quasi-induced exposure method in which rear-end struck crashes 
are used as the control group. Leslie et al. found that LDW reduced 
lane departure crashes of all severities by 10 percent, and that LKA 
(which includes LDW) reduced these crashes by 20 percent.
    Kusano et al. (2014) developed a comprehensive crash and injury 
simulation model to estimate the potential safety impacts of LDW. The 
model simulated results from 481 single-vehicle collisions documented 
in the NASS-CDS data-base for the year 2012. Each crash was simulated 
as it actually occurred and again as it would occur had the vehicles 
been equipped with LDW. Crashes were simulated multiple times to 
account for variation in driver reaction, roadway, and vehicle 
conditions. Kusano et al. found that LDW could reduce all roadway 
departure crashes caused by the driver drifting from his or her lane by 
28.9 percent, resulting in 24.3 percent fewer serious injuries.
    Kusano and Gabler (2015), simulated single-vehicle roadway 
departure crashes based on a sample of 478 crashes in the 2012 NASS-
CDS. Modelling was based on 54 model year 2010-2014 vehicles that were 
evaluated in NHTSA's New Car Assessment Program (NCAP). Kusano and 
Gabler found LDW systems could prevent 11-23 percent of drift-out-of-
lane crashes and 13-22 percent of serious to fatally injured drivers.
    As noted previously for frontal crash technologies, the agencies 
will base our projections on Leslie et al. because they are the most 
recent study, thereby reflecting the most current versions of these 
systems in the largest number of vehicles, and because they arguably 
have the most precise identification of the presence of the specific 
technologies in the vehicle fleet. However, unlike forward crash 
technologies, lane change technologies are operational at travel speeds 
where fatalities are likely to occur. Both LDW and LKA typically 
operate at speeds above roughly 35 mph. For this reason, and because 
the research noted in Table VI-246 indicates similar effectiveness 
against fatalities, injuries, and crashes, the agencies believe it is 
reasonable to assume the Leslie et al. crash reduction estimates are 
generally applicable to all crash severities, including fatal crashes. 
Our central effectiveness estimates are thus 10 percent for LDW and 20 
percent for LKA. For sensitivity analysis, the agencies adopt the 95 
percent confidence intervals from Flannagan & Leslie. For LKA this 
range is 14.95-25.15 percent. For LDW, the upper range was 4.95-13.93 
percent.
(c) Blind Spot Crash Technologies
    To address blind spot crashes, manufacturers are currently 
equipping vehicles with BSD, which detects vehicles in either of the 
adjacent lanes that may not be apparent to the driver. The system warns 
the driver of an approaching vehicle's presence to help facilitate safe 
lane changes and avoid crashes. A more advanced version of this, LCA, 
also detects vehicles that are rapidly approaching the driver's blind 
spot. Table VI-247 summarizes studies which have measured effectiveness 
for BSD and LCA.2071 2072 2073
---------------------------------------------------------------------------

    \2071\ Cicchino, J.B. (2017b). Effects of blind spot monitoring 
systems on police-reported lane-change crashes. Insurance Institute 
for Highway Safety, August 2017.
    \2072\ Leslie et al., supra note 2063.
    \2073\ Isaksson-Hellman, I., Lindman, M., An evaluation of the 
real-world safety effect of a lane change driver support system and 
characteristics of lane change crashes based on insurance claims. 
Traffic Injury Prevention, February 28, 2018: 19 (supp. 1).

---------------------------------------------------------------------------

[[Page 24806]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.458

[GRAPHIC] [TIFF OMITTED] TR30AP20.459

    Cicchino (2017) used Poisson regression to compare crash 
involvement rates per insured vehicle year in police-reported lane-
change crashes in 26[thinsp]U.S. States during 2009-2015 between 
vehicles with blind spot monitoring and the same vehicle models without 
the optional system, controlling for other factors that can affect 
crash risk. Systems designs across the 10 different manufacturers 
included in the study varied regarding the extent to which the size of 
the adjacent lane zone that they covered exceeded the blind spot area, 
speed differentials at which vehicles could be detected, and their 
ability to detect rapidly approaching vehicles, but these different 
systems were not examined separately. The study examined 4,620 lane 
change crashes, including 568 injury crashes. Cicchino found an overall 
reduction of 14 percent in blind spot related crashes of all 
severities, with a non-significant 23 percent reduction in injury 
crashes.
    Leslie et al. (2019) analyzed the relative crash performance of 
123,377 2013-2017 General Motors (GM) vehicles linked to State police-
reported crashes by Vehicle Identification numbers (VIN). GM provided 
VIN-linked safety content information for these vehicles to enable 
precise identification of safety technology content. The authors 
analyzed the effectiveness of a variety of crash avoidance technologies 
including both BSD and LCA separately. They estimated effectiveness 
comparing system-relevant crashes to baseline (control group) crashes 
using a quasi-induced exposure method in which rear-end struck crashes 
are used as the control group. Flannagan and Leslie found that BSD 
reduced lane departure crashes of all severities by 3 percent (non-
significant), and that LCA (which includes BSD) reduced these crashes 
by 26 percent.
    Isaksson-Hellman and Lindman (2018) evaluated the effect of the 
Volvo Blind Spot Information System (BLIS) on lane change crashes. 
Volvo's BLIS functions as an LCA, detecting vehicles approaching the 
blind spot as well as those already in it. The authors analyzed crash 
rate differences in lane change situations for cars with and without 
the BLIS system based on a population of 380,000 insured vehicle years. 
The authors found the BLIS system did not significantly reduce the 
overall number of lane change crashes of all severities, but they did 
find a significant 31 percent reduction in crashes with a repair cost 
exceeding $1250, and a 30 percent lower claim cost across all lane 
change crashes, indicating a reduced crash severity effect.
    Like lane change technologies, blind spot technologies are 
operational at travel speeds where fatalities are likely to occur. The 
agencies therefore assume the Leslie et al. crash reduction estimates 
are generally applicable to all crash severities, including fatal 
crashes. Our central effectiveness estimates are thus 3 percent for BSD 
and 26 percent for LCA. For sensitivity analysis, the agencies adopt 
the 95 percent confidence intervals from Flannagan & Leslie. For LCA 
this range is 16.59-33.74 percent. For BSD, the upper range was 14.72 
percent, but the findings were not statistically significant. The 
agencies therefore limit the range to 0-14.72 percent.
    Table VI-248 summarizes the effectiveness rates calculated in 
Leslie et al. and used in this analysis. Differences between the rates 
listed as ``Used in CAFE Fatality Analysis'' and those computed from 
Leslie et al. are explained in the above discussion.

[[Page 24807]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.460

(ii) Target Populations for Crash Avoidance Technologies
    The impact on fatality rates that will occur due to these 
technologies will be a function of both their effectiveness rate and 
the portion of occupant fatalities that occur under circumstances that 
are relevant to the technologies function. The agencies base our target 
population estimates on a recent study that examined these portions 
specifically for a variety of crash avoidance technologies including 
those analyzed here. Wang (2019) documented target populations for five 
groups of collision avoidance technologies in passenger vehicles 
including forward collisions, lane keeping, blind zone detection, 
forward pedestrian impact, and backing collision avoidance. The first 
three of these affect the light occupant target population examined in 
this analysis. Wang separately examined crash populations stratified by 
severity including fatal injuries, non-fatal injuries, and property 
damaged only (PDO) vehicles. She based her analysis on 2011-2015 data 
from NHTSA's Fatality Analysis Reporting System (FARS), National 
Automotive Sampling System (NASS), and General Estimates System (GES). 
FARS data was the basis for fatal crashes while nonfatal injuries and 
PDOs were derived from the NASS and GES.
    Wang followed the pre-crash typology concept initially developed by 
the Volpe National Transportation Systems Center (Volpe). Under this 
concept, crashes are categorized into mutually exclusive and distinct 
scenarios based on vehicle movements and critical events occurring just 
prior to the crash. Table VI-249 summarizes the portion of total annual 
crashes and injuries for each crash severity category that is relevant 
to the three crash scenarios examined.
[GRAPHIC] [TIFF OMITTED] TR30AP20.461

    The relevant proportions vary significantly depending on the 
severity of the crash. The rear-end crashes that are addressed by FCW 
and AEB technologies tend to be low-speed crashes and thus account for 
a larger portion of non-fatal injury and PDO crashes than for 
fatalities. Only 4 percent of fatal crashes occur in front-to-rear 
crashes, but over 30 percent of nonfatal crashes are this type. By 
contrast, fatal crashes are highly likely to involve inadvertent lane 
departure, 44 percent of all light vehicle occupant fatalities occur in 
crashes that involve lane departure, but only 17 percent of non-fatal 
injuries and 12 percent of PDOs involve this crash scenario. Blind spot 
crashes account for only about 2 percent of fatalities, 7 percent of 
MAIS1-5 injuries, and 12 percent of PDOs.
    The target population of this analysis is occupants of the light 
vehicles subject to CAFE. The values in Table VI-249 are portions of 
all crashes that occur annually. These include crashes of motor 
vehicles not subject to the current CAFE rulemaking such as medium and 
large trucks, buses, motorcycles, bicycles, etc. To adjust for this, 
the values in Wang were normalized to represent their portion of all 
light passenger vehicle (PV) crashes, rather than all crashes of any 
type. Wang provides total PV fatalities consistent with her technology 
numbers which are used as a baseline for this process. Based on 2011-
2015 FARS data, Wang

[[Page 24808]]

found an average of 29,170 PV occupant fatalities occurred annually.
    A second adjustment to Wang's results was made to make them 
compatible with the effectiveness estimates found in Leslie et al. In 
her target population estimate for lane departure warning, Wang 
included both head-on collisions and rollovers, but Leslie et al. did 
not. The Leslie et al. effectiveness rate is thus applicable to a 
smaller target population than that examined by Wang. To make these 
numbers more compatible, counts for these crash types were removed from 
Wang's lane departure totals.
    Electronic Stability Control (ESC) has been standard equipment in 
all light vehicles in the U.S. since the 2012 model year. ESC is highly 
effective in reducing roadway departure and traction loss crashes, and 
although it will be present in all future model year vehicles, it was 
present in only about 30 percent of the 2011-2015 on-road fleet 
examined by Wang. To reflect the impact of ESC on future on-road fleets 
therefore, the agencies further adjusted Wang's numbers to reflect a 
100 percent ESC presence in the on-road fleet. The agencies allocated 
the reduced roadway departure fatalities to the LDW target population, 
and the reduced traction loss fatalities to the AEB target population. 
This has the effect of reducing the total fatalities in both groups as 
well as in the total projected fatalities baseline.
    Table VI-250 summarizes the revised incidence counts and re-
calculated proportions of total PV occupant crash/injury. Revised 
totals are derived from original totals referenced in Table 1-3 in Wang 
(2019).
[GRAPHIC] [TIFF OMITTED] TR30AP20.462

(iii) Fleet Penetration Schedules
    The third element of the rule's safety projections is the fleet 
technology penetration schedules. Advanced safety technologies (ADAS) 
will only influence the safety of future MY fleets to the extent that 
they are installed and used in those fleets. These technologies are 
already being installed on some vehicles to varying degrees, but the 
agencies expect that over time, they will become standard equipment due 
to some combination of market pressure and/or safety regulation. The 
agencies adopt this assumption based on the history of most previous 
vehicle safety technologies, which are now standard equipment on all 
new vehicles sold in the U.S.
    The pace of technology adoption is estimated based on a variety of 
factors, but the most fundamental is the current pace of adoption in 
recent years. These published data were obtained from Ward's Automotive 
Reports for each technology.\2074\ Since these technologies are 
relatively recent, only a few years of data--typically 2 or 3 years--
were available from which to derive a trend. This makes these 
projections uncertain, but under these circumstances, a continuation of 
the known trend is the baseline assumption, which the agencies modify 
only when there is a rationale to justify it.
---------------------------------------------------------------------------

    \2074\ Derived from Ward's Automotive Yearbooks, 2014 through 
2018, % Factory Installed Electronic ADAS Equipment tables, 
weighting domestic and imported passenger cars and light trucks by 
sales volume.
---------------------------------------------------------------------------

    The technologies were examined in pairs reflecting their mutual 
target populations. Both FCW and AEB affect the same target 
population--frontal collisions. Both systems have been installed in 
some current MY vehicles, but their relative paces are expected to 
diverge significantly due to a formal agreement brokered by NHTSA and 
IIHS involving nearly all auto manufacturers, to have AEB installed in 
100 percent of their vehicles by September 2022 (MY 2023).\2075\ Wards 
first published installation rates for FCW and AEB for the 2016 model 
year and as of this analysis the 2017 MY is the latest data they have 
published. The agencies thus have data indicating that FCW was 
installed in 17.6 percent of MY 2016 vehicles and 30.5 percent of MY 
2017 vehicles. AEB was installed in 12.0 percent of MY 2016 vehicles 
and 27.0 percent of MY 2017 vehicles. AEB was installed in 12.0 percent 
of MY 2016 vehicles and 27.0 percent of MY 2017 vehicles. More recent 
reports submitted by manufacturers to the Federal Register indicate 
that installation rates accelerated in MY 2018 and 2019

[[Page 24809]]

vehicles. Four manufacturers, Tesla, Volvo, Audi, and Mercedes, have 
already met their voluntary commitment of 100 percent installation 3 
years ahead of schedule. During the period September 1, 2018 through 
August 31, 2019, 12 of the 20 manufacturers equipped more than 75 
percent of their new passenger vehicles with AEB, and overall 
manufacturers equipped more than 9.5 million new passenger vehicles 
with AEB.\2076\
---------------------------------------------------------------------------

    \2075\ See https://www.nhtsa.gov/press-releases/nhtsa-iihs-announcement-aeb.
    \2076\ See NHTSA Announces Update to Historic AEB Commitment by 
20 Automakers. December 17, 2019. https://www.nhtsa.gov/press-releases/nhtsa-announces-update-historic-aeb-commitment-20-automakers.
---------------------------------------------------------------------------

    Because of the NHTSA/IIHS agreement, the agencies assume that AEB 
will be in 100 percent of light vehicles by the 2023 MY. To derive 
installation rates for MYs 2018 through 2022, the agencies interpolate 
between the MY 2017 rate of 27 percent and the MY 2023 rate of 100 
percent. To derive a MY 2015 estimate, the agencies modelled the 
results for MYs 2016-2023 and calculated a value for year x=0, 
essentially extending the model results back one year on the same 
trendline.
    For FCW, the agencies used the same interpolation/modeling method 
as was used for AEB to derive an initial baseline trend. However, while 
both systems are available on some portion of the current MY fleet, the 
agencies anticipate that by MY 2023, all vehicles will have AEB systems 
that essentially encompass both FCW and AEB functions. The agencies 
therefore project a gradual increase in both systems until the sum of 
both systems penetration rates exceeds 100 percent. At that point, the 
agencies project a gradual decrease in FCW only installations until FCW 
only systems are completely replaced by AEB systems in MY 2023.
    For LDW, Wards penetration data were available as far back as MY 
2013, giving a total of 5 data points through MY 2017. The projection 
for LDW was derived by modelling these data points. The data indicate a 
near linear trend and our initial projections of future years were 
derived directly from this model. Wards did not report any of the more 
advanced LKA systems until MY 2016, leaving only 2 data points. The 
agencies modelled a simple trendline through these data points to 
estimate the pace of future LKA installations. As with Frontal crashes, 
the agencies assume a gradual phase-in of the most effective 
technology, LKA, will eventually replace the lesser technology, LDW, 
and the agencies allow gradual increases in both systems penetration 
until their sum exceeds 100 percent, at which point LDW penetration 
begins to decline to zero while LKA penetration climbs to 100 percent.
    For blind spot crashes, Wards data was available for MYs 2013-2017 
for BSD, but no data was available to distinguish LCA systems. LCA 
systems were available as optional equipment on at least 10 MY 2016 
vehicles.\2077\ In addition, Flannagan and Leslie found numerous cases 
in State data-bases involving vehicles with LCA. Because LCA data is 
not specifically identified, the agencies will estimate its frequency 
based on the samples found in Flannagan & Leslie. In that study, 62 
percent of vehicles with blind spot technologies has BSD alone, while 
38 percent had LCA (which includes BSD). The agencies employ this ratio 
to establish the relative frequency of these technologies in our 
projections. As with frontal and lane change technologies, the agencies 
assume a gradual phase-in of the most effective technology, LCA, will 
eventually replace the lesser technology, BSD, and the agencies allow 
gradual increases in both systems penetration until their sum exceeds 
100 percent, at which point BSD penetration begins to decline to zero 
while LCA penetration climbs to 100 percent.
---------------------------------------------------------------------------

    \2077\ See, e.g. https://www.autobytel.com/car-buying-guides/features/10-cars-with-lane-change-assist-using-cameras-or-sensors-130847.
---------------------------------------------------------------------------

(iv) Impact Calculations
    Table VI-251, Table VI-252, and Table VI-253 summarize the 
resulting estimates of impacts on fatality rates for frontal crash 
technologies, lane change technologies, and blind spot technologies 
respectively for MYs 2016-2035. All previously discussed inputs are 
shown in the tables. The effect of each technology is the product of 
its effectiveness, it's percent installation in the MY fleet, and the 
portion of the total light vehicle occupant target population that each 
technology might address. Since installation rates for each technology 
apply to different portions of the vehicle fleet (i.e., vehicles have 
either the more basic or more advanced version of the technology), the 
effect of the two technologies combined is a simple sum of the two 
effects. Likewise, since each crash type addresses a unique target 
population, there is no overlap among the three crash types and the sum 
of the normalized crash impacts across all three crash types represents 
the total impact on fatality rates from these 6 technologies for each 
model year. These cumulative results are shown in the last column of 
Table VI-253. As technologies phase in to newer MY fleets,\2078\ their 
impact on the light vehicle occupant fatality rate increases 
proportionally to roughly 8.5 percent before levelling off. That is, 
eventually, by approximately MY 2026, these technologies are expected 
to reduce fatalities and fatality rates for new vehicles by roughly 8.5 
percent below their initial baseline levels.
---------------------------------------------------------------------------

    \2078\ While it is technically possible to retrofit these 
systems into the on-road fleet, such retrofits would be 
significantly more expensive than OEM installations. The agencies 
thus assume all on-road fleet penetration of these technologies will 
come through new vehicle sales.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24810]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.463

[GRAPHIC] [TIFF OMITTED] TR30AP20.464


[[Page 24811]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.465

BILLING CODE 4910-59-C

[[Page 24812]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.466

(b) Fatality Trend Model
    The revised fatality trend model differs from the model employed in 
the NPRM in four main respects:
     The fatality rates for individual model years and ages 
were re-calculated to correct the counts of fatalities to occupants of 
light-duty vehicles and to reflect the revised VMT estimates, the 
latter of which incorporate revisions to both vehicle registration 
counts and the estimated relationship between vehicle age and annual 
use; \2079\
---------------------------------------------------------------------------

    \2079\ These revised estimates of the number of miles traveled 
by vehicles of each model year during past calendar years were 
developed from the expanded sample of vehicles' odometer readings 
obtained by NHTSA.
---------------------------------------------------------------------------

     In response to comments on the version used in the NPRM, t 
model adds controls for changes to factors (such as driver demographics 
and behavior, and geographic patterns of travel) that can affect 
fatality rates for vehicles of all model years and ages;
     The revised analysis clusters past model years into 
``safety cohorts,'' which are groups of successive model years that 
exhibit similar fatality rates during their first years of use, in 
order to represent the actual historical pattern of safety improvements 
more realistically; and
     The model employs a slightly less complex mathematical 
relationship between a model year's age and its fatality rate 
(fatalities per mile driven), which still describes the observed 
relationship accurately.
    Similar to the fatality trend model employed in the proposal, the 
revised estimates of annual travel were combined with tabulations of 
annual fatalities occurring among occupants of light-duty vehicles of 
each model year during past calendar years, tabulated from NHTSA's FARS 
data. Fatalities occurring in vehicles produced during each model year 
making up a calendar year's light-duty vehicle fleet are divided by the 
estimated number of miles they were driven during that calendar year to 
calculate historical fatality rates by model year and calendar year, 
measured as fatalities per billion miles traveled. These data represent 
the dependent variable in the revised statistical model of fatality 
rates.
    Longitudinal or time-series analyses such as the model of 
historical variation in fatality rates for individual model years need 
to incorporate three separate effects to account for all potential 
sources of variation. First, they need to employ model year in some 
form as an explanatory variable, to account for improvements in the 
safety of vehicles produced during successive model years that persist 
throughout their lifetimes in the vehicle fleet. This is an example of 
a ``cohort effect'' in the age-period-cohort framework that is widely 
used to of analysis of population-wide behavior.\2080\ Second, such a 
model must account for the effect of age on the safety of each 
individual model year as it grows older, accumulates mileage, and in 
most cases changes ownership one or more times during its expected 
service lifetime (the ``aging effect'' in age-period-cohort analysis).
---------------------------------------------------------------------------

    \2080\ For a detailed explanation of the rationale and methods 
for age-period-cohort analysis, see for example Columbia University 
Mailman School of Public Health, Population Health Methods: Age-
Period-Cohort Analysis, available at https://www.mailman.columbia.edu/research/population-health-methods/age-period-cohort-analysis (accessed February 12, 2020); and Kupper, 
Lawrence L. et al., ``Statistical age-period-cohort analysis: A 
review and critique,'' Journal of Chronic Diseases 38:10 (1985), at 
811-830, available at https://www.sciencedirect.com/science/article/abs/pii/0021968185901055#! (accessed February 12, 2020).
---------------------------------------------------------------------------

    Finally, most longitudinal analyses, including the historical 
safety model developed here, need to account explicitly for factors 
that vary over time--in this case, calendar years. By doing so, they 
can affect the safety of vehicles of all model years and ages making up 
the fleet during successive calendar years, or change the composition 
of total travel by vehicles of different model years and ages. In 
either case, such time-related factors--often referred to as ``period 
effects''--can change the overall safety performance of the entire 
fleet from one calendar year to the next, independently of and in 
addition to the changes that would result from the combination of new 
model years entering the fleet while older ones are retired from 
service (the cohort effect), and the aging of all model years making up 
the fleet. For example, an increase in seat belt use among all drivers 
during a calendar year would be expected to reduce the fatality rates 
of vehicles of all model years and ages in use during that year, while 
an economic recession may change the composition of drivers and 
vehicles on the road during a calendar year. In either case, one result 
will be a change in the fleet-wide composite fatality rate for that 
calendar year.
    Figure VI-83 below illustrates the contributions of cohort, aging, 
and time-period effects to changes over time in population-wide 
behavior. As the figure indicates, these effects are conceptually 
independent, but interact in ways that combine to produce the observed 
historical evolution of the fleet-wide fatality rate for light-duty 
vehicle occupants. Again, calendar year or time-period factors can 
affect the safety performance of the entire fleet independently of the 
effect that would result from the combination of changes in the 
specific model years making up the fleet and the advancing ages of all 
model years, and any ``period effect'' effect attributable to factors 
that vary over time is in addition to cohort and aging effects.

[[Page 24813]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.467

    To introduce such period effects into the fatality trend model, 
which were absent from the NPRM analysis, the agencies obtained 
historical data on factors that varied by calendar year, and were 
expected to be responsible for such effects. As indicated previously, 
these included the following:
     Seat belt use, as measured by the fraction of drivers 
observed to be wearing lap and shoulder belts, estimated by NHTSA's 
National Occupant Protection Survey (NOPUS);
     Driving under the influence of alcohol or drugs, measured 
by the fraction of drivers reporting having recently done so in surveys 
conducted by the U.S. Centers for Disease Control (CDC); \2081\
---------------------------------------------------------------------------

    \2081\ The agencies also experimented with measures of drivers 
appearing to be under the influence of alcohol or drugs included in 
NHTSA's NOPUS, available at https://crashstats.nhtsa.dot.gov/#/PublicationList/18.
---------------------------------------------------------------------------

     Use of hand-held electronic devices, measured by the 
fraction of drivers visually observed to be doing so in NHTSA's NOPUS;
     The fraction of licensed drivers who are male and under 
the age of 25 (historically the riskiest cohort of drivers), as 
reported by the FHWA's annual Highway Statistics publication; \2082\
---------------------------------------------------------------------------

    \2082\ Federal Highway Administration, Highway Statistics, 
various years, Table DL-20, available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
---------------------------------------------------------------------------

     The fraction of miles traveled in rural areas, also as 
reported by FHWA; \2083\ and
---------------------------------------------------------------------------

    \2083\ Federal Highway Administration, Highway Statistics, 
various years, Table VM-1, available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
---------------------------------------------------------------------------

     The overall performance of the U.S. economy, as measured 
by the annual rate of unemployment.\2084\
---------------------------------------------------------------------------

    \2084\ See Bureau of Labor Statistics, historical data series 
LNS14000000, available at https://data.bls.gov/cgi-bin/surveymost?ln.
---------------------------------------------------------------------------

    The agencies were unable to obtain useful measures of roadway 
design parameters or road conditions that would be expected to affect 
safety. Although such measures exist, they tend to be reported for 
individual road and highway segments or routes, and it is difficult to 
combine these data into meaningful, aggregate measures that describe 
overall driving conditions that are likely to vary by calendar year. 
Nor could they identify satisfactory measures of incident response time 
or the effectiveness of emergency medical treatment in reducing the 
consequences of injuries occurring in motor vehicle crashes.
    An important challenge to incorporating these time-period effects 
into the fatality trend model arose from the fact that their patterns 
of variation over the historical period the agencies analyzed (which 
extended from calendar year 1995 to 2017) were extremely closely 
correlated, making it virtually impossible to distinguish their 
independent contributions to improvements in fleet-wide safety over 
time. Table VI-254 below reports the pairwise correlation coefficients 
among the potential measures of period effects listed above. As it 
suggests, patterns of variation about their respective mean values over 
the period analyzed were very similar (with the exception of the 
unemployment rate), and the resulting high statistical correlations (or 
``collinearity'') among them made it nearly impossible to identify 
their independent effects on variation in safety over time, even when 
controlling for the effects of model year and vehicle age.

[[Page 24814]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.468

    To address this difficulty, the agencies substituted a time trend--
that is, a variable that takes the value of one in the first calendar 
year and increases by one in each successive calendar year--in an 
effort to capture the joint movements in the variables that were 
intended to measure time-period effects on safety. The agencies 
experimented with both linear and more complex time trends to capture 
the apparently declining rate of improvement in fleet-wide safety over 
time, but found that the linear trend captured the combined effects 
most reliably. Because the model's dependent variable is the natural 
logarithm of model year and age-specific fatality rates, using a linear 
time trend corresponds to assuming a constant percentage decline in 
fatality rates each year (rather than a constant absolute decline each 
year), and this pattern appeared to provide the best fit to the 
observed historical pattern of safety improvements. Finally, after 
noting that the linear time trend did not fully capture the effects on 
fleet-wide safety associated with the economic recessions in 2001 and 
2007-11, the agencies supplemented the time trend with indicator (or 
``dummy'') variables for these years, finding that only those for 2008, 
2009, and 2010 improved its explanatory power significantly.
    Another significant improvement to the NPRM analysis was to group 
model years into ``safety cohorts'' on the basis of similarity in their 
fatality rates when new (that is, during their first year in service), 
rather than treating each model year as a separate cohort. Groupings 
were created through a combination of identifying years when new safety 
regulations initially took effect or were phased in, examining of 
first-year fatality rates, and limited statistical experimentation. 
Grouping successive model years reduces the number of cohorts 
significantly, since similar fatality rates were typically observed for 
at least five, and sometimes as many as ten, consecutive model years 
over the historical period the agencies examined. Grouping model years 
into a smaller number of cohorts rather than treating each model year 
as a separate cohort offers the advantage of introducing some variation 
in the ages of vehicles making up the same cohort during a calendar 
year, which improves the statistical reliability with which the 
independent effect of age itself can be estimated.
    Figure VI-84 below shows historical variation in the fatality rates 
of past model years when each one was newly-introduced (i.e., during 
its first year in use).\2085\ It clearly displays the significant 
improvement in the safety of new vehicles over time in response to 
improvements in safety features, including those required by NHTSA's 
safety regulations. The figure also clearly documents the natural 
clustering of fatality rates for successive model years that was used 
to identify and define the safety cohorts used in the revised model. In 
the panel structure of the model, which combines time-series and cross-
section variation in fatality rates for individual model years as their 
ages vary across calendar years, the clustering of first-year fatality 
rates for successive model years is captured by using separate ``fixed 
effects'' for each safety cohorts illustrated in the figure. Some 
judgment is inevitably required to distinguish between successive 
cohorts and identify when the fatality rate for new model years has 
changed significantly; the agencies experimented with using from five 
to eight cohorts, ultimately finding that the agencies could 
distinguish most reliably among the fatality rates for five cohorts.
---------------------------------------------------------------------------

    \2085\ For simplicity, the figure assumes that each model year's 
first year of use was the calendar year identical to its designated 
model year; for example, the first full year of use for model year 
2000 was assumed to be calendar year 2000. In fact, new vehicles 
frequently become available for purchase during the calendar year 
preceding their designated model year and continue to be sold 
through the calendar year following it, although most sales occur 
during the calendar year matching their designated model year.

---------------------------------------------------------------------------

[[Page 24815]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.469

    A final revision to the NPRM model was to employ a slightly less 
complex mathematical relationship between a model year's age and its 
fatality rate than had been used in the NPRM version. Specifically, the 
revised model relates fatality rates to age itself as well as the 
second and third powers of age (that is, age squared and age cubed), 
but omits the fourth power of age, which was included in the model 
developed for the NPRM. This slightly simpler relationship proved 
adequate to capture fully the complex--but strongly recurring--pattern 
of fatality rates for past model years as they aged. Specifically, as 
Figure VI-85 below shows, fatality rates have tended to remain 
approximately constant for the first few years of most recent model 
years' lifetimes, before increasing steadily through age 15-20 and then 
declining gradually over the remainder of their lifetimes.
    As discussed previously, the increase in fatality rates through 
approximately age 20 is generally thought to result primarily from the 
fact that used vehicles are commonly purchased and driven by members of 
households whose demographic characteristics, driving behavior, and 
geographic locations are associated with more risky driving behavior 
and thus more frequent or severe crashes. Of course, increased 
frequency of mechanical failures as vehicles age and accumulate mileage 
also seems likely to contribute to this pattern. In contrast, the 
consistent tendency for fatality rates to decline after about age 20 is 
less well understood, but may owe partly to the demographic 
characteristics and driving behavior of owners of very old vehicles. 
Whatever its source, the number of vehicles remaining in service past 
age 20 is so small and their use typically so limited that their 
contribution to the fleet-wide fatality rate is minimal.

[[Page 24816]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.470

    Figure VI-85 documents the relationship between age and fatality 
rate for selected past model years.\2086\ As it shows, fatality rates 
for recent model years follow a complex but strikingly similar pattern 
of increase and subsequent decline with increasing age, although the 
figure also shows that the earliest model years included in the sample 
(1975-1980) tended not to display increasing fatality rates in the 
first half of their lifetimes. At the same time, the figure illustrates 
the gradual downward shift in fatality rates at all ages for successive 
past model years, although there is considerable variation in the 
extent of this shift for individual model years, particularly when they 
are examined at specific ages. That is, the downward shift in fatality 
rates for successive model years is not necessarily ``monotonic,'' 
particularly when it is examined at specific individual ages.
---------------------------------------------------------------------------

    \2086\ For a color version, see the corresponding safety 
discussion in the accompanying FRIA.
---------------------------------------------------------------------------

    The agencies believe that the increase in fatality rates for cars 
and light trucks produced during recent model years through 
approximately age 20 reflects the fact that as aging vehicles change 
ownership via the used car market, they are often purchased and driven 
by households whose demographic characteristics and locations are 
associated with riskier driving behavior and conditions. The decline in 
vehicles' fatality rates after this age is not well understood, but 
seems likely to reflect the fact that the relatively small fraction of 
those originally produced in a model year that survive beyond age 20-25 
are owned and driven by households that maintain them carefully, are 
likely to reside in areas where driving conditions are safest, and 
whose members engage in less risky driving behavior.
    After examining the information summarized in Figure VI-85, the 
agencies conclude that the effect of increasing age on vehicle safety 
appears to be largely independent of the improvement in new cars' 
fatality rates over successive model years, and appears to operate 
similarly for all except the earliest model years in our historical 
sample (which includes model years 1975-2017).\2087\ As a formal 
statistical test, the agencies experimented with allowing the aging 
effect to change across model years when the agencies estimated the 
revised model, anticipating that newer safety technologies and vehicle 
designs might ``flatten'' the relationship between fatality rates and 
age--that is, reduce the degree to which fatality rates increased over 
the 5-20 year range of vehicle ages--for newer model years. However, 
the agencies found no evidence that the effect of age on safety changed 
significantly for more recent model years compared to older ones, so 
the agencies retained the assumption of identical aging effects for all 
model

[[Page 24817]]

years in the revised model.\2088\ Thus the revised model shows 
progressively lower fatality rates for more recent model years when 
they are new, but fatality rates for all model years increase with age 
and subsequently decline according to the same non-linear pattern 
displayed in Figure VI-85. On a related question, the agencies also 
found that including the squared and cubed values of age in addition to 
age itself as explanatory variables in the model, while excluding the 
fourth power of age, which had been included in the NPRM model, proved 
adequate to capture the pattern of variation in fatality rates with 
increasing age that most past model years have exhibited. Table VI-255 
below reports the estimated parameter values for alternative 
specifications of the model, together with various goodness-of-fit and 
other diagnostic measures. The analysis described in the following 
section uses the estimated time trend from Model 2 in the table, which 
implies annual reduction in fatality rates for all model years of 2.14 
percent.
---------------------------------------------------------------------------

    \2087\ Of course, the agencies cannot observe the safety 
performance of all model years included in the agencies' data sample 
over their entire lifetimes, because the data the agencies use to 
estimate the model start in calendar year 1990, by which time all 
model years before 1990 were no longer new--for example, MY1975 cars 
are already 15 years old by then--while the newest model years in 
the agencies' sample are still very ``young'' when the agencies' 
data ends in calendar year 2017. Thus, the agencies have only 
incomplete information about the relationship of fatality rates to 
age over the entire lifetimes of these model years, so it is 
possible that this relationship differs at particularly early or 
advanced ages for the oldest and newest model years in the agencies' 
sample.
    \2088\ Specifically, the agencies tested for interactions 
between the age and model year variables, which would reveal changes 
in the relationship between fatality rates and age for more recent 
model years, but found that such interaction effects were generally 
not statistically significant. Allowing for interactions between age 
and the indicator variables for safety cohorts (recall that these 
represent groupings of successive model years) produced this same 
result--few of the interaction effects were statistically 
significant.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24818]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.471

BILLING CODE 4910-59-C
BILLING CODE 4910-59-P

[[Page 24819]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.472


[[Page 24820]]


BILLING CODE 4910-59-C
Using the Model and Technology Analysis to Forecast Fatality Rates
    The newest safety cohort includes model years from 2009 to 2017, so 
in effect the agencies estimate that all those model years have 
essentially the same fatality rate in their first year of use. The 
agencies apply the estimated effectiveness of crash avoidance 
technologies in reducing fatal crashes to the observed fatality rate 
for model years 2009 to 2017 vehicles during their first year in use to 
estimate fatality rates for future model years during the first year 
each one is introduced. Figure VI-86 below shows the result of this 
process; as it indicates, fatality rates for new model years decline 
gradually through 2035 and then stabilize, reflecting the fact that the 
agencies are only able to project the effectiveness of emerging crash 
avoidance technologies on the safety of new vehicles through that year.
[GRAPHIC] [TIFF OMITTED] TR30AP20.473

    The next step in constructing the forecast of fleet-wide fatality 
rates is to apply the age-related increases in the fatality rate for 
each model year making up the previous calendar year's fleet. For 
example, the agencies assume that the fatality rates for all model 
years comprising the light-duty vehicle fleet in 2017 increase with age 
according to the relationship captured by the estimated coefficients on 
the age variables in the preferred model specification shown in Table 
VI-255. The same assumption is applied to all new model years 
introduced in subsequent years. Finally, the agencies also assume that 
the historical decline in fatality rates observed over past calendar 
years (the ``period effect'' captured by the time trend variable) will 
continue into the future. This implies that fatality rates for all 
model years and ages will decline by an additional 2.41 percent in each 
successive future calendar year from the rates that would have resulted 
from the combined effects of continuing improvements in the safety of 
newly-introduced model years and the effect of increasing age.\2089\
---------------------------------------------------------------------------

    \2089\ The agencies do not apply this trend reduction to the 
fatality rates for the newest model year in each calendar year's 
fleet, because it is assumed to be independent of both the decline 
in new-car fatality rates and the aging effect.
---------------------------------------------------------------------------

    This process produces an estimate of the fatality rate for each 
model year making up the fleet during each future calendar year. That 
estimate reflects the combination of (1) reductions in fatality rates 
for new cars, reflecting the continued improvements in their safety due 
to crash avoidance technologies (through MY2035); (2) increases in the 
fatality rates for each model year in the fleet from the previous 
calendar year, which represent the effect of age estimated by the 
historical model; and (3) the continuing downward trend in fatality 
rates for all vehicles except the newest model year in each calendar 
year's fleet, which is derived from the historical model.
    The agencies then weight the fatality rate for each model year 
making up a future year's fleet by the fraction of total fleet-wide VMT 
it accounts for, and sum

[[Page 24821]]

the results to produce an estimate of the fleet-wide fatality rate. The 
CAFE model does not actually use this fleet-wide fatality rate, because 
all of the fatality calculations are performed separately for each 
individual model year making up the fleet, which are then aggregated; 
nevertheless, the agencies provide the fleet-wide rate as a useful 
check on the reasonableness of our fatality rate forecasts for 
individual model years as they enter the fleet and age over their 
respective lifetimes. Figure VI-87 displays the projected fleet-wide 
fatality rates for future calendar years, as well as the trend in their 
recent historical values.
[GRAPHIC] [TIFF OMITTED] TR30AP20.474

(d) Impact of Advanced Technologies on Older Vehicle Fatality Rates
    In the NPRM, the agencies calculated the potential safety impacts 
of delayed purchases of vehicles with new safety technology that might 
result from higher vehicles prices associated with more stringent CAFE 
standards. A number of commenters noted that since these improvements 
will be driven by crash avoidance technologies, they will also benefit 
older vehicles and reduce their fatality rates as well. For example, 
CARB noted that ``safety improvements generally provide systematic 
safety benefits to all vehicles in the on-road fleet, not only to new 
vehicles. However, NHTSA's safety model assigns safety coefficients to 
vehicles solely based on their model year and it fails to incorporate 
the effect that new safety designs and technologies will have on 
systematically improving fleet-wide on-road safety.'' IPI similarly 
noted that should ``new safety technologies be adopted, the predicted 
fatalities for all the older vehicle vintages will have to be lowered 
as well because effective crash avoidance technologies will lower all 
vehicles' fatality costs.''
    The agencies agree that the users of older vehicles will also 
benefit from crash avoidance technologies on newer vehicles. In 
response, the agencies have modified our methodology to reflect lower 
fatality rates on older vehicles resulting from the new crash avoidance 
technologies. Crash avoidance technologies prevent crashes from 
happening and thus benefit both the vehicle with the technology and any 
other vehicles that it might have collided with. However, the scope of 
these impacts on older vehicle's fatality rates are somewhat limited 
due to several factors:
    Single vehicle crashes, which make up about half of all fatal 
crashes, will not be affected. Only multi-vehicle crashes involving a 
newer vehicle with the advanced technology and an older vehicle will be 
affected. Multi-vehicle crashes account for roughly half of all light 
vehicle occupant fatalities.
     For a new safety technology to benefit an older vehicle in 
a multi-vehicle crash, the vehicle with the technology must have been 
in a position to control, or prevent the crash. For example, in front-
to-rear crashes which can be addressed by FCW and AEB, the older 
vehicle would only benefit if it was the vehicle struck from behind. If 
the struck vehicle were the newer vehicle, its AEB technology would not 
prevent the crash. Logically this would occur in roughly half of two-
vehicle crashes and a third of all three-vehicle crashes. Since most 
multi-vehicle crashes involve only two vehicles, roughly half of all 
multi-vehicle crashes might qualify.
     The benefits experienced by older vehicles are 
proportional to the probability that the vehicles they collide with are 
newer vehicles with advanced crash avoidance technology. The

[[Page 24822]]

agencies estimate that the probability that this would occur is a 
function of the relative exposure of vehicles by age, measured by the 
portion of total VMT driven by vehicles of that age. Based on VMT 
schedules (see CY 2016 example in Table VI-256), new (current MY) 
vehicles account for about 9.6 percent of annual fleet VMT. The 
relevant portion would increase over time as additional MY vehicles are 
produced with advanced technologies. However, the portion of older 
vehicle crashes that might be affected by newer technologies is 
initially very small--only about 2 percent (.5*.5*.096) of older 
vehicles involved in crashes might benefit from advanced crash 
avoidance technologies in other vehicles in the first year.
[GRAPHIC] [TIFF OMITTED] TR30AP20.475


[[Page 24823]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.476

    To reflect this safety benefit for older vehicles, the agencies 
calculated a revised fatality rate for each older MY vehicle on the 
road based on its interaction with each new MY starting with MY 2021 
vehicles based on the following relationship:

Revised fatality rate = Fm-((x-y)mnp) + F(1-m)

Where: F = initial fatality rate for each MY

x = baseline MY fatality rate
y = current MY fatality rate
m = proportion of occupant fatalities that occur in multi-vehicle 
crashes (52 percent)
n = probability that crash is with a new MY vehicle containing 
advanced technologies
p = probability that new vehicle is ``striking'' vehicle

    The initial fatality rate for each vehicle MY (F) was derived by 
combining fatality counts from NHTSA's Fatality Analysis Reporting 
System (FARS) with VMT data from IHS/Polk.
    The baseline MY fatality rate (x) represents the baseline rate over 
which the impact of new crash avoidance technologies should be measured 
It establishes the baseline rate for each MY that will be compared to 
the most current MY rate to determine the change in fatality rate (FR) 
for each MY. The relative effectiveness of new crash-avoidance 
technologies in modifying the fatality rate of older model vehicles is 
measured differently depending on the age of the older vehicle. The 
fatality rate is a historical measure that reflects safety differences 
due to both crashworthiness technologies such as air bags and crash 
avoidance technologies such as electronic stability control, but up 
through MY 2017, crashworthiness standards are the predominant cause of 
these differences.
    The most recent significant crashworthiness safety standard, which 
upgraded roof strength standards which was effective in all new 
passenger vehicles in MY 2017. Crashworthiness standards would not have 
secondary benefits for older MY vehicles. Post MY 2017, the agencies 
believe crash avoidance technologies will drive safety improvements. To 
isolate the added crash avoidance safety expected in newer vehicles, 
the marginal impact of the difference between the MY 2017 fatality rate 
and the most current MY fatality rate represents the added marginal 
effectiveness of new crash-avoidance technologies of each subsequent MY 
for MYs 2017 and earlier. Beginning with MY 2018, the difference 
between the older MY fatality rate and most current MY rate determines 
the potential safety benefit for the older vehicles.

[[Page 24824]]

    The current MY fatality rate (y), represents the projected fatality 
rate of future MY vehicles after adjustment for the impacts of the 
advanced crash avoidance technologies and projected improvements in 
non-technology factors examined in this analysis. This process was 
discussed in detail in the previous section.
    The proportion of passenger vehicle occupant fatalities that occur 
in multi-vehicle crashes (m), was derived from an analysis of occupants 
of fatal passenger vehicle crashes from 2002-2017 FARS. The analysis 
indicated that 47.8 percent of fatal crash occupants were in single 
vehicle crashes, 40.2 percent were in two vehicle crashes, and 12 
percent were in crashes involving 3 or more vehicles. Overall, 52.2 
percent were in multi-vehicle crashes.
    The portion of older vehicle crashes involving newer vehicles 
containing advanced crash avoidance technologies (n), is assumed to be 
equal to the cumulative risk exposure of vehicles that have these 
technologies. This exposure is measured by the product of annual VMT by 
vehicle age and registrations of vehicles of that age. The CAFE model 
calculates this dynamically, but as an example, based on 2016 
registration data (see Table VI-256 above), the most current MY would 
represent 9.6 percent of all VMT in a calendar year, implying a 9.6 
percent probability that the vehicle encountered would be from the most 
current MY. This percentage would increase for each CY as more MY 
vehicles adopt advanced crashworthiness technologies. The agencies note 
that other factors such as uneven concentrations of newer vs. older 
vehicles or improved crash avoidance in the younger vehicles already on 
the road that are the basis for the agencies' VMT proportion table 
might disrupt this assumption, but it is likely that this would only 
serve to slow the probability of these encounters, making this a 
conservative assumption in that it maximizes the probability that older 
vehicles might benefit from newer technologies.
    The probability that the vehicle with advanced crash avoidance 
technology is the controlling or striking vehicle (p), was calculated 
using the relative frequency of fatal crash occupants in multi-vehicle 
crashes. As noted previously, 40.2 percent were in two vehicle crashes, 
and 12 percent were in crashes involving 3 or more vehicles. The 
agencies assume a probability of 50 percent for two vehicle crashes and 
33 percent for crashes with 3 or more vehicles. Weighted together the 
agencies estimate a 46.1 percent probability that, given a multi-
vehicle crash involving a vehicle with advanced technologies and an 
older vehicle without them, the newer vehicle will be the striking 
vehicle or in a position where its crash avoidance technologies might 
influence the outcome of the crash with the older vehicle.
    This process is illustrated in Table VI-257 below for adjustments 
due to improvements in MY 2021 vehicles back through MY 1995. In Table 
VI-257, the actual model year fatality rate is shown in the second 
column. As noted above, the base fatality rate, shown in column 3, is 
the MY 2017 rate for all MYs prior to 2018, after which it becomes the 
actual MY rate. Column 4 shows the difference between the fatality rate 
for MY 2021 and the base rate for each MY. Column 5 shows the resulting 
revised fatality rate that would be used for each older MY, and column 
6 and 7 list the change in that rate. The various factors noted in the 
above formula are applied in column 5. The results indicate a 0.006 
decrease in pre-2018 MY vehicles fatality rates, with declining impacts 
going forward to MY 2021. In subsequent years, this impact would grow 
to reflect the both the increased probability that an older vehicle 
would crash with vehicles containing advanced technology, as well as 
the increased technology levels in progressively newer vehicles. This 
table was created using NPRM inputs and is provided for explanatory 
purposes only. The actual impacts are dynamically calculated within the 
Volpe model and reflect revised fatality rate trends going forward and 
cover even older model years.

[[Page 24825]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.477

(e) Dynamic Fleet Composition
    As described in the sales discussion in Section Dynamic Fleet Share 
(DFS), the standards may impact the distribution of cars and trucks 
purchased. As light trucks, SUVs and passenger cars respond differently 
to technology applied to meet the standards--namely mass reduction--
fleets with different compositions of body styles will have varying 
amounts of fatalities. Since mass-safety fatalities are calculated by 
multiplying mass point-estimates by VMT, which implicitly captures the 
impact of the dynamic fleet share model, the estimates of mass-safety 
fatalities in the previous section include the impact of vehicle prices 
on fleet composition.
(c) Impact of Rebound Effect on Fatalities
    The ``rebound effect'' is a measure of the additional driving that 
occurs when the cost of driving declines. More stringent standards 
reduce vehicle operating costs, and in response, some consumers may 
choose to drive more. Driving more increases exposure to risks 
associated with on-road transportation, and this added exposure 
translates into higher fatalities. The agencies have calculated this 
impact by estimating the change in VMT that results from alternative 
standards.
    As noted previously, rebound miles are not imposed on consumers by 
regulation. They are a freely chosen activity resulting from reduced 
vehicle operational costs. As such, the agencies believe a large 
portion of the safety risks associated with additional driving are

[[Page 24826]]

offset by the benefits drivers gain from added driving. For the 
proposal, the agencies assumed that, in deciding to drive more, drivers 
internalize the full cost to themselves and others, including the cost 
of accidents, associated with their additional driving.
    In response to the NPRM, EDF noted that consumers may not fully 
value the added safety risk, such as risk to other drivers.\2090\ In 
making this point, EDF suggested a value of 50 percent would be 
conservative, but did not provide supporting evidence for that value. 
The agencies agree that the level of risk internalized by drivers is 
uncertain, and for the final rule have revised the portion of the added 
monetized safety risk that consumers internalize to 90 percent, which 
mostly offsets the societal impact of any added fatalities from this 
voluntary consumer choice.
---------------------------------------------------------------------------

    \2090\ EDF, Appendix B, NHTSA-2018-0067-12108, at 101.
---------------------------------------------------------------------------

    The actual portion of risk from crashes that drivers internalize is 
unknown. The agencies suspect that drivers are more likely to 
internalize serious crash consequences than minor ones, and some 
drivers may not perfectly internalize injury consequences to other 
individuals, especially occupants of other vehicles and pedestrians. 
However, legal consequences from crash liability, both criminal and 
civil, should also act as a caution for drivers considering added crash 
risk exposure. The agencies considered several approaches to estimating 
internalized crash risk. The first assumes that drivers value harm to 
themselves as well as legal liability for causing harm to others. It 
considers that all fatalities in single vehicle crashes are fully 
valued, that there is roughly a 50 percent chance that each driver 
would be the one killed in multi-vehicle crashes, and that there is 
roughly a 50 percent chance that each driver would be at-fault in a 
multi-vehicle crash that they survived. This produces an estimate of 
roughly 87 percent. Another approach assumes that drivers fully value 
all damage in single vehicle crashes, and only discount property damage 
incidents in multi-vehicle crashes. Based on data in Blincoe, et al. 
(2015),\2091\ multi-vehicle property-damage-only crashes account for 
about 7 percent of all societal crash costs, leaving 93 percent 
recognized under this approach. Yet another approach would assume 
drivers value injury crashes, but discount non-injury related costs 
such as property damage and traffic congestion. This approach results 
in roughly an 88 percent estimate of costs internalized. Overall, while 
the agencies recognize this proportion is uncertain, the agencies 
believe it is reasonable to assume that drivers internalize 90 percent 
of the crash risk that results from added driving.
---------------------------------------------------------------------------

    \2091\ Blincoe, L., Miller, T.R., Zaloshnja, E., Lawrence, B.A., 
(May 2015, Revised) The Economic and Societal Impact of Motor 
Vehicle Crashes, 2010, (DOT HS 812 012), National Highway Traffic 
Safety Administration, Washington, DC.
---------------------------------------------------------------------------

    IPI commented that additional mileage attributable to the scrappage 
and dynamic fleet model is ``inexplicably and unjustifiably not offset 
by countervailing mobility benefits in the benefit cost analysis--and 
the agencies inappropriately claim that these traffic fatalities--which 
comprise the other half of the 12,700 projection--also justify the roll 
back.'' \2092\ In this comment, IPI has erroneously conflated the 
rebound effect and the scrappage effect. The agencies have 
appropriately accounted for the additional value consumers get out of 
increases in fuel efficiency, which manifest in two ways: Reductions in 
fuel costs, and the additional driving resulting from the reductions in 
per-mile fuel costs. The agency cannot appropriately consider one 
without the other, as the two effects trade off, one against the other, 
according to consumer preferences between the two.
---------------------------------------------------------------------------

    \2092\ IPI, Appendix, NHTSA-2018-0067-12213, at 12 (internal 
citation omitted).
---------------------------------------------------------------------------

    The scrappage effect represents the behavior of consumers when 
their choices are restricted by more stringent fuel economy standards. 
For instance, the consumer loses lower-price and less fuel-efficient 
bundles of vehicle attributes that would be available in the absence of 
more stringent alternatives. If anything, these consumers experience an 
un-estimated cost regarding the lost utility from being priced out of 
the new car market and being forced to drive an older, less safe--and 
likely less fuel efficient--vehicle. That the agencies have assessed 
the benefits of the rebound effect by assuming they are at least as 
great as 90 percent of the additional safety costs of rebound driving, 
does not mean that other channels of safety effects must be offset. 
However, the agencies did evaluate whether the sales, scrappage, and 
dynamic fleet share model could lead to changes in fuel economy in the 
legacy fleet that may result in significant changes in VMT and/or fuel 
economy. Upon further review, the agencies determined that such an 
effect--if it were to exist--would be very small and would not impact 
the analysis meaningfully, so the agencies declined to include this 
effect in the final rule's analysis.
d) Fatalities by Source
    For the NPRM, the agencies calculated rebound fatalities by running 
the model with a 20 percent rebound assumption and again with a 0 
percent rebound assumption. The following difference was assumed to 
assign the change in fatalities of the rule due to rebound:

Rebound Fatalities = (FatalitiesAlt,20 - 
FatalitiesAlt,0) - 
(FatalitiesAug,20 - 
FatalitiesAug,0)

    Similarly, the agencies calculated mass reduction fatalities by 
running the model using the central assumptions about coefficients on 
delta curb weight and again setting these coefficients to 0, so that a 
change in mass reduction would not affect the fatality rate of a 
vehicle. The following difference assigned the change in fatalities of 
the rule due to changes in mass reduction levels:

[Delta]CW Fatalities = (FatalitiesAlt,MR - FatalitiesAlt,NoMR) - 
(FatalitiesAug,MR) - (FatalitiesAug,NoMR)

Where ``Alt'' represents the alternative being estimated, ``Aug'' is 
the augural or baseline, ``MR'' stands for mass reduction, and 
``NOMR'' means no mass reduction or mass reduction equaling zero.

    The NPRM modeling then assumed that the remaining incremental 
fatalities were due to changes in sales, scrappage, and the dynamic 
fleet share. This can be represented by the following:

Sales/Scrap Fatalities = (FatalitiesAlt - FatalitiesAug) - Rebound 
Fatalities - [Delta]CW Fatalities

    The changes to the VMT model (mainly the constraint that fixes 
total non-rebound VMT to be constant across alternatives) necessitated 
revising how fatalities are partitioned by source. The number of 
vehicles of each regulatory class and age changes in each regulatory 
alternative. Because of this, taking the increment of the rebound 
fatalities solved in each scenario as described above would capture 
changes both to the usage per vehicle from rebound, but also 
differences in the number of vehicles. This would wrongly attribute 
some of the sales and scrappage fatalities to rebound. Similarly, 
taking the increment of the mass reduction fatalities solved in each 
scenario as described above would capture the changes both to the 
fatality rate for vehicles (from mass reduction) and the difference in 
the number of vehicles across alternatives. This would likewise have 
the potential of wrongly attributing the source of sales and scrappage 
fatalities to mass reduction.

[[Page 24827]]

    Instead of computing the fatalities due to rebound in each scenario 
and then taking the incremental values across alternatives, the 
agencies compute rebound fatalities by taking the difference in per 
vehicle rebound miles in the regulatory alternative and the augural 
case multiplied by the augural fatality rate per mile and augural 
vehicle count. Holding the number of vehicles constant addresses the 
concern about the NPRM fatality allocation method wrongly attributing 
rebound fatalities to the sales and scrappage models. Fatalities due to 
rebound are computed as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.478

Where ``RVMT'' is VMT including rebound miles, ``NRVMT'' is VMT 
excluding rebound miles, ``Veh'' is the quantity of vehicles, and 
``Alt'' and ``Aug'' have the same meaning described above. The 
rebound fatalities will show as zero for the augural scenario, and 
all alternatives will show fatalities due to rebound miles using the 
augural vehicle counts.

    The fatalities due to mass reduction will use the augural vehicle 
counts, augural per vehicle VMT including rebound--this simplifies to 
total VMT including rebound, as shown below. Using a constant vehicle 
count addresses the concern of the NPRM method wrongly assigning some 
mass reduction fatalities to the sales and scrappage models. As with 
the fatalities attributable to rebound, the fatalities attributable to 
changes in mass reduction are calculated inherently as incremental 
values, relative to the augural standards (the values will appear as 
zero for augural standards in the outputs). The equation used to 
calculate the fatalities due to curb weight changes is as follows:

[Delta]CW FatalitiesAlt = (Fatality RateAlt - Fatality RateAug) * R 
VMTAug
    The agencies then computed the sales/scrappage fatalities as the 
remainder, as was done in the NPRM.

Sales/Scrap Fatalities = (FatalitiesAlt-FatalitiesAug)-Rebound 
Fatalities-[Delta]CW Fatalities
(e) Adjustment for Non-Fatal Crashes
    Fatalities are valued as a societal cost within the CAFE models' 
cost and benefit accounting. Their value is based on the comprehensive 
value of a fatality, which includes lost quality of life and is 
quantified in the value of a statistical life (VSL) as well as economic 
consequences such as medical and emergency care, insurance 
administrative costs, legal costs, and other economic impacts not 
captured in the VSL alone. These values were derived from data in 
Blincoe et al. (2015), adjusted to 2018 economics, and updated to 
reflect the official DOT guidance on the value of a statistical life. 
This gives a societal value of $10.4 million for each fatality, which 
is an update to the value used in the NPRM.\2093\ The CAFE safety model 
estimates traffic fatalities but does not directly estimate the 
corresponding non-fatal injuries and property damage that would result 
from the same factors that influence fatalities. To address this, the 
agencies developed an adjustment factor applied to fatality costs that 
accounts for these crashes and related costs. The agencies' approach to 
estimating non-fatal costs remains relatively unchanged from the 
proposal, however the agencies have made one minor adjustment to 
account for advance crash technologies as advocated by commenters.
---------------------------------------------------------------------------

    \2093\ The NPRM used a societal value of $9,900,000 in 2016 
dollars.
---------------------------------------------------------------------------

    In the proposal, development of this factor was premised on the 
assumption that non-fatal crashes would be affected by the standards in 
proportion to their current nationwide rate of incidence and severity. 
The agencies assumed the injury profile--the relative number of crashes 
of each injury severity level that occur nationwide--would increase or 
decrease congruent with changes in fatalities, meaning that the ratio 
between fatal and non-fatal costs remained constant across 
alternatives. The agencies recognized that this may not be the case, 
but did not have data to support individual injury estimates across 
injury severities. The agencies provided several explanations as to why 
a proportionality assumption may be an oversimplification.\2094\ For 
example, the agencies reviewed NHTSA's separate analysis of traffic 
crash data showing that older model year vehicles are generally less 
safe than newer vehicles, meaning fatalities would comprise a larger 
portion of the total injury picture for older vehicles. This would 
imply lower ratios across the non-fatal injury and property damage only 
(PDO) crash profiles and would imply the adjustment overstates total 
societal impacts.
---------------------------------------------------------------------------

    \2094\ See 83 FR 43146 (Aug. 24, 2018).
---------------------------------------------------------------------------

    As noted previously, in response to requests by commenters, the 
agencies have added the estimated impact of six advanced crash 
avoidance technologies that are currently being deployed commercially 
to their analysis of future fatality rates. The same data and methods 
described previously in this section to compute the impact of advanced 
crash avoidance technologies on fatalities can also be used to examine 
the effectiveness of these technologies against non-fatal and PDO 
crashes. The inputs and results are summarized for nonfatal injuries in 
Table VI-258 through Table VI-260, and for PDOs in Table VI-261 through 
Table VI-263.\2095\
---------------------------------------------------------------------------

    \2095\ See previous discussion in this section for the studies 
and methodology used to create these estimates.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24828]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.479

[GRAPHIC] [TIFF OMITTED] TR30AP20.480


[[Page 24829]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.481

[GRAPHIC] [TIFF OMITTED] TR30AP20.482


[[Page 24830]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.483

[GRAPHIC] [TIFF OMITTED] TR30AP20.484


[[Page 24831]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.485

[GRAPHIC] [TIFF OMITTED] TR30AP20.486


[[Page 24832]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.487

BILLING CODE 4910-59-C
    Based on a comparison of the combined average effectiveness impacts 
for the three crash severity groups (fatalities, non-fatal injuries, 
and property damage), it is apparent that these advanced crash 
avoidance technologies would reduce non-fatal injuries and property 
damage crashes by even more than they would fatalities.\2096\ To 
explore the scope of this impact, the agencies developed an adjustment 
factor that reflects the ratio of the decline in the rate of non-fatal 
crashes to that of fatal crashes. This factor would hypothetically 
affect the portion of safety improvement that is attributable to safety 
technologies. The adjustments were based on the cumulative fatality 
rates (for all three technology groups) by model year, noted in Table 
VI-251 (Phased Impact of Crashworthiness Technologies on Fatality 
Rates, Forward Collision Crashes) for fatalities, Table VI-260 for non-
fatal injuries, and Table VI-263 for PDOs, which are listed by MY in 
the last column of Table VI-260 and Table VI-263. These factors would 
modify the original non-fatal impacts--which were derived using an 
assumption that they were proportional to fatal impacts--to reflect the 
higher effectiveness of these technologies against non-fatal crashes.
---------------------------------------------------------------------------

    \2096\ For example, for MY 2035, the combined effectiveness for 
PDO crashes is .224784, as shown in the second to last column of 
Table VI-6, which is 2.613 times the .0860 combined effectiveness 
for fatalities, as seen in the final table from the Crash Avoidance 
discussion above, which shows the disproportional impact of crash 
avoidance technologies on non-fatal accidents.
---------------------------------------------------------------------------

    The agencies considered including this additional adjustment factor 
to account for the additional cost savings attributable to advance 
crash avoidance technologies. The impact of such a factor would 
decrease the incidence and severity, and thus the costs of nonfatal 
crashes in regulatory alternatives where new vehicle sales increase, 
including the preferred alternative. The agencies ultimately erred on 
the side of caution for this rulemaking and have excluded this factor. 
Therefore, today's analysis assumes that advance crash avoidance 
technologies impact non-fatal and PDO crashes to the same extent as 
fatal crashes. The agencies will consider including an adjustment for 
non-fatal and PDO crashes in future rulemakings.

[[Page 24833]]

    The original proportionality-based adjustment factor, which is 
described in detail in the following paragraphs, was derived from 
Tables 1-8 and I-3 in Blincoe et al. (2015). Incidence in Table I-3 in 
Blincoe et al. reflects the Abbreviated Injury Scale (AIS), which ranks 
nonfatal injury severity based on an ascending 5 level scale with the 
most severe injuries ranked as level 5.\2097\
---------------------------------------------------------------------------

    \2097\ More information on the basis for these classifications 
is available from the Association for the Advancement of Automotive 
Medicine at https://www.aaam.org/abbreviated-injury-scale-ais/.
---------------------------------------------------------------------------

    Table 1-3 in Blincoe et al. lists injured persons with their 
highest (maximum) injury determining the AIS level. This scale is 
represented in terms of maximum abbreviated injury scale (MAIS) level. 
MAIS0 refers to uninjured occupants in injury vehicles, MAIS1 injuries 
are generally considered minor (e.g., a superficial laceration) with no 
probability of death, MAIS2 injuries are generally considered moderate 
(e.g., a fractured sternum) with a 1-2 percent probability of death, 
MAIS3 injuries are serious (e.g., open fracture of the humerus) with an 
8-10 percent probability of death, MAIS4 injuries are severe (e.g., 
perforated trachea) with a 5-50 percent probability of death, and MAIS5 
injuries are critical (e.g., rupture liver with tissue loss) with a 5-
50 percent probability of death. Counts for PDO's refer to vehicles in 
which no one was injured. From Table VI-264, ratios of injury 
incidence/fatality are derived for each injury severity level as 
follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.488

[GRAPHIC] [TIFF OMITTED] TR30AP20.489

    For each fatality that occurs nationwide in traffic crashes, there 
are 561 vehicles involved in PDOs, 139 uninjured occupants in crashes 
which resulted in at least one injury,\2098\ 105 minor injuries, 10 
moderate injuries, 3 serious injuries, and fractional numbers of the 
most serious categories which include severe and critical nonfatal 
injuries. For each fatality ascribed to the standards, it is assumed 
there will be non-fatal crashes in these same ratios.
---------------------------------------------------------------------------

    \2098\ Uninjured passengers incur a cost despite being 
uninjured. For example, they are often transported to emergency care 
even tough uninjured resulting in lost time and productivity; 
furthermore, their vehicle might be damaged even though they are 
uninjured.
---------------------------------------------------------------------------

    Property damage costs associated with delayed fleet turnover must 
be treated differently than rebound- and mass-related costs because 
crashes that involve vehicles that are retained longer due to the 
standards involve damage to older, used vehicles instead of newer 
vehicles.\2099\ Used vehicles are worth less and will cost less to 
repair, if they are repaired at all. The consumer's property damage 
loss is thus reduced by longer retention of these vehicles. To estimate 
this loss, average new and used vehicle prices were compared. New 
vehicle transaction prices were estimated from a study published by 
Kelley Blue Book.\2100\ Based on this data, the average new vehicle 
transaction price in January 2017 was $34,968. Used vehicle transaction 
prices were obtained from Edmonds Used Vehicle Market Report published 
in February of 2017.\2101\ Edmonds data indicate the average used 
vehicle transaction price was $19,189 in 2016. There is a minor timing 
discrepancy in these data because the new vehicle data represent 
January 2017, and the used vehicle price is for the average over 2016. 
The agencies were unable to locate exact matching data, but believe the 
difference is minor and negligible.
---------------------------------------------------------------------------

    \2099\ The agencies note that property damage costs are the 
costs realized given an accident has occurred. The disparity of 
incidence rates between new and older vehicles is accounted for 
above in the fatality calculations.
    \2100\ Press Release, ``New-Car Transaction Prices Remain High, 
Up More Than 3 Percent Year-Over-Year in January 2017, According to 
Kelley Blue Book,'' February 1, 2017, available at https://mediaroom.kbb.com/2017-02-01-New-Car-Transaction-Prices-Remain-High-Up-More-Than-3-Percent-Year-Over-Year-In-January-2017-According-To-Kelley-Blue-Book.
    \2101\ Edmonds Used Vehicle Market Report, February 2017. 
Available at https://dealers.edmunds.com/static/assets/articles/2017_Feb_Used_Market_Report.pdf.
---------------------------------------------------------------------------

    Based on these data, new vehicles are on average worth 82 percent 
more than used vehicles. To estimate the effect of higher property 
damage costs for newer vehicles in crashes, the per unit property 
damage costs from Table I-9 in Blincoe et al. (2015) were multiplied by 
this factor.\2102\ Results are illustrated in Table VI-265.
---------------------------------------------------------------------------

    \2102\ The original unit costs were derived from vehicles 
involved in crashes, which are predominately used vehicles. While 
not precise, we assume this average cost is a reasonable proxy for 
the property damage to a used vehicle.

---------------------------------------------------------------------------

[[Page 24834]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.490

    The total property damage cost reduction was then calculated as a 
function of the number of increased fatalities due to stricter CAFE and 
CO2 standards as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.491

Where:

 S = total property damage reductions from retaining used 
vehicles longer
 F = increase in fatalities estimated due to used vehicles 
being retained longer because of stricter standards
 r = ratio of non-fatal injuries or PDO vehicles to 
fatalities
 p = value of property damage prevented by retaining older 
vehicle
 n = the 8 injury severity categories

    The number of fatalities ascribed to the standards because of 
slower fleet turnover was multiplied by the unit cost per fatality from 
Table I-9 in Blincoe et al. (2015) to determine the societal impact of 
fatalities.\2103\ After subtracting the total reductions in property 
damage from this value, the agencies divided the fatality cost by it to 
estimate that overall, fatalities account for 39 percent of the total 
costs that would result from older vehicle retention.
---------------------------------------------------------------------------

    \2103\ Note--These calculations used the original values in the 
Blincoe et al. (2015) tables without adjusting for economics. These 
calculations produce ratios and are thus not sensitive to 
adjustments for inflation.
---------------------------------------------------------------------------

    These calculations are summarized as follows:

SV = Fv/x-s

Where:

 SV = Value of societal impacts of all crashes resulting 
from changes to fleet turnover
 F = Increase in fatalities estimated due to retaining used 
vehicles longer because of stricter standards
 v = Comprehensive societal value of preventing 1 fatality
 x = Percent of total societal loss from crashes 
attributable to fatalities
 S = total property damage reductions from retaining used 
vehicles longer

    For the fatalities that occur because of mass effects or to the 
rebound effect, the calculation was more direct, a simple application 
of the ratio of the portion of costs produced by fatalities to the 
change in fatalities; there is no need to adjust for property damage 
because all impacts were derived from the mix of vehicles in the on-
road fleet. Again, from Table I-8 in Blincoe et al. (2015), the 
agencies derived this ratio based on all cost factors including 
property damage to be 36 percent.
    For purposes of application in the CAFE model, these two factors 
(the factor for sales/scrappage, and the factor for mass and rebound) 
were combined based on the relative contribution to total fatalities of 
different factors. As noted previously, although a safety impact from 
the rebound effect is calculated, these impacts are considered to be 
freely chosen rather than imposed by the standards and imply personal 
benefits at least equal to the sum of their added operational costs and 
the portion of safety consequences internalized. However, the agencies 
still calculate and report the impacts of the rebound effect to provide 
a comprehensive view of the impacts of the standards. There are two 
different factors depending on which metric is considered (total 
impacts or CAFE imposed impacts). The agencies created these two 
adjustment factors by weighting components by the relative contribution 
to changes in fatalities associated with each component. This process 
and results are shown in Table VI-266. Note that due to programming 
constraints, the agencies applied the average weighted factor to all 
fatalities. This will tend to overstate costs slightly because of sales 
and scrappage and to understate costs associated with mass and rebound.

[[Page 24835]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.492

f) Summary of Safety Impacts
    Table VI-267 through Table VI-270 summarize the safety effects of 
CAFE standards across the various alternatives under the 3 percent and 
7 percent discount rates.
    Table VI-271 through Table VI-274 summarize these impacts for 
CO2 standards. As noted in Section VI.D.2.e), societal 
impacts are valued using a $10.4 million value per statistical life 
(VSL). Note that fatalities in these tables are undiscounted--only the 
monetized societal impact is discounted.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.493

[GRAPHIC] [TIFF OMITTED] TR30AP20.494


[[Page 24836]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.495

[GRAPHIC] [TIFF OMITTED] TR30AP20.496


[[Page 24837]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.497


[[Page 24838]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.498


[[Page 24839]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.499


[[Page 24840]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.500


[[Page 24841]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.501


[[Page 24842]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.502


[[Page 24843]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.503

BILLING CODE 4910-59-C
    These tables present aggregations or averages of results for 
calendar years through 2050. Underlying model output files provide 
results for each model year in each calendar year.\2104\ These results 
can be used for more detailed review and analysis of estimated trends. 
For example, for each calendar year through 2050, the following two 
tables--one for CAFE standards and one for CO2 standards--
show (a) the number of light-duty vehicles in service, (b) the travel 
accumulated by those vehicles,

[[Page 24844]]

and (c) the total number fatalities among the types included in today's 
analysis.
---------------------------------------------------------------------------

    \2104\ FOOTNOTE 2104???
---------------------------------------------------------------------------

    The analysis shows the annual number of fatalities for the final 
standards growing more slowly than under the baseline standards, 
reflecting the combined effects of fleet turnover, mass reduction, and 
shifts between passenger cars and light trucks in the new vehicle 
fleet.
    Table VI-274 summarizes the non-fatal safety impacts under 
alternative CAFE and CO2 standards:
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.504


[[Page 24845]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.505

BILLING CODE 4910-59-C
    The Pennsylvania Department of Environmental Protection commented 
that the agencies did not fully account for safety improvements 
associated with the augural standards.\2105\ The agencies note that the 
analysis accounts for the safety impacts of mass reduction, sales and 
scrappage, rebound, vehicle model year and vehicle age for each of the 
alternatives relative to the augural baseline. The commenter did not 
provide any specific items that were omitted from the analysis. The 
agencies believe the analysis thoroughly assesses the safety effects of 
all the alternatives.
---------------------------------------------------------------------------

    \2105\ NOT ON MANUSCRIPT.
---------------------------------------------------------------------------

Simulating Environmental Impacts of Regulatory Alternatives
    This final rulemaking predominantly addresses fuel economy of the 
light-duty vehicle fleet in the United States through different 
technologies to improve efficiency. Inherently, these technologies will 
reduce the fuel consumed and therefore impact CO2 and other 
greenhouse gases foremost. Certain technologies will also impact air 
quality through changes to criteria pollutants and air toxics emitted 
at the tailpipe as well as upstream of the fuel source. Upstream 
emissions for conventional fuels occur during crude oil extraction, 
transportation, refining, and the transportation, storage, and 
distribution of the finished fuel. For electricity, upstream emissions 
are dependent on the mix of feedstocks such as coal, natural gas, 
nuclear, and renewable sources for power generation. Similarly, 
specific hydrogen production pathways such as natural gas reforming or 
electrolysis of water molecules will determine the upstream emissions 
of hydrogen fuel. Emission impacts are described in greater detail in 
the following sections.\2106\
---------------------------------------------------------------------------

    \2106\ NHTSA also uses the results of the CAFE model to analyze 
the potential environmental impacts of the regulatory alternatives 
in its Environmental Impact Statement (EIS). That EIS informs the 
agency's decision-making process.
---------------------------------------------------------------------------

    The impacts of both greenhouse gases (GHGs) and criteria pollutant 
emissions that result from changes in vehicle usage and fuel 
consumption were estimated and considered as part of this analysis. 
GHGs are gaseous constituents in the atmosphere, both natural and 
anthropogenic, and absorb infrared radiation. Primary GHGs in the 
atmosphere are water vapor, CO2, nitrous oxide 
(N2O), methane (CH4), and ozone. Criteria air 
pollutants include carbon monoxide (CO), nitrogen dioxide 
(NO2) (one of several oxides of nitrogen), ozone, sulfur 
dioxides (SO2), particulate matter (including fine 
particulate matter, or PM2.5), and lead. Vehicles do not 
directly emit ozone, but ozone impacts are evaluated based on emissions 
of the ozone precursor pollutants nitrogen oxides (NOX) and 
volatile organic compounds (usually referred to as VOC). These 
pollutants are emitted during vehicle storage and use, as well as 
throughout the fuel production and distribution system. While increases 
in domestic fuel refining, storage, and distribution that result from 
higher fuel consumption will increase emissions of these pollutants, 
reduced vehicle use associated with the fuel economy rebound effect 
will decrease their emissions. The net effect of CAFE and 
CO2 standards on total emissions of each criteria pollutant 
depends on the relative magnitudes of increases in its emissions during 
fuel refining and distribution, and decreases in its emissions 
resulting from vehicle use. Because the relationship between emissions 
in fuel refining and vehicle use is different for each criteria 
pollutant, the net effect of fuel consumption on total emissions of 
each pollutant differs between regulatory alternatives.
Climate Change and CO2 Emissions Considered in This Rule
    The NPRM described how both agencies consider climate change and 
GHG emissions under their respective programs for fuel economy and 
CO2. As noted in the NPRM, ``In 1988, NHTSA included climate 
change concepts in its CAFE notices and prepared its first 
environmental assessment addressing that subject.'' \2107\ 
Additionally, NHTSA ``cited concerns about climate change as one of its 
reasons for limiting the extent of its reduction of the CAFE standard 
for MY 1989 passenger cars.'' \2108\ As stated in the NPRM, ``Since 
then, NHTSA has considered the effects of reducing tailpipe emissions 
of CO2 in its fuel economy rulemakings pursuant to the need 
of the United States to conserve energy by reducing petroleum 
consumption.\2109\
---------------------------------------------------------------------------

    \2107\ 83 FR 43211 (citing 53 FR 33080, 33096 (Aug. 29, 1988)).
    \2108\ Id. (citing 53 FR 39275, 39302 (Oct. 6, 1988)).
    \2109\ 83 FR 43211.
---------------------------------------------------------------------------

    Similarly, in the NPRM, EPA described that ``the primary purpose of 
Title II of the Clean Air Act is the protection of public health and 
welfare. EPA's light-duty vehicle GHG standards serve this purpose, as 
the GHG emissions from light-duty vehicles have been found by EPA to 
endanger public health and welfare (see EPA's 2009 Endangerment Finding 
for on-highway motor vehicles), and the goal of these standards is to 
reduce these emissions that contribute to climate change.'' \2110\ In 
the NPRM, EPA summarized its purpose for establishing CO2 
standards as follows:
---------------------------------------------------------------------------

    \2110\ 83 FR 4228 (citing 74 FR 66496 (Dec. 15, 2009)).

    Section 202(a)(1) of the Clean Air Act (CAA) states that ``the 
Administrator shall by regulation prescribe (and from time to time 
revise) . . . standards applicable to the emission of any air 
pollutant from any class or classes of new motor vehicles . . . , 
which in his judgment cause, or contribute to, air

[[Page 24846]]

pollution which may reasonably be anticipated to endanger public 
health or welfare.'' If EPA makes the appropriate endangerment and 
cause or contribute findings, then section 202(a) authorizes EPA to 
issue standards applicable to emissions of those pollutants. Indeed, 
EPA's obligation to do so is mandatory: Coalition for Responsible 
Regulation, 684 F.3d at 114; Massachusetts v. EPA, 549 U.S. at 
533.\2111\
---------------------------------------------------------------------------

    \2111\ 83 FR 43228.

    The agencies modeled the estimated physical changes in quantity of 
CO2, CH4, and NO2 emissions in the 
NPRM analysis, and conducted additional modeling of climate-related 
impacts, including sea-level rise, global temperate increases, and 
ocean pH changes in the Draft EIS accompanying the NPRM. The Draft EIS 
also included a comprehensive discussion of climate change impacts, 
drawing from various Intergovernmental Panel on Climate Change (IPCC) 
reports, the U.S. Global Change Research Program (USGCRP) National 
Climate Assessment (NCA) reports, and other peer-reviewed reports and 
assessment reports. The agencies also considered the increase in 
climate damages from an increase in CO2 emissions,\2112\ 
also known as the social cost of carbon and discussed previously in 
Section VI.D.1, above.
---------------------------------------------------------------------------

    \2112\ 83 FR 43106.
---------------------------------------------------------------------------

    Many commenters expressed a desire for more information on the 
rule's potential climate impacts, so the discussion has been expanded 
here and in the Final EIS. Specifically, commenters stated that the 
agencies failed to address climate change in the proposal, and that the 
proposal ignored ``scores of studies and reports'' on climate change 
published since EPA's 2009 Endangerment Finding and promulgation of the 
existing CO2 and CAFE standards.\2113\ Several commenters 
presented summaries of climate impacts, citing IPCC, USGCRP, and other 
reports explicitly relied on in the DEIS, on temperature increases, 
increases in extreme weather events, ocean warming, acidification, and 
sea level rise, impacts on the United States' water supply, human 
health impacts, impacts to crop productivity and global food security, 
potential increases in the spread of infectious disease, national 
security impacts, and impacts to animal and plant species, including 
Federally protected species, among other impacts.\2114\
---------------------------------------------------------------------------

    \2113\ NHTSA-2018-0067-12088.
    \2114\ NHTSA-2018-0067-11735; NHTSA-2018-0067-11926; NHTSA-2018-
0067-11972; NHTSA-2018-0067-12088; NHTSA-2018-0067-12127; NHTSA-
2018-0067-12303; NHTSA-2018-0067-12378; NHTSA-2018-0067-12436.
---------------------------------------------------------------------------

    In addition to comments stating the agencies had presented too 
little information on climate change in the NPRM, some commenters 
disagreed with how the agencies framed the impact of the rule on 
climate change. Many commenters cited IPCC and USGCRP to reinforce 
their understanding that human activities are the dominant cause of 
global warming since the mid-20th century. NHTSA considered both the 
IPCC and USGCRP reports in the DEIS accompanying the NPRM and in this 
final rule, and did not dispute those findings. Commenters also cited 
IPCC and the National Climate Assessments, among other reports, as 
support to their understanding that regardless of the perceived 
magnitude of the rule on total CO2 emissions, any additional 
actions taken now to reduce CO2 emissions would affect the 
degree of climate impacts in the future. Further discussion of these 
comments occurs in Section VIII.
    Just as NHTSA does with both the draft and final EIS, and as EPA 
did for its Endangerment and Cause or Contribute Findings for 
Greenhouse Gases under the Clean Air Act, for this rule, both agencies 
relied on existing studies and reports to summarize the current state 
of climate science and provide a framework for the analysis of impacts. 
The agencies drew primarily on panel-reviewed synthesis and assessment 
reports from the Intergovernmental Panel on Climate Change (IPCC) and 
the U.S. Global Change Research Program (GCRP), supplemented with past 
reports from the U.S. Climate Change Science Program (CCSP), the 
National Research Council, and the Arctic Council and EPA's Technical 
Support Document for Endangerment and Cause or Contribute Findings for 
Greenhouse Gases under the Clean Air Act,\2115\ which, as stated above, 
relied on past major international or national scientific assessment 
reports.
---------------------------------------------------------------------------

    \2115\ EPA Technical Support Document for Endangerment and Cause 
or Contribute Findings for Greenhouse Gases under Section 202(a) of 
the Clean Air Act. December 7, 2009. U.S. Environmental Protection 
Agency, Office of Atmospheric Programs, Climate Change Division: 
Washington, DC. Available at: https://www.epa.gov/sites/production/files/2016-08/documents/endangerment_tsd.pdf.
---------------------------------------------------------------------------

    Assessment reports assess numerous individual studies to draw 
general conclusions about the potential impacts of climate change. Even 
where assessment reports include consensus conclusions of expert 
authors, uncertainty still exists, as with all assessments of 
environmental impacts. Given the global nature of climate change and 
the need to communicate uncertainty to a variety of decision-makers, 
IPCC has focused considerable attention on developing a systematic 
approach to characterize and communicate this information. The IPCC is 
a United Nations panel, founded in 1988, which evaluates climate 
science by assessing research on climate change and synthesizing 
relevant research into major assessment reports. The IPCC provides 
regular assessments on climate impacts and future risks, and options 
for adaptation and risk mitigation. The agencies used the system 
developed by IPCC to describe uncertainty associated with various 
climate change impacts.
    The IPCC reports communicate uncertainty and confidence bounds 
using commonly understood but carefully defined words in italics to 
represent likelihood of occurrence. The referenced IPCC documents 
provide a full understanding of the meaning of those uncertainty terms 
in the context of the IPCC findings. The IPCC notes that there are two 
primary uncertainties with climate modeling: Model uncertainties and 
scenario uncertainties: \2116\
---------------------------------------------------------------------------

    \2116\ IPCC. Climate Change 2013: The Physical Science Basis. 
Contribution of Working Group I to the Fifth Assessment Report of 
the Intergovernmental Panel on Climate Change. Stocker, T.F., D. 
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, 
Y. Xia, V. Bex and P.M. Midgley (Eds.). Cambridge University Press: 
Cambridge, United Kingdom and New York, NY, USA. pp. 1535. Available 
at: http://www.ipcc.ch/report/ar5/wg1/. [hereinafter IPCC 2013].
---------------------------------------------------------------------------

     Model uncertainties. These uncertainties occur when a 
climate model might not accurately represent complex phenomena in the 
climate system. For some processes, the scientific understanding could 
be limited regarding how to use a climate model to ``simulate'' 
processes in the climate system.
     Scenario uncertainties. These uncertainties arise because 
of uncertainty in projecting future GHG emissions, concentrations, and 
forcings (e.g., from solar activity).
    According to IPCC, these types of uncertainties are described by 
using two metrics for communicating the degree of certainty: Confidence 
in the validity of findings, expressed qualitatively, and quantified 
measures of uncertainties, expressed probabilistically.\2117\ The 
confidence levels synthesize the judgments about the validity of the 
findings, determined through evaluation of the evidence and the degree 
of scientific agreement. The qualitative expression of confidence 
ranges are described, in italics, from very low to very high, with 
higher confidence levels assigned to findings that are supported by 
high scientific agreement. The quantitative expression of confidence 
ranges from exceptionally unlikely to

[[Page 24847]]

virtually certain, with higher confidence representing findings 
supported by robust evidence. Table VI-276 shows that the degree of 
confidence increases as evidence becomes more robust and agreement is 
greater.
---------------------------------------------------------------------------

    \2117\ IPCC 2013.
    [GRAPHIC] [TIFF OMITTED] TR30AP20.506
    
    As described in more detail in the Final EIS, the process known as 
the greenhouse effect is responsible for trapping a portion of a 
planet's heat in the planet's atmosphere, rather than allowing all of 
that heat to be radiated into space. GHGs trap heat in the lower 
atmosphere (the atmosphere extending from Earth's surface to 
approximately 4 to 12 miles above the surface), absorb heat energy 
emitted by Earth's surface and lower atmosphere, and reradiate much of 
it back to Earth's surface, thereby causing warming. Human activities, 
particularly fossil-fuel combustion, lead to the presence of increased 
concentrations of GHGs in the atmosphere; this buildup of GHGs is 
changing the Earth's energy balance. IPCC states the warming 
experienced over the past century is due to the combination of natural 
climatic forcers (e.g., natural GHGs, solar activity) and human-made 
climate forcers.\2118\ IPCC concluded, ``[h]uman influence has been 
detected in warming of the atmosphere and the ocean, in changes in the 
global water cycle, in reductions in snow and ice, in global mean sea-
level rise, and in changes in some climate extremes. . . . This 
evidence for human influence has grown since [the IPCC Working Group 1 
(WG1) Fourth Assessment Report (AR4)]. IPCC reports that it is 
extremely likely that human influence has been the dominant cause of 
the observed warming since the mid-20th century.'' \2119\
---------------------------------------------------------------------------

    \2118\ IPCC 2013.
    \2119\ IPCC 2013.
---------------------------------------------------------------------------

    Although the climate system is complex, IPCC has identified the 
following drivers of climate change:
     GHGs. Primary GHGs in the atmosphere are water vapor, 
atmospheric CO2, N2O (nitrous oxide), 
CH4 (methane), and ozone.\2120\
---------------------------------------------------------------------------

    \2120\ IPCC 2013.
---------------------------------------------------------------------------

     Aerosols. Aerosols are natural (e.g., from volcanoes) and 
human-made particles in the atmosphere that scatter incoming sunlight 
back to space, causing cooling. Some aerosols are hygroscopic (i.e., 
attract water) and can affect the formation and lifetime of clouds. 
Large aerosols (more than 2.5 micrometers in size) modify the amount of 
outgoing long-wave radiation.\2121\ Other particles, such as black 
carbon, can absorb outgoing terrestrial radiation, causing warming. 
Natural aerosols have had a negligible cumulative impact on climate 
change since the start of the industrial era.\2122\ Further discussion 
of black carbon and other aerosols is located in Chapter 4 of the FEIS.
---------------------------------------------------------------------------

    \2121\ IPCC 2013.
    \2122\ IPCC 2013.
---------------------------------------------------------------------------

     Clouds. Depending on cloud height, cloud interactions with 
terrestrial and solar radiation can vary. Small changes in the 
properties of clouds can have important implications for both the 
transfer of radiative energy and weather.\2123\
---------------------------------------------------------------------------

    \2123\ IPCC 2013.
---------------------------------------------------------------------------

     Ozone. Ozone is created through photochemical reactions 
from natural

[[Page 24848]]

and human-made gases. In the troposphere, ozone absorbs and reemits 
long-wave radiation. In the stratosphere, the ozone layer absorbs 
incoming short-wave radiation.\2124\
---------------------------------------------------------------------------

    \2124\ IPCC 2013.
---------------------------------------------------------------------------

     Solar radiation. Solar radiation, the amount of solar 
energy that reaches the top of Earth's atmosphere, varies over time. 
Solar radiation has had a negligible impact on climate change since the 
start of the industrial era compared to other main drivers.\2125\
---------------------------------------------------------------------------

    \2125\ IPCC 2013.
---------------------------------------------------------------------------

     Surface changes. Changes in vegetation or land surface 
properties, ice or snow cover, and ocean color can affect surface 
albedo.\2126\ The changes are driven by natural seasonal and diurnal 
changes (e.g., snow cover) as well as human influences (e.g., changes 
in vegetation type).\2127\
---------------------------------------------------------------------------

    \2126\ Surfaces on Earth (including land, oceans, and clouds) 
reflect solar radiation back to space. This reflective 
characteristic, known as albedo, indicates the proportion of 
incoming solar radiation the surface reflects. High albedo has a 
cooling effect because the surface reflects rather than absorbs most 
solar radiation.
    \2127\ IPCC 2013.
---------------------------------------------------------------------------

    Effects of emissions and the corresponding processes that affect 
climate are highly complex and variable, which complicates the 
measurement and detection of change. However, IPCC indicates that an 
increasing number of studies conclude that anthropogenic GHG emissions 
are affecting climate in detectable and quantifiable 
ways.2128 2129 GHGs occur naturally and because of human 
activity. Other GHGs, such as the fluorinated gases,\2130\ are 
primarily anthropogenic in origin and are used in commercial 
applications such as refrigeration and air conditioning and industrial 
processes such as aluminum production.
---------------------------------------------------------------------------

    \2128\ IPCC. Summary for Policymakers. In: Change 2013: The 
Physical Science Basis. Contribution of Working Group I to the Fifth 
Assessment Report of the Intergovernmental Panel on Climate Change. 
Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. 
Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (Eds.). 
Cambridge University Press: Cambridge, United Kingdom and New York, 
NY, USA. 1535 pp. Available at: http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_SPM_FINAL.pdf.
    \2129\ GCRP. 2017. Climate Science Special Report: Fourth 
National Climate Assessment. U.S. Global Change Research Program. 
[Wuebbles, D.J., D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. 
Stewart, and T.K. Maycock (Eds.)]. U.S. Government Printing Office: 
Washington, DC 477 pp. doi:10.7930/J0J964J6. Available at: https://science2017.globalchange.gov/downloads/CSSR2017_FullReport.pdf. 
[hereinafter GCRP 2017].
    \2130\ Fluorinated GHGs or gases include PFCs, HFCs, 
SF6, and NF3.
---------------------------------------------------------------------------

    In its most recent assessment of climate change (IPCC WG1 AR5), 
IPCC states that, ``Warming of the climate system is unequivocal, and 
since the 1950s, many of the observed changes are unprecedented over 
decades to millennia. The atmosphere and ocean have warmed, the amounts 
of snow and ice have diminished, sea level has risen, and the 
concentrations of greenhouse gases have increased.'' \2131\ IPCC 
concludes that, at continental and global scales, numerous long-term 
changes in climate have been observed. To be more specific, IPCC and 
the GCRP include the following trends observed over the 20th century as 
further supporting the evidence of climate-induced changes:
---------------------------------------------------------------------------

    \2131\ IPCC 2013.
---------------------------------------------------------------------------

     Most land areas have very likely experienced warmer and/or 
fewer cold days and nights along with warmer and/or more frequent hot 
days and nights.2132 2133 From 1880 to 2016, the global mean 
surface temperature rose by about 0.9 [deg]C (1.6 [deg]F).\2134\ Air 
temperatures are warming more rapidly over land than over 
oceans.2135 2136 Similar to the global trend, the U.S. 
average temperature is about 1.8 [deg]F warmer than it was in 1895, and 
this rate of warming is increasing--most of the warming has occurred 
since 1970.\2137\ IPCC projects a continuing increase in surface 
temperature between 2081 and 2100, with a likely range between 0.3 
[deg]C (0.5 [deg]F) and 4.8 [deg]C (8.6 [deg]F), compared with 1986 
through 2005, where the lower value corresponds to substantial future 
mitigation of carbon emissions.\2138\
---------------------------------------------------------------------------

    \2132\ IPCC Climate Change 2014: Impacts, Adaptation, and 
Vulnerability. Part A: Global and Sectoral Aspects. Contribution of 
Working Group II to the Fifth Assessment Report of the 
Intergovernmental Panel on Climate Change. Field, C.B., V.R. Barros, 
D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, 
K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. 
Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (Eds.). 
Cambridge University Press: Cambridge, United Kingdom and New York, 
NY, USA, 1132 pp. Available at: http://ipcc-wg2.gov/AR5/report/. 
[hereinafter IPCC 2014].
    \2133\ GCRP 2017.
    \2134\ GCRP 2017.
    \2135\ IPCC 2013.
    \2136\ GCRP 2017.
    \2137\ GCRP 2017.
    \2138\ IPCC 2013.
---------------------------------------------------------------------------

     Cold-dependent habitats are shifting to higher altitudes 
and latitudes, and growing seasons are becoming 
longer.2139 2140 According to the IPCC, ``it is virtually 
certain that there will be more frequent hot and fewer cold temperature 
extremes over most land areas on daily and seasonal timescales'' and it 
is very likely that heat wave frequency and duration will also 
increase.\2141\
---------------------------------------------------------------------------

    \2139\ IPCC 2014.
    \2140\ GCRP 2017.
    \2141\ IPCC 2014.
---------------------------------------------------------------------------

     Sea level is rising, caused by thermal expansion of the 
ocean and melting of snowcaps and ice sheets.2142 2143 
Between 1971 and 2010, global ocean temperature warmed by approximately 
0.25 [deg]C (0.45 [deg]F) in the top 200 meters (approximately 660 
feet).\2144\ IPCC concludes that mountain glaciers, ice caps, and snow 
cover have declined on average, further contributing to sea-level rise. 
Losses from the Greenland and Antarctic ice sheets very likely 
contributed to sea-level rise from 1993 to 2010, and satellite 
observations confirm that they have contributed to sea-level rise in 
subsequent years.\2145\ IPCC projects that the global temperature 
increase will continue to affect sea level, causing a likely rise of 
0.26 meter (0.85 foot) to 0.82 meter (2.7 feet) in the next 
century.\2146\
---------------------------------------------------------------------------

    \2142\ IPCC 2013.
    \2143\ GCRP 2017.
    \2144\ IPCC 2013.
    \2145\ IPCC 2013.
    \2146\ IPCC 2013.
---------------------------------------------------------------------------

     More frequent weather extremes such as droughts, floods, 
severe storms, and heat waves have been observed.2147 2148 
Average atmospheric water vapor content has increased since at least 
the 1970s over land and the oceans, and in the upper troposphere, 
largely consistent with air temperature increases.\2149\ Because of 
changes in climate, including increased moisture content in the 
atmosphere, heavy precipitation events have increased in frequency over 
most land areas.2150 2151 Observations of increased dryness 
since the 1950s suggest that some regions of the world have experienced 
longer, more intense droughts caused by higher temperatures and 
decreased precipitation, particularly in the tropics and 
subtropics.\2152\ Heavy precipitation events have increased globally 
since 1951, with some regional and subregional variability.\2153\ A 
warmer atmosphere holds more moisture and increases the energy 
available for convection, causing stronger storms and heavier 
precipitation.2154 2155
---------------------------------------------------------------------------

    \2147\ IPCC 2013.
    \2148\ GCRP 2017.
    \2149\ IPCC 2013.
    \2150\ IPCC 2013.
    \2151\ Min, S.-K., Zhang, X., Zwiers, F.W., & Hegerl, G.C. 2011. 
Human contribution to more-intense precipitation extremes. Nature, 
470(7334), pp. 378-81. Available at: https://doi.org/10.1038/nature09763.
    \2152\ IPCC 2013.
    \2153\ IPCC 2013.
    \2154\ GCRP 2017.
    \2155\ Gertlet, C., O'Gorman, P. 2019. Changing available energy 
for extratropical cyclones and associated convection in the Northern 
Hemisphere summer, PNAS 116(10):4105-4110.

---------------------------------------------------------------------------

[[Page 24849]]

    Many commenters urged the agencies to consider more stringent 
standards to address GHG emissions. The Northeast States for 
Coordinated Air Use Management (NESCAUM) stated that ``effectively 
combatting climate change requires GHG reductions on a national and 
international scale. Maintaining an aggressive downward trend in 
transportation sector GHG emissions will not occur in the absence of 
strong national GHG emission reductions.'' \2168\ Similarly, the Center 
for Biological Diversity et al. stated ``the scientific record is now 
overwhelming that climate change poses grave harm to public health and 
welfare; that its hazards have become even more severe and urgent than 
previously understood; and that avoiding devastating harm requires 
substantial reductions in greenhouse gas emissions, including from the 
critically important transport sector, within the next decade.'' \2169\ 
Minnesota Pollution Control Agency (MPCA), the Minnesota Department of 
Transportation (MnDOT), and the Minnesota Department of Health (MDH) 
stated ``Tackling climate change will require aggressive and immediate 
action on reducing emissions from the transportation sector. The 
existing GHG and CAFE standards are a critical piece to the 
multifaceted and global effort to reduce GHG emissions.'' \2170\
---------------------------------------------------------------------------

    \2168\ NESCAUM, NHTSA-2018-0067-11691.
    \2169\ Center for Biological Diversity et al., NHTSA-2018-0067-
12000.
    \2170\ MPCA, MnDOT, and MDH, NHTSA-2018-0067-11706.
---------------------------------------------------------------------------

    Commenters also expressed concerns that the agencies did not 
accurately consider the effects of climate change resulting from the 
rulemaking. Pennsylvania Department of Environmental Protection (PA 
DEP) stated ``the Proposed Rule does not fully consider the potential 
effects of global climate change resulting from these forgone 
reductions or the interests of states in preventing or mitigating the 
impacts of climate change on their citizens and environment.'' \2171\ 
The Center for Biological Diversity et al. stated ``the agencies 
callously disregard the demonstrated need to reduce emissions sharply 
over the next decade if severe impacts of a destabilized climate are to 
be avoided.'' \2172\ Similarly, the Joint Submission from the States of 
California et al. and the Cities of Oakland et al. stated ``discussion 
of the effect of the Proposed Rollback on GHG emissions significantly 
understates the outcome,'' and ``the overwhelming scientific consensus 
is that immediate and continual progress toward a near-zero GHG-
emission economy by mid-century is necessary to avoid truly 
catastrophic climate change impacts.'' \2173\
---------------------------------------------------------------------------

    \2171\ PA DEP, NHTSA-2018-0067-11956.
    \2172\ Center for Biological Diversity et al., NHTSA-2018-0067-
12000.
    \2173\ Joint Submission from the States of California et al. and 
the Cities of Oakland et al., NHTSA-2018-0067-11735.
---------------------------------------------------------------------------

    The agencies have carefully considered these comments in the 
context of the information on climate change summarized in the NPRM and 
DEIS, and have updated information for this final rule. The agencies 
drew upon updates to climate science and impacts for the analysis from 
reports and studies that were updated or released since the NPRM, 
including IPCC's Global Warming of 1.5 degrees C report, Volume 2 of 
the 4th National Climate Assessment, and IPCC's Special Report on 
Climate Change and Land, and the IPCC's Special Report on the Ocean and 
Cryosphere in a Changing Climate.
    The following sections also provide additional context about 
climate impacts from this final rule; the results of the agencies' 
quantitative analysis presented in Section VII shows estimated 
CO2, CH4, and N2O emissions resulting 
from the rule, and the discussion of how each agency balanced climate 
change as a factor considered in decision-making is presented in 
Section VIII. The Final EIS accompanying today's rule also includes a 
comprehensive discussion of climate impacts, and additional climate 
modeling that estimates climate-related effects. As discussed in more 
detail in the FEIS and following sections, but relevant for placing the 
following discussion in context, climate modeling performed for this 
final rule shows the following impacts as a result of the final 
standards selected: CO2 Concentrations of 789.80 ppm in 
2100, compared with 789.11 ppm under the augural standards; global mean 
surface temperature increases of 3.487 [deg]C in 2100, compared with 
3.484 [deg]C under the augural standards; sea-level rise increases of 
76.34 cm in 2100, compared with 76.28 cm under the augural standards; 
and ocean pH of 8.2172 in 2100, compared with 8.2176 under the augural 
standards. These equal differences of 0.69 ppm, 0.003 [deg]C, 0.06 cm, 
and -0.0004, respectively. Additionally, the agencies valued 
anticipated climate-related economic effects in accordance with E.O. 
13783, as discussed in Section VI.D.1.
(1) Global Greenhouse Gas Emissions
    According to NOAA and IPCC, Global atmospheric CO2 
concentrations have increased 46.4 percent, from approximately 278 
parts per million (ppm) in 1750 \2174\ to approximately 407 ppm in 
2018.\2175\ According to IPCC and WRI, in 2014, CO2 
emissions \2176\ accounted for 76 percent of global GHG emissions on a 
global warming potential (GWP)-weighted basis,\2177\ followed by 
CH4 (16 percent), N2O (6 percent), and 
fluorinated gases (2 percent).2178 2179 IPCC notes that 
atmospheric concentrations of CH4 and N2O 
increased approximately 150 and 20 percent, respectively, over roughly 
the same period.\2180\
---------------------------------------------------------------------------

    \2174\ IPCC 2013.
    \2175\ NOAA. Globally Averaged Marine Surface Annual Mean 
CO2 Data. Available at: ftp://aftp.cmdl.noaa.gov/products/trends/co2/co2_annmean_gl.txt.
    \2176\ These global GHG estimates do not include contributions 
from land-use change and forestry or international bunker fuels.
    \2177\ Each GHG has a different radiative efficiency (the 
ability to absorb infrared radiation) and atmospheric lifetime. To 
compare their relative contributions, GHG emission quantities are 
converted to carbon dioxide equivalent (CO2e) using the 
100-year time horizon global warming potential (GWP) as reported in 
IPCC's Second Assessment Report (AR2): The Science of Climate Change 
in Sections B.7 Summary of Radiative Forcing and B.8 Global Warming 
Potential.
    \2178\ IPCC. 1996. Second Assessment: Climate Change 1995. 
Inventories. Available at: https://www.ipcc.ch/site/assets/uploads/2018/06/2nd-assessment-en.pdf.
    \2179\ WRI (World Resources Institute). 2018. Climate Analysis 
Indicators Tool (CAIT) 2.0: WRI's Climate Data Explorer. Available 
at: http://cait.wri.org/. [hereinafter WRI 2018].
    \2180\ IPCC 2013.
---------------------------------------------------------------------------

    According to WRI, developed countries, including the United States, 
have been responsible for the majority of historical GHG emissions 
since the mid-1800s and still have some of the highest GHG emissions 
per capita.\2181\ While annual emissions from developed countries have 
been relatively flat over the last few decades, world population 
growth, industrialization, and increases in living standards in 
developing countries are expected to cause global fossil-fuel use and 
resulting GHG emissions to grow substantially. According to IPCC, 
global GHG emissions since 2000 have been increasing nearly three times 
faster than in the 1990s.\2182\ This is further illustrated in Figure 
VI-88 showing carbon dioxide emissions since 1990 by world region: 
\2183\
---------------------------------------------------------------------------

    \2181\ WRI 2018.
    \2182\ IPCC 2013.
    \2183\ EPA's Climate Change Indicators in the United States, 
2016: www.epa.gov/climate-indicators. Data source: WRI, 2015.

---------------------------------------------------------------------------

[[Page 24850]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.507

    GHGs are emitted from a wide variety of sectors, including energy, 
industrial processes, waste, agriculture, and forestry. According to 
WRI, the energy sector is the largest contributor of global GHG 
emissions, accounting for 72 percent of global emissions in 2014; other 
major contributors of GHG emissions are agriculture (10 percent) and 
industrial processes (6 percent).\2184\ Transportation CO2 
emissions--from the combustion of petroleum-based fuels--account for 
roughly 15 percent of total global GHG emissions, and have increased by 
64 percent from 1990 to 2014.2185 2186
---------------------------------------------------------------------------

    \2184\ WRI 2018.
    \2185\ The energy sector is largely composed of emissions from 
fuels consumed in the electric power, transportation, industrial, 
commercial, and residential sectors. The 15 percent value for 
transportation is therefore included in the 72 percent value for 
energy.
    \2186\ WRI 2018.
---------------------------------------------------------------------------

    In general, global GHG emissions continue to increase, although 
annual increases vary according to factors such as weather, energy 
prices, and economics. Comparing observed carbon emissions to projected 
emissions, the current global trajectory is similar to the most fossil 
fuel-intensive emissions scenario (A1Fi) in the IPCC Special Report on 
Emissions Scenarios (2000) and the highest emissions scenario (RCP8.5) 
represented by the more recent Representative Concentration Pathways 
(RCP).2187 2188
---------------------------------------------------------------------------

    \2187\ The Representative Concentration Pathways (RCPs) were 
developed for the IPCC AR5 report. They define specific pathways to 
emission concentrations and radiative forcing in 2100. The RCPs 
established four potential emission concentration futures, a 
business-as-usual pathway (RCP8.5), two stabilization pathways 
(RCP6.0, 4.5), and an aggressive reduction pathway (RCP2.6).
    \2188\ IPCC 2013.
---------------------------------------------------------------------------

(2) U.S. Greenhouse Gas Emissions and the Transportation Sector
    Most GHG emissions in the United States are from the energy sector, 
with the majority of those being CO2 emissions coming from 
the combustion of fossil fuels. Fossil fuel combustion CO2 
emissions alone account for 76 percent of total U.S.GWP-weighted 
emissions, with the remaining 24 percent contributed by other sources 
such as industrial processes and product use, agriculture and forestry, 
and waste.\2189\ CO2 emissions due to combustion of fossil 
fuels are from fuels consumed in the transportation (37 percent of 
fossil fuel combustion CO2 emissions), electric power (35 
percent), industrial (16 percent), residential (6 percent), and 
commercial (5 percent) sectors.\2190\ In 2017, U.S. GHG emissions were 
estimated to be 6,456.7 MMTCO2e,\2191\ or approximately 14 
percent of global GHG emissions.2192 2193
---------------------------------------------------------------------------

    \2189\ EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks: 
1990-2017. EPA 430-R-19-001. U.S. Environmental Protection Agency. 
Washington DC Available at: https://www.epa.gov/sites/production/files/2019-04/documents/us-ghg-inventory-2019-main-text.pdf. 
[hereinafter EPA 2019].
    \2190\ EPA 2019.
    \2191\ Most recent year for which an official EPA estimate is 
available. EPA 2019.
    \2192\ Based on global and U.S. estimates for 2014, the most 
recent year for which a global estimate is available. Excluding 
emissions and sinks from land-use change and forestry and 
international bunker fuels.
    \2193\ WRI 2018.
---------------------------------------------------------------------------

    Similar to the global trend, CO2 is by far the primary 
GHG emitted in the U.S.,

[[Page 24851]]

representing 82 percent of U.S. GHG emissions in 2017 (on a GWP-
weighted basis),\2194\ and accounting for 15 percent of total global 
CO2 emissions.2195 2196 Although CO2 
is the GHG with the largest contribution to warming, methane accounts 
for 10.2 percent of U.S. GHGs on a GWP-weighted basis, followed by 
N2O (5.6 percent) and the fluorinated gases (2.6 
percent).\2197\
---------------------------------------------------------------------------

    \2194\ EPA 2019.
    \2195\ The estimate for global emissions from the World 
Resources Institute is for 2014, the most recent year with available 
data for all GHGs. It excludes emissions and sinks from land use 
change and forestry.
    \2196\ WRI 2018.
    \2197\ EPA 2019.
---------------------------------------------------------------------------

    When U.S. CO2 emissions are apportioned by end use, 
transportation is the single leading source of U.S. emissions from 
fossil fuels, causing over one-third of total CO2 emissions 
from fossil fuels.\2198\ Passenger cars and light trucks account for 59 
percent of total U.S. CO2 emissions from transportation, an 
increase of 14 percent since 1990.\2199\ This increase in emissions is 
attributed to about 50 percent increase in vehicle miles traveled (VMT) 
because of population growth and expansion, economic growth, and low 
fuel prices. Additionally, the rising popularity of sport utility 
vehicles and other light trucks with lower fuel economy than passenger 
cars has contributed to higher emissions.2200 2201 Although 
emissions typically increased over this period, emissions declined from 
2008 to 2009 because of decreased economic activity associated with the 
most recent recession.\2202\
---------------------------------------------------------------------------

    \2198\ Apportioning by end use allocates emissions associated 
with electricity generation to the sectors (residential, commercial, 
industrial, and transportation) where it is used. EPA 2019.
    \2199\ EPA 2019.
    \2200\ EPA 2019.
    \2201\ DOT. 2016. Table 4-23: Average Fuel Efficiency of U.S. 
Light Duty Vehicles. U.S. Department of Transportation, Bureau of 
Transportation Statistics. Available at: https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_statistics/html/table_04_23.html.
    \2202\ EPA 2019.
---------------------------------------------------------------------------

    Today's rule addresses light-duty vehicle fuel economy and 
CO2 emissions from new-model passenger cars and light 
trucks. Several commenters observed that the transportation sector 
accounted for a large, if not the largest, portion of the United States 
greenhouse gas emissions, and that light-duty vehicle emissions 
contributed to a large fraction of that portion.\2203\ Many commenters 
referenced the IPCC Report from 2018 on Global Warming of 1.5 Degrees 
Celsius, which considered transportation sector greenhouse gas 
emissions in describing pathways to limit climate impacts.
---------------------------------------------------------------------------

    \2203\ NHTSA-2018-0067-11284; NHTSA-2018-0067-10966; NHTSA-2018-
0067-11691; NHTSA-2018-0067-11735; NHTSA-2018-0067-11765; NHTSA-
2018-0067-11921; NHTSA-2018-0067-12000; NHTSA-2018-0067-12021; 
NHTSA-2018-0067-12022; NHTSA-2018-0067-12088; NHTSA-2018-0067-12303; 
NHTSA-2018-0067-4159.
---------------------------------------------------------------------------

    Graphically, historical trends in U.S. GHG emissions reported by 
EPA appear as follows.\2204\
---------------------------------------------------------------------------

    \2204\ Historical data from https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks. The asterisk 
indicates that the chart does not include reported emissions changes 
attributable to land use, land use change, and forestry (LULUCF).

---------------------------------------------------------------------------

[[Page 24852]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.508

    Notably, light-duty vehicle CO2 emissions outweigh other 
GHG emissions from light-duty vehicles, and light-duty vehicle 
CO2 emissions have been relatively stable over a nearly 30-
year period during which highway vehicles miles traveled has increased 
by about 50 percent.\2205\ Without fuel economy increases that have 
accumulated since EPCA's passage in 1975, recent light-duty vehicle 
CO2 emissions would have been 50 percent greater than shown 
above.\2206\
---------------------------------------------------------------------------

    \2205\ https://www.fhwa.dot.gov/policyinformation/travel_monitoring/historicvmt.pdf.
    \2206\ DOT reports fuel economy levels of the historical on-road 
fleet at https://www.bts.gov/content/average-fuel-efficiency-us-light-duty-vehicles.
---------------------------------------------------------------------------

    For fuel combustion, EIA's National Energy Modeling System (NEMS), 
which EIA uses to produce its Annual Energy Outlook (AEO) forecasts of 
U.S. energy consumption and supply, provides corresponding estimates of 
CO2 emissions. For the final rule, modeling conducted by the 
agencies using the AEO2019 version of NEMS shows the following levels 
of future CO2 emissions from sectors other than light-duty 
vehicles (which this rule impacts directly) and refineries (which this 
rule is estimated to impact through changes in fuel consumption):

[[Page 24853]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.509

    As this chart indicates, EIA's representation of laws and 
regulations current as of AEO2019 shows aggregate emissions from these 
sectors remaining remarkably stable through 2050, despite projected 
growth in the U.S. population and economy.
    The agencies agree with commenters that the transportation sector, 
and specifically light-duty vehicle emissions, contribute to the 
largest portion of the United States' greenhouse gas emissions.\2207\ 
However, the fuel economy and CO2 of vehicles, regulated in 
this rulemaking, is not the only determining factor for whether the 
light-duty transportation sector would see a rise or decline in 
CO2 emissions. As discussed elsewhere in this rule, the 
standards from the final rule affect only new vehicles, which are 
responsible for approximately 3.5 percent of on-road VMT in any year. 
The agencies recognize that the revised standards result in additional 
CO2 emissions, and these emissions are accounted for in the 
analysis. It is worthwhile to note that the difference between the 
augural standard and the new standard is a small change to a small 
fraction of total VMT, and it is important to consider in context the 
different mechanisms that contribute to transportation sector 
greenhouse gas emissions. These mechanisms are considered in the 2018 
IPCC special report cited by commenters as well; in addition to vehicle 
fuel efficiency, IPCC considers preventing (or reducing) the need for 
transport,\2208\ as ``increasingly efficient fleets of vehicles over 
time . . . does not necessarily limit the driven distance.'' (internal 
citations omitted).\2209\
---------------------------------------------------------------------------

    \2207\ See U.S. Energy Information Administration available at 
https://www.eia.gov/todayinenergy/detail.php?id=29612 and EPA, 
Sources of Greenhouse Gas Emissions available at https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions.
    \2208\ IPCC 2018 at 349 (citing Gota et al., 2018).
    \2209\ IPCC 2018 at 377 (citing Ajanovic and Haas, 2017; Sen et 
al., 2017).
---------------------------------------------------------------------------

b) Air Quality
    This section discusses the health and environmental effects 
associated with exposure to some of the criteria and air toxic 
pollutants impacted by the proposed vehicle standards. The agencies 
note that these impacts are, compared to the impacts on vehicular fuel 
consumption and CO2 emissions, small and mixed. CAFE and 
CO2 standards directly impact vehicular fuel consumption and 
CO2 emissions. Notwithstanding modest indirect impacts, such 
as impacts on vehicle

[[Page 24854]]

sales, retention, and mileage accumulation, one can ``draw a direct 
line'' between CAFE/CO2 standards and resultant changes in 
overall fuel consumption and CO2 emissions, and these follow 
the expected trends.
    Changes in emissions of criteria pollutants due to these rules will 
impact air quality. The Clean Air Act (CAA) is the primary federal 
statute that addresses air quality. Pursuant to its CAA authority, the 
EPA has established National Ambient Air Quality Standards (NAAQS) for 
six criteria pollutants: CO, NO2, ozone, SO2, 
particulate matter (PM), and lead. Vehicles do not directly emit ozone, 
but ozone impacts are evaluated based on emissions of the ozone 
precursor pollutants nitrogen oxides (NOX) and volatile 
organic compounds (VOC). When the measured concentrations of a criteria 
pollutant in a geographic region are less than those permitted by 
NAAQS, EPA designates the region as an attainment area for that 
pollutant; regions where concentrations of criteria pollutants exceed 
Federal standards are called nonattainment areas. Former nonattainment 
areas that are now in compliance with NAAQS are designated as 
attainment areas and are commonly referred to as maintenance areas. 
Each state with a nonattainment area is required to develop and 
implement a State Implementation Plan (SIP) documenting how the region 
will reach attainment levels within periods specified in the CAA. For 
maintenance areas, the SIP must document how the State intends to 
maintain compliance with NAAQS. When EPA changes a NAAQS, each State 
must revise its SIP to address how it plans to attain the new standard. 
In addition to analyzing criteria pollutants, the agencies considered 
hazardous air pollutants emitted from vehicles that are known or 
suspected to cause cancer or other serious health and environmental 
impacts and are referred to as mobile source air toxics, as further 
discussed in this section. Table VI-277 below provides an overview of 
criteria pollutants and mobile source air toxics with a high level 
overview of health effects. See further within this section for details 
on the pollutants and toxics.
BILLING CODE 4910-59-P

[[Page 24855]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.510


[[Page 24856]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.511

BILLING CODE 4910-59-C
    The CAA requires the EPA to review periodically the NAAQS and the 
supporting science, and to revise the standards as appropriate.\2210\ 
Schedules for recently completed and ongoing reviews are summarized 
here. In February 2019, the EPA issued a decision to retain the 
existing primary NAAQS for SO2.\2211\ For the ongoing 
reviews of the NAAQS for PM and ozone, the EPA intends to issue 
proposed decisions in early 2020 and final decisions in late 2020.
---------------------------------------------------------------------------

    \2210\ https://www.epa.gov/criteria-air-pollutants/naaqs-table.
    \2211\ 84 FR 9866 (March 18, 2019).
---------------------------------------------------------------------------

    Nationally, levels of PM2.5, ozone, NO2, 
SO2, CO and air toxics have declined significantly in the 
last 30 years. However, as of January 31, 2020, more than 130 million 
people lived in counties designated nonattainment for one or more of 
the NAAQS, and this figure does not include the people living in areas 
with a risk of exceeding a NAAQS in the future. Many Americans continue 
to be exposed to ambient concentrations of air toxics at levels which 
have the potential to cause adverse health effects. In addition, 
populations who live, work, or attend school near major roads 
experience elevated exposure concentrations to a wide range of air 
pollutants. As discussed in the FEIS, concentrations of many air 
pollutants are elevated near high-traffic roadways. If minority 
populations and low-income populations disproportionately live near 
such roads, then an issue of environmental justice (EJ) may be present. 
Comments were received from multiple entities expressing concern about 
emissions and EJ communities. The agencies considered EJ when 
considering the effects of this rule; EJ considerations and EJ-related 
comments received on the NPRM and DEIS are discussed in Section X and 
the FEIS.
    Total emissions from on-road mobile sources (highway vehicles) have 
declined dramatically since 1970 because of pollution controls on 
vehicles and regulation of the chemical content of fuels, despite 
continuing increases in vehicle miles traveled (VMT). From 1970 to 
2016, emissions from on-road mobile sources declined 89 percent for CO, 
71 percent for NOX, 59 percent for PM2.5, 40 
percent for PM10, 93 percent for SO2, and 90 
percent for VOCs.\2212\ The figure below further shows the highway 
vehicle emissions trends that indicate reduced pollutants regulated 
under NAAQS.
---------------------------------------------------------------------------

    \2212\ See https://www.epa.gov/transportation-air-pollution-and-climate-change/accomplishments-and-success-air-pollution-transportation https://gispub.epa.gov/air/trendsreport/2019/#home.

---------------------------------------------------------------------------

[[Page 24857]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.410

    Many commenters expressed concerns about the increase of emissions 
leading to regions in nonattainment for ozone and particulate matter 
and concerns regarding the inability to meet the NAAQS. The Center for 
Biological Diversity et al., and a number of State and local 
governments and government agencies asserted that State and local 
jurisdictions would be at jeopardy of becoming nonattainment areas 
under the proposed rule.\2213\ CARB and the joint submission from the 
States of California and Cities of Oakland stated that the proposed 
rule would result in ``increases in emissions [which] will undermine 
state implementation plans'' and the proposed rule ``would create an 
additional 1.24 tons per day of NOX emissions in the South 
Coast basin.'' \2214\ The South Coast Air Quality Management District 
(SCAQMD) stated ``[a]s a regional air quality district, we have limited 
authority to control emissions from mobile sources, and rely on the 
Federal government to take action,'' and they expressed concern about 
meeting the NAAQS under the proposed rule because, to meet that 
standard, the Basin would have to ``reduce NOX emissions by 
45% beyond existing requirements.'' \2215\
---------------------------------------------------------------------------

    \2213\ Center for Biological Diversity, et al., NHTSA-2018-0067-
12123.
    \2214\ CARB, NHTSA-2018-0067-11873, Joint Submission from States 
of California and Cities of Oakland, NHTSA-2018-0067-11735.
    \2215\ SCAQMD, NHTSA-2018-0067-11813.
---------------------------------------------------------------------------

    In particular, commenters including PA DEP, the Regional Air 
Pollution Control Agency (RAPCA), and CARB, expressed the importance of 
existing CAFE standards in meeting the NAAQS.\2216\ The Northeast 
States for Coordinated Air Use Management (NESCAUM) also asserted that 
regulation and reduction of GHG was necessary to meet the NAAQS, and 
``[o]ur states recognize the urgent need to reduce GHG emissions across 
all sectors of our economy.'' \2217\ Similarly, the agencies from 
Minnesota stated that ``[t]he existing standards are critical for 
states to attain and maintain the NAAQS because vehicles account for 
about 24% of Minnesota's overall air pollution emissions.'' \2218\ The 
Pima County Department of Environmental quality stated that 
``[f]reezing emission reductions for six years could put this region in 
jeopardy of being designated as non-attainment of the ozone standard 
and impact the health of many of our most vulnerable residents.'' 
\2219\ The Washington State Department of Ecology stated that increases 
in NOX and VOC would increase ozone levels in two areas at 
rise of ozone nonattainment in the Puget Sound and the Tri-Cities.'' 
\2220\ The Pennsylvania Department of Environmental Protection stated 
``[r]emoving currently realized emissions reductions and forgoing 
future achievable emissions reductions may make it more difficult for 
areas to attain and maintain the NAAQS. PADEP relies on emission 
reductions from mobile sources as part of its SIP planning to attain 
and maintain the

[[Page 24858]]

NAAQS.'' \2221\ The North Carolina Department of Environmental Quality 
asserted that based on modeling analysis conducted by NCDEQ, ``we 
believe that the fleet changes predicted by the CAFE modeling would 
lead to emissions increases that would interfere with the ability of 
some ozone maintenance areas to meet transportation conformity budgets 
and maintain compliance with the NAAQS.'' \2222\
---------------------------------------------------------------------------

    \2216\ PA DEP, NHTSA-2018-0067-11956, RAPCA NHTSA-2018-0067-
11620, and CARB NHTSA-2018-0067-11873.
    \2217\ NESAUM, NHTSA-2018-0067-11691.
    \2218\ Minnesota Pollution Control Agency(MPCA), the Minnesota 
Department of Transportation (MnDOT), and the Minnesota Department 
of Health(MDH), NHTSA-2018-0067-11706.
    \2219\ Pima County Department of Environmental Quality, NHTSA-
2018-0067-11876.
    \2220\ Washington State Department of Ecology, NHTSA-2018-0067-
11926.
    \2221\ PA DEP, NHTSA-2018-0067-11956.
    \2222\ North Carolina Department of Environmental Quality, 
NHTSA-2018-0067-12025.
---------------------------------------------------------------------------

    Many State commenters also expressed concern about their ability to 
conform with their State Implementation Plan (SIP) after this rule, as 
the Federal vehicle emissions standards previously set were 
incorporated into the SIPs and a rollback could result in further 
increased emissions.\2223\ CARB stated that its ``2016 SIP calls for 
reducing NOX emissions by approximately 6 tons per day,'' 
and according to CARB, the proposed rule would not allow California to 
achieve its South Coast SIP commitments without dramatic 
countermeasures to reduce emissions elsewhere.\2224\ Similarly, other 
agencies expressed concern about SIP requirements, such as PA DEP, who 
stated that ``[b]y flatlining emissions standards at the MY 2020 level, 
the agencies' Proposed Rule increases vehicle emissions. The Proposed 
Rule would interfere with Pennsylvania's SIP planning requirements.'' 
\2225\
---------------------------------------------------------------------------

    \2223\ CARB NHTSA-2018-0067-11873, SCAQMD NHTSA-2018-0067-11813, 
NESCAUM NHTSA-2018-0067-11691, Joint Submission from Colorado local 
governments NHTSA-2018-0067-11929, PA DEP NHTSA-2018-0067-11956, and 
Joint Submission from the States of California et al. and the Cities 
of Oakland et al. NHTSA-2018-0067-11735.
    \2224\ CARB NHTSA-2018-0067-11873.
    \2225\ PA DEP NHTSA-2018-0067-11956.
---------------------------------------------------------------------------

    The commenters expressed concerns that this final rule will present 
challenges in fulfilling existing SIP requirements and in attaining or 
maintaining the NAAQS, resulting in the need for emission reductions to 
offset increases due to this rule. This final rulemaking predominantly 
addresses fuel economy and CO2 emissions of the light-duty 
vehicle fleet. It does not affect EPA's Tier 3 vehicle and gasoline 
(Tier 3) standards or California's low emission vehicle III (LEV III) 
emission standards. Tier 3 and LEV III regulations are predominantly 
responsible for regulating criteria pollutant emissions (e.g. 
NOX, VOCs, and carbon monoxide) from light-duty vehicles. 
While this final rulemaking will result in increases in the amount of 
gasoline produced, the number of vehicle re-fueling events and 
emissions of certain criteria pollutants and precursors the emissions 
impact will vary from area to area depending on factors such as the 
composition of the local vehicle fleet and the amount of gasoline 
produced in the area. The agencies expect that states will evaluate any 
adverse emissions or air quality impacts that result from the 
finalization of this rule in the context of state implementation plan 
development for relevant NAAQS, such as the relevant ozone and 
PM2.5 NAAQS.
    CARB, the joint submission from the States of California and Cities 
of Oakland, and other commenters also stated that the rulemaking 
``fails to meet the general conformity requirements under the Clean Air 
Act.'' \2226\ Similarly, the Center for Biological Diversity, et al., 
stated ``it is highly unlikely that the Proposal would not violate 
general conformity.'' \2227\ The states and cities expressed that the 
General Conformity rule applies to this action because ``[f]irst, an 
increase in criteria pollutants is reasonably foreseeable as the 
agencies quantified those emissions as part of this rulemaking. Second, 
the agencies can practically control those emissions as they possess 
ultimate regulatory authority over standards that govern vehicle 
operation.'' \2228\ CARB stated ``NHTSA's determination regarding its 
own conformity obligations . . . does not address conformity-related 
obligations EPA may have that flow from the joint rulemaking.'' \2229\ 
SCAQMD similarly stated that ``EPA counts as a federal agency that must 
comply with general conformity requirements. The proposal leaves 
unclear whether EPA also determined its actions comply with the general 
conformity requirements under 40 CFR 93.150 and general conformity SIP 
revisions allowed under 40 CFR 51.851.'' \2230\ SCAQMD concluded that 
EPA must make its own conformity determination, ``and it is not clear 
that EPA can rely on NHTSA's analysis given its dissimilar position in 
having continuing program responsibility over mobile source 
emissions.'' \2231\
---------------------------------------------------------------------------

    \2226\ CARB, NHTSA-2018-0067-11873, Joint Submission from States 
of California and Cities of Oakland, NHTSA-2018-0067-11735.
    \2227\ Center for Biological Diversity, et al., NHTSA-2018-0067-
12123.
    \2228\ Joint Submission from States of California and Cities of 
Oakland, NHTSA-2018-0067-11735.
    \2229\ CARB, NHTSA-2018-0067-11873.
    \2230\ SCAQMD, NHTSA_2018-0067-11813.
    \2231\ SCAQMD, NHTSA_2018-0067-11813.
---------------------------------------------------------------------------

    EPA and NHTSA disagree with the commenters that this rule is 
subject to the CAA section 176(c) conformity requirement and the 
General Conformity regulations. A General Conformity evaluation is 
required for a general Federal action proposed to occur within specific 
nonattainment or maintenance areas. For a General Conformity evaluation 
to be necessary, the action must cause emissions of the criteria and 
precursor pollutants for which the areas are nonattainment or 
maintenance, and the emissions must originate within those areas. 
Further, the evaluation would require a demonstration that the action 
conforms to a specific State Implementation Plan's strategy for air 
pollution prevention and control applicable to the nonattainment and 
maintenance areas. In addition, any mitigation or offsets required to 
demonstrate conformity may require written commitments that must be 
fulfilled, and offsets must occur during the same calendar year as the 
emission increases from the action.
    While the EPA established the framework of methods and procedures 
that Federal agencies must follow when General Conformity applies to 
their actions, it is the responsibility of each Federal agency to 
prepare its own General Conformity evaluation for actions the agency 
supports, funds, permits or approves. When the EPA functions as a lead 
agency for actions that are subject to General Conformity, such as 
water projects, and the agency may issue permits or approve actions 
that require a General Conformity evaluation, EPA is responsible for 
and sometimes is required to prepare its own General Conformity 
evaluation. For the reasons specified here and in Section X.E.2, a 
General Conformity evaluation is not necessary for either agency.
    As stated in section 4.1.1.4 of the DEIS and in section 4.1.1.4 of 
the FEIS, the agencies do not believe the proposed rule would result in 
either direct or indirect emissions as defined for General Conformity 
at 40 CFR 93.152 or as required for applicability of the rule under 
section 93.153(b). Furthermore, as described in the proposal, emissions 
from operation of vehicles produced during the model years covered by 
this rule, while reasonably foreseeable, cannot be quantified with any 
certainty in any particular nonattainment or maintenance area. In 
addition, while the emissions rates from MY 2021-2026 vehicles are 
projected for future years in this rule, neither NHTSA nor EPA has 
control over where, when or how many of the vehicles will operate 
during a given future year or within a certain geographical area. 
Therefore, the emissions are not quantifiable. Furthermore, the General 
Conformity

[[Page 24859]]

applicability analysis requires an analytical comparison of the 
emissions from MY 2021-2026 vehicles in some specific nonattainment or 
maintenance area in a specific future year, to the emissions projected 
from the operation of vehicles produced in other model years that would 
otherwise operate in that same area in the same future year. Without 
the identity of the future year vehicle fleet by type/make/model (which 
depends on a specific nonattainment or maintenance location and year), 
the net emissions, or total of direct and indirect emissions, cannot be 
quantified. Thus, this rule, in and of itself, is not subject to a 
General Conformity evaluation.
    CARB stated that this rulemaking would, if finalized, invalidate 
the model underlying California's SIPs (the EMFAC 2014 model), which 
would result in the SIPs being disapproved by EPA.\2232\ CARB expressed 
further concern that as a result of the Clean Air Act's conformity 
requirements, this disapproval would put significant limits on new 
RTPs, TIPS, or regionally significant transportation projects being 
adopted or approved in California.\2233\
---------------------------------------------------------------------------

    \2232\ CARB, NHTSA-2018-0067-11873.
    \2233\ CARB, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    The commenter expressed the opinion that if this rule is finalized, 
EPA would disapprove its SIPs because its on-road emission factor model 
(EMFAC) would be invalidated. The commenter also opined that such 
disapprovals would limit the ability of metropolitan planning 
organizations in California to make transportation conformity 
determinations for metropolitan transportation plans, transportation 
improvement programs and certain transportation projects. It is 
premature to assume that EPA will disapprove SIPs because they are 
based on EMFAC2014 or EMFAC2017. EPA will evaluate and address, as 
appropriate, the impact of the SAFE action on future SIP approval 
actions EMFAC2014 and EMFAC2017 remain approved emission factor models 
for SIPs and transportation conformity analyses in California. EPA is 
aware that California released adjustment factors to be applied to 
EMFAC2014 and EMFAC2017 model results to account for impacts of the 
SAFE Part 1 rule for on-road criteria pollutant emissions from light-
duty vehicles. EPA will work with CARB and DOT on the appropriate 
implementation of federal requirements based on current and available 
information.
    Because passenger cars and light trucks are subject to gram-per-
mile emissions standards for criteria pollutants, more fuel-efficient 
(and, correspondingly, less CO2-intensive) vehicles are not, 
from the standpoint of air quality, ``cleaner'' vehicles. Therefore, to 
the extent that CAFE/CO2 standards lead to changes in 
overall quantities of vehicular emissions that impact air quality, 
these are dominated by induced changes in highway travel. Changes in 
overall fuel consumption do lead to changes in emissions from 
``upstream'' processes involved in supplying fuel to vehicles. 
Depending on how total vehicular emissions and total upstream emissions 
change in response to less stringent standards, overall emissions could 
increase or decrease. While small in magnitude, net impacts could also 
vary considerably among different geographic areas. In other words, 
CAFE and CO2 standards impact fuel consumption and 
CO2 emissions in ways that are direct and unambiguous, and 
impact air quality in ways that are indirect and ambiguous.
    The following sections, included in prior rules setting fuel 
economy and CO2 standards and updated based on EPA's latest 
scientific assessments, describe the criteria and air toxics considered 
in this rule, and their health and environmental effects. Additionally, 
the section that follows describes how the estimated effects of each 
pollutant were modeled in this rulemaking. Section VII discusses the 
interactions between upstream, tailpipe, and highway travel that result 
in the net emissions of criteria and air toxic pollutants estimated as 
a result of this rule.
(1) Particulate Matter
(a) Background
    Particulate matter (PM) is a complex mixture of solid particles and 
liquid droplets distributed among numerous atmospheric gases which 
interact with solid and liquid phases. Particles range in size from 
those smaller than 1 nanometer (10-\9\ meter) to over 100 
micrometers ([micro]m, or 10-\6\ meter) in diameter (for 
reference, a typical strand of human hair is 50-70 [micro]m in diameter 
and a grain of fine beach sand is about typically 90 [micro]m in 
diameter). Atmospheric particles can be grouped into several classes 
according to their aerodynamic and physical sizes. Generally, the three 
broad classes of particles include ultrafine particles (UFPs, generally 
considered as particulates with a diameter less than or equal to 0.1 
[micro]m [typically based on physical size, thermal diffusivity or 
electrical mobility]), ``fine'' particles (PM2.5; particles 
with a nominal mean aerodynamic diameter less than or equal to 2.5 
[micro]m), and ``thoracic'' particles (PM10; particles with 
a nominal mean aerodynamic diameter less than or equal to 10 [micro]m). 
Particles that fall within the size range between PM2.5 and 
PM10 are referred to as ``thoracic coarse particles'' 
(PM10-2.5 particles with a nominal mean aerodynamic diameter 
greater than 2.5 [micro]m and less than or equal to 10 [micro]m). EPA 
currently has standards that regulate PM2.5 and 
PM10.\2234\
---------------------------------------------------------------------------

    \2234\ Regulatory definitions of PM size fractions and 
information on reference and equivalent methods for measuring PM in 
ambient air are provided in 40 CFR parts 50, 53, and 58. With regard 
to national ambient air quality standards (NAAQS) which provide 
protection against health and welfare effects, the 24-hour 
PM10 standard provides protection against effects 
associated with short-term exposure to thoracic coarse particles 
(i.e. PM10--2.5).
---------------------------------------------------------------------------

    Most particles are found in the lower troposphere, where they can 
have residence times ranging from a few hours to weeks. Particles are 
removed from the atmosphere by wet deposition, such as when they are 
carried by rain or snow, or by dry deposition, when particles settle 
out of suspension due to gravity. Atmospheric lifetimes are generally 
longest for PM2.5, which often remains in the atmosphere for 
days to weeks before being removed by wet or dry deposition. \2235\In 
contrast, atmospheric lifetimes for UFP and PM10-2.5 are 
shorter. Within hours, UFP can undergo coagulation and condensation 
that lead to formation of larger particles in the accumulation mode, or 
can be removed from the atmosphere by evaporation, deposition, or 
reactions with other atmospheric components. PM10-2.5 are 
also generally removed from the atmosphere within hours, through wet or 
dry deposition.\2236\
---------------------------------------------------------------------------

    \2235\ U.S. EPA. Integrated Science Assessment (ISA) for 
Particulate Matter (Final Report. 2019), U.S. Environmental 
Protection Agency, Washington DC, EPA/600/R-19/188, 2019. Table 2-1.
    \2236\ U.S. EPA. Integrated Science Assessment (ISA) for 
Particulate Matter (Final Report, 2019). U.S Environmental 
Protection Agency, Washington, DC, EPA/600/R-19/188, 2019. Table 2-
1.
---------------------------------------------------------------------------

    Particulate matter consists of both primary and secondary 
particles. Primary particles are emitted directly from sources, such as 
combustion-related activities (e.g., industrial activities, motor 
vehicles, biomass burning), while secondary particles are formed 
through atmospheric chemical reactions of gaseous precursors (e.g., 
sulfur oxides (SOx), nitrogen oxides (NOx) and volatile organic 
compounds (VOCs) and ammonia). From 2000 to 2017, national annual 
average PM2.5 concentrations have declined by over 
40%,\2237\ largely reflecting reductions in emissions of precursor 
gases.
---------------------------------------------------------------------------

    \2237\ See https://www.epa.gov/air-trends/particulate-matter-pm25-trends and https://www.epa.gov/air-trends/particulate-matter-pm25-trends#pmnat for more information.

---------------------------------------------------------------------------

[[Page 24860]]

(b) Health Effects of PM
    Scientific evidence spanning animal toxicological, controlled human 
exposure, and epidemiologic studies shows that exposure to ambient PM 
is associated with a broad range of health effects. The Integrated 
Science Assessment for Particulate Matter (PM ISA) (U.S. EPA 2009) 
synthesizes the toxicological, clinical and epidemiological evidence to 
determine whether each pollutant is causally related to an array of 
adverse human health outcomes associated with either acute (i.e., hours 
or days-long) or chronic (i.e. years-long) exposure; for each outcome, 
the ISA reports this relationship to be causal, likely to be causal, 
suggestive of a causal relationship, inadequate to infer a causal 
relationship or not likely to be a causal relationship.
    In brief, the ISA for PM2.5 found acute exposure to 
PM2.5 to be causally related to cardiovascular effects and 
mortality (i.e., premature death), and respiratory effects as likely-
to-be-causally related. The ISA identified cardiovascular effects and 
total mortality as being causally related to long-term exposure to 
PM2.5 and respiratory effects as likely-to-be-causal; and 
the evidence was suggestive of a causal relationship for reproductive 
and developmental effects as well as cancer, mutagenicity and 
genotoxicity. The ISA for ozone found acute exposure to ozone to be 
causally related to respiratory effects, a likely-to-be-causal 
relationship with cardiovascular effects and total mortality and a 
suggestive relationship for central nervous system effects. Among 
chronic effects, the ISA reported a likely-to-be-causal relationship 
for respiratory outcomes and respiratory mortality, and suggestive 
relationship for cardiovascular effects, reproductive and developmental 
effects, central nervous system effects, and total mortality. DOT 
follows EPA's approach of estimating the incidence of air pollution 
effects for those health effects above where the ISA classified as 
either causal or likely-to-be-causal.
    EPA's more recent Integrated Science Assessment for Particulate 
Matter (PM ISA), which was finalized in December 2019,\2238\ summarizes 
the most recent health effects evidence for short- and long-term 
exposures to PM2.5, PM10-2.5, and ultrafine 
particles, characterizing the strength of the evidence and whether the 
relationship is likely to be causal nature in nature. The 2019 P.M. ISA 
reinforces the findings of the 2009 ISA, and supports the decision to 
continue monetizing the respiratory and cardiovascular health endpoints 
monetized in the current analysis. EPA is currently in the process of 
considering how the 2019 ISA and eventual decision by the Administrator 
regarding the National Ambient Air Quality Standards for particulate 
matter will be used to update forthcoming regulatory impact analysis.
---------------------------------------------------------------------------

    \2238\ U.S. EPA. Integrated Science Assessment (ISA) for 
Particulate Matter (Final Report, 2019). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
---------------------------------------------------------------------------

(c) Current Concentrations
    There are two primary NAAQS for PM2.5: an annual 
standard (12.0 micrograms per cubic meter ([mu]g/m\3\)) set in 2012 and 
a 24-hour standard (35 [mu]g/m\3\) set in 2006, and two secondary NAAQS 
for PM2.5: an annual standard (15.0 [mu]g/m\3\) set in 1997 
and a 24-hour standard (35 [mu]g/m\3\) set in 2006.\2239\
---------------------------------------------------------------------------

    \2239\ The EPA is currently reviewing the PM NAAQS and 
anticipates completing this review in late 2020 Available at https://www.epa.gov/naaqs/particulate-matter-pm-air-quality-standards).
---------------------------------------------------------------------------

    There are many areas of the country that are currently in 
nonattainment for the annual and 24-hour primary PM2.5 
NAAQS. As of January 31, 2020, more than 19 million people lived in the 
4 areas that are designated as nonattainment for the 1997 annual 
PM2.5 NAAQS. These PM2.5 nonattainment areas are 
comprised of 14 full or partial counties. As of January 31, 2020, 6 
areas are designated as nonattainment for the 2012 annual 
PM2.5 NAAQS; these areas are composed of 16 full or partial 
counties with a population of more than 20 million. As of January 31, 
2020, 14 areas are designated as nonattainment for the 2006 24-hour 
PM2.5 NAAQS; these areas are composed of 41 full or partial 
counties with a population of more than 31 million. In total, there are 
currently 17 PM2.5 nonattainment areas with a population of 
more than 32 million people.
    The EPA has already adopted many mobile source emission control 
programs that are expected to reduce ambient PM concentrations. As a 
result of these and other federal, state and local programs, the number 
of areas that fail to meet the PM2.5 NAAQS in the future is 
expected to decrease. However, even with the implementation of all 
current state and federal regulations, there are projected to be 
counties violating the PM2.5 NAAQS well into the future.
(2) Ozone
(a) Background
    Ground-level ozone pollution is typically formed through reactions 
involving VOC and NOX in the lower atmosphere in the 
presence of sunlight. These pollutants, often referred to as ozone 
precursors, are emitted by many types of sources, such as highway and 
nonroad motor vehicles and engines, power plants, chemical plants, 
refineries, makers of consumer and commercial products, industrial 
facilities, and smaller area sources.
    The science of ozone formation, transport, and accumulation is 
complex. Ground-level ozone is produced and destroyed in a cyclical set 
of chemical reactions, many of which are sensitive to temperature and 
sunlight. When ambient temperatures and sunlight levels remain high for 
several days and the air is relatively stagnant, ozone and its 
precursors can build up and result in more ozone than typically occurs 
on a single high-temperature day. Ozone and its precursors can be 
transported hundreds of miles downwind from precursor emissions, 
resulting in elevated ozone levels even in areas with low local VOC or 
NOX emissions.
(b) Health Effects of Ozone
    This section provides a summary of the health effects associated 
with exposure to ambient concentrations of ozone.\2240\ The information 
in this section is based on the information and conclusions in the 
February 2013 Integrated Science Assessment for Ozone (Ozone ISA), 
which formed the basis for EPA's revision to the primary and secondary 
standards in 2015.\2241\ The Ozone ISA concludes that human exposures 
to ambient concentrations of ozone are associated with a number of 
adverse health effects and characterizes the weight of evidence for 
these health effects.\2242\ The discussion below

[[Page 24861]]

highlights the Ozone ISA's conclusions pertaining to health effects 
associated with both short-term and long-term periods of exposure to 
ozone.
---------------------------------------------------------------------------

    \2240\ Human exposure to ozone varies over time due to changes 
in ambient ozone concentration and because people move between 
locations which have notable different ozone concentrations. Also, 
the amount of ozone delivered to the lung is not only influenced by 
the ambient concentrations but also by the individuals breathing 
route and rate.
    \2241\ U.S. EPA. Integrated Science Assessment of Ozone and 
Related Photochemical Oxidants (Final Report). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA 
is available at http://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=247492#Download.
    \2242\ The ISA evaluates evidence and draws conclusions on the 
causal nature of relationship between relevant pollutant exposures 
and health effects, assigning one of five ``weight of evidence'' 
determinations: causal relationship, likely to be a causal 
relationship, suggestive of, but not sufficient to infer, a causal 
relationship, inadequate to infer a causal relationship, and not 
likely to be a causal relationship. For more information on these 
levels of evidence, please refer to Table II in the Preamble of the 
ISA.
---------------------------------------------------------------------------

    For short-term exposure to ozone, the Ozone ISA concludes that 
respiratory effects, including lung function decrements, pulmonary 
inflammation, exacerbation of asthma, respiratory-related hospital 
admissions, and mortality, are causally associated with ozone exposure. 
It also concludes that cardiovascular effects, including decreased 
cardiac function and increased vascular disease, and total mortality 
are likely to be causally associated with short-term exposure to ozone 
and that evidence is suggestive of a causal relationship between 
central nervous system effects and short-term exposure to ozone.
    For long-term exposure to ozone, the Ozone ISA concludes that 
respiratory effects, including new onset asthma, pulmonary inflammation 
and injury, are likely to be causally related with ozone exposure. The 
Ozone ISA characterizes the evidence as suggestive of a causal 
relationship for associations between long-term ozone exposure and 
cardiovascular effects, reproductive and developmental effects, central 
nervous system effects and total mortality. The evidence is inadequate 
to infer a causal relationship between chronic ozone exposure and 
increased risk of lung cancer.
    Finally, inter-individual variation in human responses to ozone 
exposure can result in some groups being at increased risk for 
detrimental effects in response to exposure. In addition, some groups 
are at increased risk of exposure due to their activities, such as 
outdoor workers or children. The Ozone ISA identified several groups 
that are at increased risk for ozone-related health effects. These 
groups are people with asthma, children and older adults, individuals 
with reduced intake of certain nutrients (i.e., Vitamins C and E), 
outdoor workers, and individuals having certain genetic variants 
related to oxidative metabolism or inflammation. Ozone exposure during 
childhood can have lasting effects through adulthood. Such effects 
include altered function of the respiratory and immune systems. 
Children absorb higher doses (normalized to lung surface area) of 
ambient ozone, compared to adults, due to their increased time spent 
outdoors, higher ventilation rates relative to body size, and a 
tendency to breathe a greater fraction of air through the mouth. 
Children also have a higher asthma prevalence compared to adults.
(c) Current Concentrations
    The primary and secondary NAAQS for ozone are 8-hour standards with 
a level of 0.07 ppm. The most recent revision to the ozone standards 
was in 2015; the previous 8-hour ozone primary standard, set in 2008, 
had a level of 0.075 ppm.\2243\ As of January 31, 2020, there were 36 
ozone nonattainment areas for the 2008 ozone NAAQS, composed of 153 
full or partial counties, with a population of more than 99 million. As 
of January 31, 2020, there were 51 ozone nonattainment areas for the 
2015 ozone NAAQS, composed of 206 full or partial countries, with a 
population of more than 122 million. In total, there are currently 59 
ozone nonattainment areas with a population of more than 127 million 
people.
---------------------------------------------------------------------------

    \2243\ The EPA is currently reviewing the PM NAAQS and 
anticipates completing this review in late 2020 Available at 
(https://www.epa.gov/naaqs/ozone-o3-air-quality-standards).
---------------------------------------------------------------------------

    States with ozone nonattainment areas are required to take action 
to bring those areas into attainment. The attainment date assigned to 
an ozone nonattainment area is based on the area's classification. The 
attainment dates for areas designated nonattainment for the 2008 8-hour 
ozone NAAQS are in the 2015 to 2032 timeframe, depending on the 
severity of the problem in each area. Nonattainment area attainment 
dates associated with areas designated for the 2015 NAAQS will be in 
the 2021-2038 timeframe, depending on the severity of the problem in 
each area.
    EPA has already adopted many emission control programs that are 
expected to reduce ambient ozone levels. As a result of these and other 
federal, state and local programs, 8-hour ozone levels are expected to 
improve in the future. However, even with the implementation of all 
current state and federal regulations, there are projected to be 
counties violating the ozone NAAQS well into the future.
(3) Nitrogen Oxides
(a) Background
    Oxides of nitrogen (NOX) refers to nitric oxide and 
nitrogen dioxide (NO2). For the NOX NAAQS, 
NO2 is the indicator. Most NO2 is formed in the 
air through the oxidation of nitric oxide (NO) emitted when fuel is 
burned at a high temperature. NOX is also a major 
contributor to secondary PM2.5 formation. NOX and 
VOC are the two major precursors of ozone.
(b) Health Effects of Nitrogen Oxides
    The most recent review of the health effects of oxides of nitrogen 
completed by EPA can be found in the 2016 Integrated Science Assessment 
for Oxides of Nitrogen--Health Criteria (Oxides of Nitrogen ISA).\2244\ 
The primary source of NO2 is motor vehicle emissions, and 
ambient NO2 concentrations tend to be highly correlated with 
other traffic-related pollutants. Thus, a key issue in characterizing 
the causality of NO2-health effect relationships was 
evaluating the extent to which studies supported an effect of 
NO2 that is independent of other traffic-related pollutants. 
EPA concluded that the findings for asthma exacerbation integrated from 
epidemiologic and controlled human exposure studies provided evidence 
that is sufficient to infer a causal relationship between respiratory 
effects and short-term NO2 exposure. The strongest evidence 
supporting an independent effect of NO2 exposure comes from 
controlled human exposure studies demonstrating increased airway 
responsiveness in individuals with asthma following ambient-relevant 
NO2 exposures. The coherence of this evidence with 
epidemiologic findings for asthma hospital admissions and ED visits as 
well as lung function decrements and increased pulmonary inflammation 
in children with asthma describe a plausible pathway by which 
NO2 exposure can cause an asthma exacerbation. The 2016 ISA 
for Oxides of Nitrogen also concluded that there is likely to be a 
causal relationship between long-term NO2 exposure and 
respiratory effects. This conclusion is based on new epidemiologic 
evidence for associations of NO2 with asthma development in 
children combined with biological plausibility from experimental 
studies.
---------------------------------------------------------------------------

    \2244\ U.S. EPA. Integrated Science Assessment for Oxides of 
Nitrogen--Health Criteria (2016 Final Report). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-15/068, 2016.
---------------------------------------------------------------------------

    In evaluating a broader range of health effects, the 2016 ISA for 
Oxides of Nitrogen concluded evidence is ``suggestive of, but not 
sufficient to infer, a causal relationship'' between short-term 
NO2 exposure and cardiovascular effects and mortality and 
between long-term NO2 exposure and cardiovascular effects 
and diabetes, birth outcomes, and cancer. In addition, the scientific 
evidence is inadequate (insufficient consistency of epidemiologic and 
toxicological evidence) to infer a causal relationship for long-term 
NO2 exposure with

[[Page 24862]]

fertility, reproduction, and pregnancy, as well as with postnatal 
development. A key uncertainty in understanding the relationship 
between these non-respiratory health effects and short- or long-term 
exposure to NO2 is copollutant confounding, particularly by 
other roadway pollutants. The available evidence for non-respiratory 
health effects does not adequately address whether NO2 has 
an independent effect or whether it primarily represents effects 
related to other or a mixture of traffic-related pollutants.
    The 2016 ISA for Oxides of Nitrogen concluded that people with 
asthma, children, and older adults are at increased risk for 
NO2-related health effects. In these groups and life stages, 
NO2 is consistently related to larger effects on outcomes 
related to asthma exacerbation, for which there is confidence in the 
relationship with NO2 exposure.
(c) Current Concentrations
    On April 6, 2018, based on a review of the full body of scientific 
evidence, EPA issued a decision to retain the current primary NAAQS for 
NO2. The EPA has concluded that the current NAAQS are 
requisite to protect the public health, including the at-risk 
populations of older adults, children and people with asthma, with an 
adequate margin of safety. The primary NAAQS for NO2 are a 
one-hour standard with a level of 100 ppb, based on the three-year 
average of 98th percentile of the annual distribution of daily maximum 
one-hour concentrations, and an annual standard at a level of 53 ppb.
(4) Sulfur Oxides
(a) Background
    Sulfur dioxide (SO2), a member of the sulfur oxide 
(SOX) family of gases, is formed from burning fuels 
containing sulfur (e.g., coal or oil derived), extracting gasoline from 
oil, or extracting metals from ore. SO2 and its gas phase 
oxidation products can dissolve in water droplets and further oxidize 
to form sulfuric acid which reacts with ammonia to form sulfates, which 
are important components of ambient PM.
(b) Health Effects of SO2
    This section provides an overview of the health effects associated 
with SO2. Additional information on the health effects of 
SO2 can be found in the 2017 Integrated Science Assessment 
for Sulfur Oxides--Health Criteria (SOX ISA).\2245\ 
Following an extensive evaluation of health evidence from animal 
toxicological, controlled human exposure, and epidemiologic studies, 
the EPA has concluded that there is a causal relationship between 
respiratory health effects and short -term exposure to SO2. 
The immediate effect or SO2 on the respiratory system in 
humans is bronchoconstriction. People with asthma are more sensitive to 
the effects of SO2, likely resulting from preexisting 
inflammation associated with this disease. In addition to those with 
asthma (both children and adults), there is suggestive evidence that 
all children and older adults may be at increased risk of 
SO2-related health effects. In free-breathing laboratory 
studies involving controlled human exposures to SO2, 
respiratory effects have consistently been observed following 5-10 min 
exposures at SO2 concentrations >= 400 ppb in people with 
asthma engaged in moderate to heavy levels of exercise, with 
respiratory effects occurring at concentrations as low as 
200 ppb in some individuals with asthma. A clear 
concentration-response relationship has been demonstrated in these 
studies following exposures to SO2 at concentrations between 
200 and 1000 ppb, both in terms of increasing severity of 
respiratory symptoms and decrements in lung function, as well as the 
percentage of individuals with asthma adversely affected. Epidemiologic 
studies have reported positive associations between short-term ambient 
SO2 concentrations and hospital admissions and emergency 
department visits for asthma and for all respiratory causes, 
particularly among children and older adults (>=65 years). The studies 
provide supportive evidence for the causal relationship.
---------------------------------------------------------------------------

    \2245\ U.S. EPA (2017). Integrated Science Assessment (ISA) for 
Sulfur Oxides. Health Criteria (Final Report). EPA 600/R-17/451. 
Washington, DC, U.S. EPA.
---------------------------------------------------------------------------

    For long-term SO2 exposure and respiratory effects, the 
EPA has concluded that the evidence is suggestive or a causal 
relationship. This conclusion is based on new epidemiologic evidence 
for positive associations between long-term SO2 exposure and 
increases in asthma incidence among children, together with animal 
toxicological evidence that provides a pathophysiologic basis for the 
development of asthma. However, uncertainty remains regarding the 
influence of other pollutants on the observed associations with 
SO2 because these epidemiologic studies have not examined 
the potential for copollutant confounding.
    Consistent associations between short-term exposure to 
SO2 and mortality have been observed in epidemiologic 
studies, with larger effect estimates reported for respiratory 
mortality than for cardiovascular mortality. While this finding is 
consistent with the demonstrated effects of SO2 on 
respiratory morbidity, uncertainty remains with respect to the 
interpretation of these observed mortality associations due to 
potential confounding by various copollutants. Therefore, the EPA has 
concluded that the overall evidence is suggestive of a causal 
relationship between short-term exposure to SO2 and 
mortality.
(c) Current Concentrations
    On February 25, 2019, the EPA announced its decision to retain, 
without revision, the existing NAAQS for SOX of 75 ppb, as 
the annual 99th percentile of daily maximum SO2 
concentrations, averaged over three years (84 FR 9866, March 18, 2019). 
The existing primary (health-based) standard provides health protection 
for the at-risk group (people with asthma) against respiratory effects 
following short-term (e.g., 5-minute) exposures to SO2 in 
ambient air. The EPA has been finalizing the initial area designations 
for the 2010 SO2 NAAQS in phases and completed designations 
for most of the country in December 2017. The EPA is under a court 
order to finalize initial designations by December 31, 2020, for a 
remaining set of about 50 areas where states have deployed new 
SO2 monitoring networks. As of January 31, 2020 there are 34 
nonattainment areas for the 2010 SO2 NAAQS. As of January 
31, 2020 there also remain eight nonattainment areas for the primary 
annual SO2 NAAQS set in 1971.
(5) Carbon Monoxide
(a) Background
    Carbon monoxide is a colorless, odorless gas emitted from 
combustion processes. Nationally, particularly in urban areas, the 
majority of CO emissions to ambient air come from mobile sources.\2246\
---------------------------------------------------------------------------

    \2246\ U.S. EPA (2010). Integrated Science Assessment for Carbon 
Monoxide (Final Report). U.S. Environmental Protection Agency, 
Washington, DC, EPA/600/R-09/019F, 2010. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686. See Section 
2.1.
---------------------------------------------------------------------------

(b) Health Effects of Carbon Monoxide
    Information on the health effects of CO can be found in the January 
2010 Integrated Science Assessment for Carbon Monoxide (CO ISA) 
associated with the 2010 evaluation of the

[[Page 24863]]

NAAQS.\2247\ The CO ISA presents conclusions regarding the presence of 
causal relationships between CO exposure and categories of adverse 
health effects. This section provides a summary of the health effects 
associated with exposure to ambient concentrations of CO, along with 
the ISA conclusions.\2248\
---------------------------------------------------------------------------

    \2247\ U.S. EPA (2010). Integrated Science Assessment for Carbon 
Monoxide (Final Report). U.S. Environmental Protection Agency, 
Washington, DC, EPA/600/R-09/019F, 2010. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686.
    \2248\ Personal exposure includes contributions from many 
sources, and in many different environments. Total personal exposure 
to CO includes both ambient and nonambient components; and both 
components may contribute to adverse health effects.
---------------------------------------------------------------------------

    Controlled human exposure studies of subjects with coronary artery 
disease show a decrease in the time to onset of exercise-induced angina 
(chest pain) and electrocardiogram changes following CO exposure. In 
addition, epidemiologic studies observed associations between short-
term CO exposure and cardiovascular morbidity, particularly increased 
emergency room visits and hospital admissions for coronary heart 
disease (including ischemic heart disease, myocardial infarction, and 
angina). Some epidemiologic evidence is also available for increased 
hospital admissions and emergency room visits for congestive heart 
failure and cardiovascular disease as a whole. The CO ISA concludes 
that a causal relationship is likely to exist between short-term 
exposures to CO and cardiovascular morbidity. It also concludes that 
available data are inadequate to conclude that a causal relationship 
exists between long-term exposures to CO and cardiovascular morbidity.
    Animal studies show various neurological effects with in-utero CO 
exposure. Controlled human exposure studies report central nervous 
system and behavioral effects following low-level CO exposures, 
although the findings have not been consistent across all studies. The 
CO ISA concludes the evidence is suggestive of a causal relationship 
with both short-and long-term exposure to CO and central nervous system 
effects.
    A number of studies cited in the CO ISA have evaluated the role of 
CO exposure in birth outcomes such as preterm birth or cardiac birth 
defects. There is limited epidemiologic evidence of a CO-induced effect 
on preterm births and birth defects, with weak evidence for a decrease 
in birth weight. Animal toxicological studies have found perinatal CO 
exposure to affect birth weight, as well as other developmental 
outcomes. The CO ISA concludes the evidence is suggestive of a causal 
relationship between long-term exposures to CO and developmental 
effects and birth outcomes.
    Epidemiologic studies provide evidence of associations between 
short-term CO concentrations and respiratory morbidity such as changes 
in pulmonary function, respiratory symptoms, and hospital admissions. A 
limited number of epidemiologic studies considered copollutants such as 
ozone, SO2, and PM in two-pollutant models and found that CO 
risk estimates were generally robust, although this limited evidence 
makes it difficult to disentangle effects attributed to CO itself from 
those of the larger complex air pollution mixture. Controlled human 
exposure studies have not extensively evaluated the effect of CO on 
respiratory morbidity. Animal studies at levels of 50-100 ppm CO show 
preliminary evidence of altered pulmonary vascular remodeling and 
oxidative injury. The CO ISA concludes that the evidence is suggestive 
of a causal relationship between short-term CO exposure and respiratory 
morbidity, and inadequate to conclude that a causal relationship exists 
between long-term exposure and respiratory morbidity.
    Finally, the CO ISA concludes that the epidemiologic evidence is 
suggestive of a causal relationship between short-term concentrations 
of CO and mortality. Epidemiologic evidence suggests an association 
exists between short-term exposure to CO and mortality, but limited 
evidence is available to evaluate cause-specific mortality outcomes 
associated with CO exposure. In addition, the attenuation of CO risk 
estimates which was often observed in copollutant models contributes to 
the uncertainty as to whether CO is acting alone or as an indicator for 
other combustion-related pollutants. The CO ISA also concludes that 
there is not likely to be a causal relationship between relevant long-
term exposures to CO and mortality.
(c) Current Concentrations
    There are two primary NAAQS for CO: an 8-hour standard (9 ppm) and 
a 1-hour standard (35 ppm). The primary NAAQS for CO were retained in 
August 2011. There are currently no CO nonattainment areas; as of 
September 27, 2010, all CO nonattainment areas have been predesignated 
to attainment.
    The past designations were based on the existing community-wide 
monitoring network. EPA made an addition to the ambient air monitoring 
requirements for CO during the 2011 NAAQS review. Those new 
requirements called for CO monitors to be operated near roads in Core 
Based Statistical Areas (CBSAs) of 1 million or more persons (76 FR 
54294, August 31, 2011).
(6) Diesel Exhaust
(a) Background
    Diesel exhaust consists of a complex mixture composed of 
particulate matter, carbon dioxide, oxygen, nitrogen, water vapor, 
carbon monoxide, nitrogen compounds, sulfur compounds, and numerous 
low-molecular-weight hydrocarbons. A number of these gaseous 
hydrocarbon components are individually known to be toxic, including 
aldehydes, benzene and 1,3-butadiene. The diesel particulate matter 
present in diesel exhaust consists mostly of fine particles (< 2.5 
[micro]m), of which a significant fraction is ultrafine particles (< 
0.1 [micro]m). These particles have a large surface area which makes 
them an excellent medium for adsorbing organics, and their small size 
makes them highly respirable. Many of the organic compounds present in 
the gases and on the particles, such as polycyclic organic matter, are 
individually known to have mutagenic and carcinogenic properties.
    Diesel exhaust varies significantly in chemical composition and 
particle sizes between different engine types (heavy-duty, light-duty), 
engine operating conditions (idle, acceleration, deceleration), and 
fuel formulations (high/low sulfur fuel). Also, there are emissions 
differences between on-road and nonroad engines because the nonroad 
engines are generally of older technology. After being emitted in the 
engine exhaust, diesel exhaust undergoes dilution as well as chemical 
and physical changes in the atmosphere. The lifetime for some of the 
compounds present in diesel exhaust ranges from hours to days.
(b) Health Effects of Diesel Exhaust
    In EPA's 2002 Diesel Health Assessment Document (Diesel HAD), 
exposure to diesel exhaust was classified as likely to be carcinogenic 
to humans by inhalation from environmental exposures, in accordance 
with the revised draft 1996/1999 EPA cancer 
guidelines.2249 2250 A number of

[[Page 24864]]

other agencies (National Institute for Occupational Safety and Health, 
the International Agency for Research on Cancer, the World Health 
Organization, California EPA, and the U.S. Department of Health and 
Human Services) had made similar hazard classifications prior to 2002. 
EPA also concluded in the 2002 Diesel HAD that it was not possible to 
calculate a cancer unit risk for diesel exhaust due to limitations in 
the exposure data for the occupational groups or the absence of a dose-
response relationship.
---------------------------------------------------------------------------

    \2249\ U.S. EPA. (1999). Guidelines for Carcinogen Risk 
Assessment. Review Draft. NCEA-F-0644, July. Washington, DC: U.S. 
EPA. Retrieved on March 19, 2009 from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=54932.
    \2250\ U.S. EPA (2002). Health Assessment Document for Diesel 
Engine Exhaust. EPA/600/8-90/057F Office of Research and 
Development, Washington DC. Retrieved on March 17, 2009 from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060. pp. 1-1 & 1-2.
---------------------------------------------------------------------------

    In the absence of a cancer unit risk, the Diesel HAD sought to 
provide additional insight into the significance of the diesel exhaust 
cancer hazard by estimating possible ranges of risk that might be 
present in the population. An exploratory analysis was used to 
characterize a range of possible lung cancer risk. The outcome was that 
environmental risks of cancer from long-term diesel exhaust exposures 
could plausibly range from as low as 10-5 to as high as 
10-3. Because of uncertainties, the analysis acknowledged 
that the risks could be lower than 10-5, and a zero risk 
from diesel exhaust exposure could not be ruled out.
    Non-cancer health effects of acute and chronic exposure to diesel 
exhaust emissions are also of concern to EPA. EPA derived a diesel 
exhaust reference concentration (RfC) from consideration of four well-
conducted chronic rat inhalation studies showing adverse pulmonary 
effects. The RfC is 5 [micro]g/m\3\ for diesel exhaust measured as 
diesel particulate matter. This RfC does not consider allergenic 
effects such as those associated with asthma or immunologic or the 
potential for cardiac effects. There was emerging evidence in 2002, 
discussed in the Diesel HAD, that exposure to diesel exhaust can 
exacerbate these effects, but the exposure-response data were lacking 
at that time to derive an RfC based on these then-emerging 
considerations. The EPA Diesel HAD stated, ``With [diesel particulate 
matter] being a ubiquitous component of ambient PM, there is an 
uncertainty about the adequacy of the existing [diesel exhaust] 
noncancer database to identify all of the pertinent [diesel exhaust]-
caused noncancer health hazards.'' The Diesel HAD also noted ``that 
acute exposure to [diesel exhaust] has been associated with irritation 
of the eye, nose, and throat, respiratory symptoms (cough and phlegm), 
and neurophysiological symptoms such as headache, lightheadedness, 
nausea, vomiting, and numbness or tingling of the extremities.'' The 
Diesel HAD noted that the cancer and noncancer hazard conclusions 
applied to the general use of diesel engines then on the market and as 
cleaner engines replace a substantial number of existing ones, the 
applicability of the conclusions would need to be reevaluated.
    It is important to note that the Diesel HAD also briefly summarized 
health effects associated with ambient PM and discusses EPA's then-
annual PM2.5 NAAQS of 15 [micro]g/m\3\. In 2012, EPA revised 
the annual PM2.5 NAAQS to 12 [micro]g/m\3\. There is a large 
and extensive body of human data showing a wide spectrum of adverse 
health effects associated with exposure to ambient PM, of which diesel 
exhaust is an important component. The PM2.5 NAAQS is 
designed to provide protection from the noncancer health effects and 
premature mortality attributed to exposure to PM2.5. The 
contribution of diesel PM to total ambient PM varies in different 
regions of the country and also, within a region, from one area to 
another. The contribution can be high in near-roadway environments, for 
example, or in other locations where diesel engine use is concentrated.
    Since 2002, several new studies have been published which continue 
to report increased lung cancer risk with occupational exposure to 
diesel exhaust from older engines. Of particular note since 2011 are 
three new epidemiology studies which have examined lung cancer in 
occupational populations, for example, truck drivers, underground 
nonmetal miners and other diesel motor-related occupations. These 
studies reported increased risk of lung cancer with exposure to diesel 
exhaust with evidence of positive exposure-response relationships to 
varying degrees.2251 2252 2253 These newer studies (along 
with others that have appeared in the scientific literature) add to the 
evidence EPA evaluated in the 2002 Diesel HAD and further reinforces 
the concern that diesel exhaust exposure likely poses a lung cancer 
hazard. The findings from these newer studies do not necessarily apply 
to newer technology diesel engines because the newer engines have large 
reductions in the emission constituents compared to older technology 
diesel engines.
---------------------------------------------------------------------------

    \2251\ Garshick, Eric, Francine Laden, Jaime E. Hart, Mary E. 
Davis, Ellen A. Eisen, and Thomas J. Smith. 2012. Lung cancer and 
elemental carbon exposure in trucking industry workers. 
Environmental Health Perspectives 120(9), 1301-06.
    \2252\ Silverman, D.T., Samanic, C.M., Lubin, J.H., Blair, A.E., 
Stewart, P.A., Vermeulen, R., & Attfield, M.D. (2012). The diesel 
exhaust in miners study: a nested case-control study of lung cancer 
and diesel exhaust. Journal of the National Cancer Institute.
    \2253\ Olsson, Ann C., et al. ``Exposure to diesel motor exhaust 
and lung cancer risk in a pooled analysis from case-control studies 
in Europe and Canada.'' American journal of respiratory and critical 
care medicine 183.7 (2011): 941-48.
---------------------------------------------------------------------------

    In light of the growing body of scientific literature evaluating 
the health effects of exposure to diesel exhaust, in June 2012 the 
World Health Organization's International Agency for Research on Cancer 
(IARC), a recognized international authority on the carcinogenic 
potential of chemicals and other agents, evaluated the full range of 
cancer-related health effects data for diesel engine exhaust. IARC 
concluded that diesel exhaust should be regarded as ``carcinogenic to 
humans.'' \2254\ This designation was an update from its 1988 
evaluation that considered the evidence to be indicative of a 
``probable human carcinogen.''
---------------------------------------------------------------------------

    \2254\ IARC (International Agency for Research on Cancer) 
(2013). Diesel and gasoline engine exhausts and some nitroarenes. 
IARC Monographs Volume 105. Available at http://monographs.iarc.fr/ENG/Monographs/vol105/index.php.
---------------------------------------------------------------------------

(c) Current Concentrations
    Because DPM is part of overall ambient PM and cannot be easily 
distinguished from overall PM, the agencies do not have direct 
measurements of DPM in the ambient air. DPM concentrations are 
estimated using ambient air quality modeling based on DPM emission 
inventories. DPM emission inventories are computed as the exhaust PM 
emissions from mobile sources combusting diesel or residual oil fuel. 
DPM concentrations were recently estimated as part of the 2014 NATA. 
Areas with high concentrations are clustered in the Northeast, Great 
Lake States, California, and the Gulf Coast States and are also 
distributed throughout the rest of the U.S.
(7) Air Toxics
(a) Background
    Light-duty vehicle emissions contribute to ambient levels of air 
toxics that are known or suspected human or animal carcinogens, or that 
have noncancer health effects. The population experiences an elevated 
risk of cancer and other noncancer health effects from exposure to the 
class of pollutants known collectively as ``air toxics.'' \2255\ These 
compounds include, but are not limited to, benzene, 1,3-

[[Page 24865]]

butadiene, formaldehyde, acetaldehyde, acrolein, polycyclic organic 
matter, and naphthalene. These compounds were identified as national or 
regional risk drivers or contributors in the 2014 or past National-
scale Air Toxics Assessment and have significant inventory 
contributions from mobile sources.2256 2257
---------------------------------------------------------------------------

    \2255\ U.S. EPA (2015). Summary of Results for the 2011 
National-Scale Assessment. http://www3.epa.gov/sites/production/files/2015-12/documents/2011-nata-summary-results.pdf.
    \2256\ U.S EPA (2018) Technical Support Document EPA's 2014 
National Air Toxics Assessment. Available at https://www.epa.gov/national-air-toxics-assessment/2014-nata-assessment-results.
    \2257\ U.S. EPA (2015). 2011 National Air Toxics Assessment. 
http://www3.epa.gov/national-air-toxics-assessment/2011-national-air-toxics-assessment.
---------------------------------------------------------------------------

(b) Benzene
    EPA's Integrated Risk Information System (IRIS) database lists 
benzene as a known human carcinogen (causing leukemia) by all routes of 
exposure, and concludes that exposure is associated with additional 
health effects, including genetic changes in both humans and animals 
and increased proliferation of bone marrow cells in 
mice.2258 2259 2260 EPA states in its IRIS database that 
data indicate a causal relationship between benzene exposure and acute 
lymphocytic leukemia and suggest a relationship between benzene 
exposure and chronic non-lymphocytic leukemia and chronic lymphocytic 
leukemia. EPA's IRIS documentation for benzene also lists a range of 
2.2 x 10-6 to 7.8 x10-6 per [micro]g/m\3\ as the unit risk estimate 
(URE) for benzene.2261 2262 The International Agency for 
Research on Cancer (IARC) has determined that benzene is a human 
carcinogen and the U.S. Department of Health and Human Services (DHHS) 
has characterized benzene as a known human 
carcinogen.2263 2264
---------------------------------------------------------------------------

    \2258\ U.S. EPA. (2000). Integrated Risk Information System File 
for Benzene. This material is available electronically at: http://www3.epa.gov/iris/subst/0276.htm.
    \2259\ International Agency for Research on Cancer, IARC 
monographs on the evaluation of carcinogenic risk of chemicals to 
humans, Volume 29, some industrial chemicals and dyestuffs, 
International Agency for Research on Cancer, World Health 
Organization, Lyon, France 1982.
    \2260\ Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry, 
V.A. (1992). Synergistic action of the benzene metabolite 
hydroquinone on myelopoietic stimulating activity of granulocyte/
macrophage colony-stimulating factor in vitro, Proc. Natl. Acad. 
Sci. 89:3691-3695.
    \2261\ A unit risk estimate is defined as the increase in the 
lifetime risk of an individual who is exposed for a lifetime to 1 
[micro]g/m\3\ benzene in air.
    \2262\ U.S. EPA (2000). Integrated Risk Information System File 
for Benzene. This material is available electronically at: http://www3.epa.gov/iris/subst/0276.htm.
    \2263\ International Agency for Research on Cancer (IARC, 2018. 
Monographs on the evaluation of carcinogenic risks to humans, volume 
120. World Health Organization--Lyon France. Available at http://publications.iarc.fr/Book-And-ReportSeries/Iarc-Monographs-On-The-ldentification-Of-Carcinogenic-Hazards-To-Humans/Benzene-2018.
    \2264\ NTP (National Toxicology Program). 2016. Report on 
Carcinogens, Fourteenth Edition.; Research Triangle Park, NC: U.S. 
Department of Health and Human Services Public Health Service. 
Available at https://ntp.niehs.nih.gov/go/roc.
---------------------------------------------------------------------------

    A number of adverse noncancer health effects including blood 
disorders, such as pre- leukemia and aplastic anemia, have also been 
associated with long-term exposure to benzene. The most sensitive 
noncancer effect observed in humans, based on current data, is the 
depression of the absolute lymphocyte count in blood. EPA's inhalation 
reference concentration (RfC) for benzene is 30 [micro]g/m\3\. The RfC 
is based on suppressed absolute lymphocyte counts seen in humans under 
occupational exposure conditions. In addition, recent work, including 
studies sponsored by the Health Effects Institute, provides evidence 
that biochemical responses are occurring at lower levels of benzene 
exposure than previously known.2265 2266 2267 2268 EPA's 
IRIS program has not yet evaluated these new data. EPA does not 
currently have an acute reference concentration for benzene. The Agency 
for Toxic Substances and Disease Registry (ATSDR) Minimal Risk Level 
(MRL) for acute exposure to benzene is 29 [micro]g/m\3\ for 1-14 days 
exposure.
---------------------------------------------------------------------------

    \2265\ Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen, 
B.; Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.; Rupa, D.; 
Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik, M.; Kwok, 
E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003). HEI Report 
115, Validation & Evaluation of Biomarkers in Workers Exposed to 
Benzene in China.
    \2266\ Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et 
al. (2002). Hematological changes among Chinese workers with a broad 
range of benzene exposures. Am. J. Industr. Med. 42: 275-285.
    \2267\ Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al. 
(2004). Hematotoxically in Workers Exposed to Low Levels of Benzene. 
Science 306: 1774-1776.
    \2268\ Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism 
in rodents at doses relevant to human exposure from Urban Air. 
Research Reports Health Effect Inst. Report No.113.
---------------------------------------------------------------------------

(c) 1,3-Butadiene
    EPA has characterized 1,3-butadiene as carcinogenic to humans by 
inhalation.2269 2270 The IARC has determined that 1,3-
butadiene is a human carcinogen and the U.S. DHHS has characterized 
1,3-butadiene as a known human 
carcinogen.2271 2272 2273 2274 There are numerous studies 
consistently demonstrating that 1,3-butadiene is metabolized into 
genotoxic metabolites by experimental animals and humans. The specific 
mechanisms of 1,3-butadiene-induced carcinogenesis are unknown; 
however, the scientific evidence strongly suggests that the 
carcinogenic effects are mediated by genotoxic metabolites. Animal data 
suggest that females may be more sensitive than males for cancer 
effects associated with 1,3-butadiene exposure; there are insufficient 
data in humans from which to draw conclusions about sensitive 
subpopulations. The URE for 1,3-butadiene is 3 x 10-5 per 
[micro]g/m\3\.\2275\ 1,3-butadiene also causes a variety of 
reproductive and developmental effects in mice; no human data on these 
effects are available. The most sensitive effect was ovarian atrophy 
observed in a lifetime bioassay of female mice.\2276\ Based on this 
critical effect and the benchmark concentration methodology, an RfC for 
chronic health effects was calculated at 0.9 ppb (approximately 2 
[micro]g/m\3\).
---------------------------------------------------------------------------

    \2269\ U.S. EPA (2002). Health Assessment of 1,3-Butadiene. 
Office of Research and Development, National Center for 
Environmental Assessment, Washington Office, Washington, DC. Report 
No. EPA600-P-98-001F. This document is available electronically at 
http://www3.epa.gov/iris/supdocs/buta-sup.pdf.
    \2270\ U.S. EPA (2002). ``Full IRIS Summary for 1,3-butadiene 
(CASRN 106-99-0)'' Environmental Protection Agency, Integrated Risk 
Information System (IRIS), Research and Development, National Center 
for Environmental Assessment, Washington, DC. Available at http://www3.epa.gov/iris/subst/0139.htm.
    \2271\ International Agency for Research on Cancer (IARC) 
(1999). Monographs on the evaluation of carcinogenic risk of 
chemicals to humans, Volume 71, Re-evaluation of some organic 
chemicals, hydrazine and hydrogen peroxide World Health 
Organization, Lyon, France.
    \2272\ International Agency for Research on Cancer (IARC). 
(2012). Monographs on the evaluation of carcinogenic risk of 
chemicals to humans, Volume 100F chemical agents and related 
occupations, World Health Organization, Lyon, France.
    \2273\ International Agency for Research on Cancer (IARC). 
(2008). Monographs on the evaluation of carcinogenic risk of 
chemicals to humans, 1,3-Butadiene, Ethylene Oxide and Vinyl Halides 
(Vinyl Fluoride, Vinyl Chloride and Vinyl Bromide) Volume 97, World 
Health Organization, Lyon, France.
    \2274\ NTP (National Toxicology Program). 201 6. Report on 
Carcinogens, Fourteenth Edition.; Research Triangle Park NC: U.S. 
Department of Health and Human Services Public Health Service. 
Available at https://ntp.niehs.nih.gov/go/rocl4.
    \2275\ U.S. EPA (2002). ``Full IRIS Summary for 1,3-butadiene 
(CASRN 106-99-0)'' Environmental Protection Agency, Integrated Risk 
Information System (IRIS), Research and Development, National Center 
for Environmental Assessment, Washington, DC http://www3.epa.gov/iris/subst/0139.htm.
    \2276\ Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996). 
Subchronic toxicity of 4-vinylcyclohexene in rats and mice by 
inhalation. Fundam. Appl. Toxicol. 32:1-10.
---------------------------------------------------------------------------

(d) Formaldehyde
    In 1991, EPA concluded that formaldehyde is a carcinogen based on 
nasal tumors in animal bioassays.\2277\ An Inhalation URE for cancer 
and a Reference Dose for oral noncancer

[[Page 24866]]

effects were developed by the agency and posted on the IRIS database. 
Since that time, the National Toxicology Program (NTP) and 
International Agency for Research on Cancer (IARC) have concluded that 
formaldehyde is a known human carcinogen.2278 2279 2280
---------------------------------------------------------------------------

    \2277\ EPA Integrated Risk Information System. Formaldehyde 
(CASRN 50-00-0) http://www3.epa.gov/iris/subst/0419/htm.
    \2278\ NTP (National Toxicology Program). 2016. Report on 
Carcinogens. Fourteenth Edition.; Research Triangle Park, NC: U.S. 
Department of Health and Human Services. Public Health Service. 
Available at https://ntp.niehs.nih.gov/go/roc 14.
    \2279\ IARC Monographs on the Evaluation of Carcinogenic Risks 
to Humans Volume 100F (2012): Formaldehyde.
    \2280\ IARC Monographs on the Evaluation of Carcinogenic Risks 
to Humans Volume 88 (2006): Formaldehyde, 2- Butoxyethanol and 1 -
tert-Butoxypropan-2-ol.
---------------------------------------------------------------------------

    The conclusions by IARC and NTP reflect the results of 
epidemiologic research published since 1991 in combination with 
previous animal, human and mechanistic evidence. Research conducted by 
the National Cancer Institute reported an increased risk of 
nasopharyngeal cancer and specific lymph hematopoietic malignancies 
among workers exposed to formaldehyde.2281 2282 2283 A 
National Institute of Occupational Safety and Health study of garment 
workers also reported increased risk of death due to leukemia among 
workers exposed to formaldehyde.\2284\ Extended follow-up of a cohort 
of British chemical workers did not report evidence of an increase in 
nasopharyngeal or lymph hematopoietic cancers, but a continuing 
statistically significant excess in lung cancers was reported.\2285\ 
Finally, a study of embalmers reported formaldehyde exposures to be 
associated with an increased risk of myeloid leukemia but not brain 
cancer.\2286\
---------------------------------------------------------------------------

    \2281\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.; 
Blair, A. 2003. Mortality from lymphohematopoietic malignancies 
among workers in formaldehyde industries. Journal of the National 
Cancer Institute 95, pp. 1615-23.
    \2282\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.; 
Blair, A. 2004. Mortality from solid cancers among workers in 
formaldehyde industries. American Journal of Epidemiology 159: 1117-
30.
    \2283\ Beane Freeman, L.E.; Blair, A.; Lubin, J.H.; Stewart, 
P.A.; Hayes, R.B.; Hoover, R.N.; Hauptmann, M. 2009. Mortality from 
lymph hematopoietic malignancies among workers in formaldehyde 
industries: The National Cancer Institute cohort. J. National Cancer 
Inst. 101: 751-61.
    \2284\ Pinkerton, L.E. 2004. Mortality among a cohort of garment 
workers exposed to formaldehyde: an update. Occup. Environ. Med. 61: 
193-200.
    \2285\ Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended 
follow-up of a cohort of British chemical workers exposed to 
formaldehyde. J National Cancer Inst. 95:1608-15.
    \2286\ Hauptmann, M.; Stewart P.A.; Lubin J.H.; Beane Freeman, 
L.E.; Hornung, R.W.; Herrick, R.F.; Hoover, R.N.; Fraumeni, J.F.; 
Hayes, R.B. 2009. Mortality from lymph hematopoietic malignancies 
and brain cancer among embalmers exposed to formaldehyde. Journal of 
the National Cancer Institute 101:1696-1708.
---------------------------------------------------------------------------

    Health effects of formaldehyde in addition to cancer were reviewed 
by the Agency for Toxics Substances and Disease Registry in 1999,\2287\ 
supplemented in 2010,\2288\ and by the World Health Organization.\2289\ 
These organizations reviewed the scientific literature concerning 
health effects linked to formaldehyde exposure to evaluate hazards and 
dose response relationships and defined exposure concentrations for 
minimal risk levels (MRLs). The health endpoints reviewed included 
sensory irritation of eyes and respiratory tract, reduced pulmonary 
function, nasal histopathology, and immune system effects. In addition, 
research on reproductive and developmental effects and neurological 
effects were discussed along with several studies that suggest that 
formaldehyde may increase the risk of asthma--particularly in the 
young.
---------------------------------------------------------------------------

    \2287\ ATSDR (1999). Toxicological Profile for Formaldehyde, 
U.S. Department of Health and Human Services (HHS), July 1999.
    \2288\ ATSDR (2010). Addendum to the Toxicological Profile for 
Formaldehyde. U.S. Department of Health and Human Services (HHS), 
October 2010.
    \2289\ IPCS (2002). Concise International Chemical Assessment 
Document 40. Formaldehyde. World Health Organization.
---------------------------------------------------------------------------

    EPA released a draft Toxicological Review of Formaldehyde--
Inhalation Assessment through the IRIS program for peer review by the 
National Research Council (NRC) and public comment in June 2010.\2290\ 
The draft assessment reviewed more recent research from animal and 
human studies on cancer and other health effects. The NRC released 
their review report in April 2011.\2291\ EPA is currently developing a 
revised draft assessment in response to this review.
---------------------------------------------------------------------------

    \2290\ EPA (2010). Toxicological Review of Formaldehyde (CAS No. 
50-00-0)-Inhalation Assessment: In Support of Summary Information on 
the Integrated Risk Information System (IRIS). External Review 
Draft. EPA/635/R-10/002A. U.S. Environmental Protection Agency, 
Washington DC. Available at http://cfpub.epa.gov/ncea/irs_drats/recordisplay.cfm?deid=223614.
    \2291\ NRC (National Research Council) (2011). Review of the 
Environmental Protection Agency's Draft IRIS Assessment of 
Formaldehyde. Washington DC: National Academies Press. http://books.nap.edu/openbook.php?record_id=13142.
---------------------------------------------------------------------------

(e) Acetaldehyde
    Acetaldehyde is classified in EPA's IRIS database as a probable 
human carcinogen, based on nasal tumors in rats, and is considered 
toxic by the inhalation, oral, and intravenous routes.\2292\ The URE in 
IRIS for acetaldehyde is 2.2 x 10-6 per [micro]g/m\3\.\2293\ 
Acetaldehyde is reasonably anticipated to be a human carcinogen by the 
U.S. DHHS in the 13th Report on Carcinogens and is classified as 
possibly carcinogenic to humans (Group 2B) by the 
IARC.2294 2295 Acetaldehyde is currently listed on the IRIS 
Program Multi-Year Agenda for reassessment within the next few years.
---------------------------------------------------------------------------

    \2292\ U.S. EPA (1991). Integrated Risk Information System File 
of Acetaldehyde. Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
electronically at http://www3.epa.gov/iris/subst/0290.htm.
    \2293\ U.S. EPA (1991). Integrated Risk Information System File 
of Acetaldehyde. This material is available electronically at http://www3.epa.gov/iris/subst/0290.htm.
    \2294\ NTP (National Toxicology Program) 2016. Report on 
Carcinogens Fourteenth Edition, Research Triangle Park, NC: U.S. 
Department of Health and Human Services. Public Health Service. 
Available at https://ntp.niehs.nih.gov/go/roc14.
    \2295\ International Agency for Research on Cancer (IARC) 
(1999). Re-evaluation of some organic chemicals, hydrazine, and 
hydrogen peroxide. IARC Monographs on the Evaluation of Carcinogenic 
Risk of Chemical to Humans, Vol 71. Lyon, France.
---------------------------------------------------------------------------

    The primary noncancer effects of exposure to acetaldehyde vapors 
include irritation of the eyes, skin, and respiratory tract.\2296\ In 
short-term (4 week) rat studies, degeneration of olfactory epithelium 
was observed at various concentration levels of acetaldehyde 
exposure.2297 2298 Data from these studies were used by EPA 
to develop an inhalation reference concentration of 9 [micro]g/m\3\. 
Some asthmatics have been shown to be a sensitive subpopulation to 
decrements in functional expiratory volume (FEV1 test) and 
bronchoconstriction upon acetaldehyde inhalation.\2299\
---------------------------------------------------------------------------

    \2296\ U.S. EPA (1991). Integrated Risk Information System File 
of Acetaldehyde. This material is available electronically at http://www3.epa.gov/iris/subst/0290.htm.
    \2297\ U.S. EPA. (2003). Integrated Risk Information System File 
of Acrolein. Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
electronically at http://www3.epa.gov/iris/subst/0364.htm.
    \2298\ Appleman, L.M., R.A. Woutersen, and V.J. Feron. (1982). 
Inhalation toxicity of acetaldehyde in rats. I. Acute and subacute 
studies. Toxicology. 23: 293-297.
    \2299\ Myou, S.; Fujimura, M.; Nishi K.; Ohka, T.; and Matsuda, 
T. (1993) Aerosolized acetaldehyde induces histamine-mediated 
bronchoconstriction in asthmatics. Am. Rev. Respir. Dis. 148(4 Pt 
1): 940-943.
---------------------------------------------------------------------------

(f) Acrolein
    EPA most recently evaluated the toxicological and health effects 
literature related to acrolein in 2003 and concluded that the human 
carcinogenic potential of acrolein could not be determined because the 
available data were inadequate. No information was available on the 
carcinogenic effects of acrolein in humans and the animal data provided 
inadequate evidence of

[[Page 24867]]

carcinogenicity.\2300\ The IARC determined in 1995 that acrolein was 
not classifiable as to its carcinogenicity in humans.\2301\
---------------------------------------------------------------------------

    \2300\ U.S. EPA (2003). Integrated Risk Information System File 
of Acrolein. Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
at http://www3.epa.gov/iris/subst/0364.htm.
    \2301\ International Agency for Research on Cancer (1995). 
Monographs on the evaluation of carcinogenic risk of chemicals to 
humans, Volume 63. Dry cleaning, some chlorinated solvents and other 
industrial chemicals, World Health Organization, Lyon, France.
---------------------------------------------------------------------------

    Lesions to the lungs and upper respiratory tract of rats, rabbits, 
and hamsters have been observed after sub-chronic exposure to 
acrolein.\2302\ The agency has developed an RfC for acrolein of 0.02 
[micro]g/m\3\ and an RfD of 0.5 [micro]g/kg-day.\2303\
---------------------------------------------------------------------------

    \2302\ U.S. EPA (2003). Integrated Risk Information System File 
of Acrolein. Office of Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
at http://www3.epa.gov/iris/subst/0364.htm.
    \2303\ U.S. EPA (2003). Integrated Risk Information System File 
of Acrolein. Office of Research and Development, National Center for 
Environmental Assessment, Washington, DC. This material is available 
at http://www3.epa.gov/iris/subst/0364.htm.
---------------------------------------------------------------------------

    Acrolein is extremely acrid and irritating to humans when inhaled, 
with acute exposure resulting in upper respiratory tract irritation, 
mucus hypersecretion and congestion. The intense irritancy of this 
carbonyl has been demonstrated during controlled tests in human 
subjects, who suffer intolerable eye and nasal mucosal sensory 
reactions within minutes of exposure.\2304\ These data and additional 
studies regarding acute effects of human exposure to acrolein are 
summarized in EPA's 2003 Toxicological Review of Acrolein.\2305\ 
Studies in humans indicate that levels as low as 0.09 ppm (0.21 mg/
m\3\) for five minutes may elicit subjective complaints of eye 
irritation with increasing concentrations leading to more extensive 
eye, nose and respiratory symptoms. Acute exposures in animal studies 
report bronchial hyper-responsiveness. Based on animal data (more 
pronounced respiratory irritancy in mice with allergic airway disease 
in comparison to non-diseased mice) \2306\ and demonstration of similar 
effects in humans (e.g., reduction in respiratory rate), individuals 
with compromised respiratory function (e.g., emphysema, asthma) are 
expected to be at increased risk of developing adverse responses to 
strong respiratory irritants such as acrolein. EPA does not currently 
have an acute reference concentration for acrolein. The available 
health effect reference values for acrolein have been summarized by EPA 
and include an ATSDR MRL for acute exposure to acrolein of 7 [micro]g/
m\3\ for 1-14 days' exposure; and Reference Exposure Level (REL) values 
from the California Office of Environmental Health Hazard Assessment 
(OEHHA) for one-hour and 8-hour exposures of 2.5 [micro]g/m\3\ and 0.7 
[micro]g/m\3\, respectively.\2307\
---------------------------------------------------------------------------

    \2304\ U.S. EPA (2003). Toxicological review of acrolein in 
support of summary information on Integrated Risk Information System 
(IRIS) National Center for Environmental Assessment, Washington, DC. 
EPA/635/R-03/003. p. 10. Available online at: http://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
    \2305\ U.S. EPA (2003). Toxicological review of acrolein in 
support of summary information on Integrated Risk Information System 
(IRIS) National Center for Environmental Assessment, Washington, DC. 
EPA/635/R-03/003. Available online at: http://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
    \2306\ Morris JB, Symanowicz PT, Olsen JE, et al. (2003). 
Immediate sensory nerve-mediated respiratory responses to irritants 
in healthy and allergic airway-diseased mice. J Appl Physiol 
94(4):1563-71.
    \2307\ U.S. EPA (2009). Graphical Arrays of Chemical-Specific 
Health Effect Reference Values for Inhalation Exposures (Final 
Report). U.S. Environmental Protection Agency, Washington, DC, EPA/
600/R-09/061, 2009. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=211003.
---------------------------------------------------------------------------

(g) Polycyclic Organic Matter
    The term polycyclic organic matter (POM) defines a broad class of 
compounds that includes the polycyclic aromatic hydrocarbon compounds 
(PAHs). One of these compounds, naphthalene, is discussed separately 
below. POM compounds are formed primarily from combustion and are 
present in the atmosphere in gas and particulate form. Cancer is the 
major concern from exposure to POM. Epidemiologic studies have reported 
an increase in lung cancer in humans exposed to diesel exhaust, coke 
oven emissions, roofing tar emissions, and cigarette smoke; all of 
these mixtures contain POM compounds.2308 2309 Animal 
studies have reported respiratory tract tumors from inhalation exposure 
to benzo[a]pyrene and alimentary tract and liver tumors from oral 
exposure to benzo[a]pyrene.\2310\ In 1997 EPA classified seven PAHs 
(benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene, 
benzo[k]fluoranthene, dibenz[a,h]anthracene, and indeno[1,2,3-
cd]pyrene) as Group B2, probable human carcinogens.\2311\ Since that 
time, studies have found that maternal exposures to PAHs in a 
population of pregnant women were associated with several adverse birth 
outcomes, including low birth weight and reduced length at birth, as 
well as impaired cognitive development in preschool children (3 years 
of age).2312 2313 These and similar studies are being 
evaluated as a part of the ongoing IRIS reassessment of health effects 
associated with exposure to benzo[a]pyrene.
---------------------------------------------------------------------------

    \2308\ Agency for Toxic Substances and Disease Registry (ATSDR). 
(1995). Toxicological profile for Polycyclic Aromatic Hydrocarbons 
(PAHs). Atlanta, GA: U.S. Department of Health and Human Services, 
Public Health Service. Available electronically at http://www.atsdr.cdc.gov/ToxProfiles/TP.asp?id=122&tid=25.
    \2309\ U.S. EPA (2002). Health Assessment Document for Diesel 
Engine Exhaust. EPA/600/8-90/057F Office of Research and 
Development, Washington DC. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.
    \2310\ International Agency for Research on Cancer (IARC). 
(2012). Monographs on the Evaluation of the Carcinogenic Risk of 
Chemicals for Humans, Chemical Agents and Related Occupations. Vol. 
100F. Lyon, France.
    \2311\ U.S. EPA (1997). Integrated Risk Information System File 
of indeno (1,2,3-cd) pyrene. Research and Development, National 
Center for Environmental Assessment, Washington, DC. This material 
is available electronically at http://www3.epa.gov/ncea/iris/subst/0457.htm.
    \2312\ Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002). Effect 
of transplacental exposure to environmental pollutants on birth 
outcomes in a multiethnic population. Environ Health Perspect. 111: 
201-05.
    \2313\ Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang, 
D.; Diaz, D.; Hoepner, L.; Barr, D.; Tu, Y.H.; Camann, D.; Kinney, 
P. (2006). Effect of prenatal exposure to airborne polycyclic 
aromatic hydrocarbons on neurodevelopment in the first 3 years of 
life among inner-city children. Environ Health Perspect 114: 1287-
92.
---------------------------------------------------------------------------

(h) Naphthalene
    Naphthalene is found in small quantities in gasoline and diesel 
fuels. Naphthalene emissions have been measured in larger quantities in 
both gasoline and diesel exhaust compared with evaporative emissions 
from mobile sources, indicating it is primarily a product of 
combustion. Acute (short-term) exposure of humans to naphthalene by 
inhalation, ingestion, or dermal contact is associated with hemolytic 
anemia and damage to the liver and the nervous system.\2314\ Chronic 
(long term) exposure of workers and rodents to naphthalene has been 
reported to cause cataracts and retinal damage.\2315\ The National 
Toxicology

[[Page 24868]]

Program listed naphthalene as ``reasonably anticipated to be a human 
carcinogen'' in 2004 on the basis of bioassays reporting clear evidence 
of carcinogenicity in rats and some evidence of carcinogenicity in 
mice.\2316\ California EPA has released a new risk assessment for 
naphthalene, and the IARC has reevaluated naphthalene and re-classified 
it as Group 2B: Possibly carcinogenic to humans.\2317\
---------------------------------------------------------------------------

    \2314\ U.S. EPA (1998). Toxicological Review of Naphthalene 
(Reassessment of the Inhalation Cancer Risk), Environmental 
Protection Agency, Integrated Risk Information System, Research and 
Development, National Center for Environmental Assessment, 
Washington, DC. This material is available electronically at http://www3.epa.gov/iris/subst/0436.htm.
    \2315\ U.S. EPA (1998). Toxicological Review of Naphthalene 
(Reassessment of the Inhalation Cancer Risk), Environmental 
Protection Agency, Integrated Risk Information System, Research and 
Development, National Center for Environmental Assessment, 
Washington, DC. This material is available electronically at http://www3.epa.gov/iris/subst/0436.htm.
    \2316\ NTP (National Toxicology Program), 2016. Report on 
Carcinogens Fourteenth Edition, Research Triangle Park NC: U.S. 
Department of Health and Human Services, Public Health Service. 
Available at https://ntp.niehs.nih.gov/go/roc14.
    \2317\ International Agency for Research on Cancer (IARC). 
(2002). Monographs on the Evaluation of the Carcinogenic Risk of 
Chemicals for Humans. Vol. 82. Lyon, France.
---------------------------------------------------------------------------

    Naphthalene also causes a number of chronic non-cancer effects in 
animals, including abnormal cell changes and growth in respiratory and 
nasal tissues.\2318\ The current EPA IRIS assessment includes noncancer 
data on hyperplasia and metaplasia in nasal tissue that form the basis 
of the inhalation RfC of 3 [micro]g/m\3\.\2319\ The ATSDR MRL for acute 
exposure to naphthalene is 0.6 mg/kg/day.
---------------------------------------------------------------------------

    \2318\ U.S. EPA (1998). Toxicological Review of Naphthalene, 
Environmental Protection Agency, Integrated Risk Information System, 
Research and Development, National Center for Environmental 
Assessment, Washington, DC. This material is available 
electronically at http://www3.epa.gov/iris/subst/0436.htm.
    \2319\ U.S. EPA (1998). Toxicological Review of Naphthalene. 
Environmental Protection Agency, Integrated Risk Information System 
(IRIS), Research and Development, National Center for Environmental 
Assessment, Washington, DC. Available at http://www3.epa.gov/iris/subst/0436.htm.
---------------------------------------------------------------------------

(i) Other Air Toxics
    In addition to the compounds described above, other compounds in 
gaseous hydrocarbon and PM emissions from motor vehicles will be 
affected by this action. Mobile source air toxic compounds that will 
potentially be impacted include ethylbenzene, propionaldehyde, toluene, 
and xylene. Information regarding the health effects of these compounds 
can be found in EPA's IRIS database.\2320\
---------------------------------------------------------------------------

    \2320\ U.S. EPA Integrated Risk Information System (IRIS) 
database is available at: www3.epa.gov/iris.
---------------------------------------------------------------------------

(j) Current Concentrations
    The most recent available data indicate that the majority of 
Americans continue to be exposed to ambient concentrations of air 
toxics at levels which have the potential to cause adverse health 
effects. The levels of air toxics to which people are exposed vary 
depending on where people live and work and the kinds of activities in 
which they engage, as discussed in detail in EPA's most recent Mobile 
Source Air Toxics Rule. According to the National Air Toxic Assessment 
(NATA) for 2014, mobile sources were responsible for 51 percent of 
outdoor anthropogenic toxic emissions and were the largest contributor 
to cancer and noncancer risk from directly emitted pollutants. Mobile 
sources are also significant contributors to precursor emissions which 
react to form air toxics. Formaldehyde is the largest contributor to 
cancer risk of all 71 pollutants quantitatively assessed in the 2014 
NATA. Mobile sources were responsible for more than 30 percent of 
primary anthropogenic emissions of this pollutant in 2014 and also 
contribute to formaldehyde precursor emissions. Benzene is also a large 
contributor to cancer risk, and mobile sources account for 
approximately 54 percent of ambient exposure. Over the years, EPA has 
implemented a number of mobile source and fuel controls which have 
resulted in VOC reductions, which also reduced formaldehyde, benzene 
and other air toxic emissions.
(k) Exposure and Health Effects Associated With Traffic
    Locations in close proximity to major roadways generally have 
elevated concentrations of many air pollutants emitted from motor 
vehicles. Hundreds of such studies have been published in peer-reviewed 
journals, concluding that concentrations of CO, NO, NO2, 
benzene, aldehydes, particulate matter, black carbon, and many other 
compounds are elevated in ambient air within approximately 300-600 
meters (approximately 1,000-2,000 feet) of major roadways. Highest 
concentrations of most pollutants emitted directly by motor vehicles 
are found at locations within 50 meters (approximately 165 feet) of the 
edge of a roadway's traffic lanes.
    A large-scale review of air quality measurements in the vicinity of 
major roadways between 1978 and 2008 concluded that the pollutants with 
the steepest concentration gradients in vicinities of roadways were CO, 
ultrafine particles, metals, elemental carbon (EC), NO, NOX, 
and several VOCs.\2321\ These pollutants showed a large reduction in 
concentrations within 100 meters downwind of the roadway. Pollutants 
that showed more gradual reductions with distance from roadways 
included benzene, NO2, PM2.5, and 
PM10. In the review article, results varied based on the 
method of statistical analysis used to determine the trend.
---------------------------------------------------------------------------

    \2321\ Karner, A.A.; Eisinger, D.S.; Niemeier, D.A. (2010). 
Near-roadway air quality: synthesizing the findings from real-world 
data. Environ Sci. Technol. 44: pp. 5334-44.
---------------------------------------------------------------------------

    For pollutants with relatively high background concentrations 
relative to near-road concentrations, detecting concentration gradients 
can be difficult. For example, many aldehydes have high background 
concentrations as a result of photochemical breakdown of precursors 
from many different organic compounds. This can make detection of 
gradients around roadways and other primary emission sources difficult. 
However, several studies have measured aldehydes in multiple weather 
conditions and found higher concentrations of many carbonyls downwind 
of roadways.2322 2323 These findings suggest a substantial 
roadway source of these carbonyls.
---------------------------------------------------------------------------

    \2322\ Liu, W.; Zhang, J.; Kwon, J.l; et l. (2006). 
Concentrations and source characteristics of airborne carbonyl 
comlbs measured outside urban residences. J Air Waste Manage Assoc. 
56: 1196-1204.
    \2323\ Cahill, T.M.; Charles, M.J.; Seaman, V.Y. (2010). 
Development and application of a sensitive method to determine 
concentrations of acrolein and other carbonyls in ambient air. 
Health Effects Institute Research Report 149. Available at http://dx.doi.org.
---------------------------------------------------------------------------

    In the past 15 years, many studies have been published with results 
reporting that populations who live, work, or go to school near high-
traffic roadways experience higher rates of numerous adverse health 
effects, compared to populations far away from major roads.\2324\ In 
addition, numerous studies have found adverse health effects associated 
with spending time in traffic, such as commuting or walking along high-
traffic roadways.2325 2326 2327 2328 The health outcomes 
with the strongest evidence linking them with traffic-associated air 
pollutants are respiratory effects, particularly in asthmatic children, 
and cardiovascular effects.
---------------------------------------------------------------------------

    \2324\ In the widely-used PubMed database of health 
publications, between January 1, 1990 and August 18, 2011, 605 
publications contained the keywords ``traffic, pollution, 
epidemiology,'' with approximately half the studies published after 
2007.
    \2325\ Laden, F.; Hart, J.E.; Smith, T.J.; Davis, M.E.; 
Garshick, E. (2007) Cause-specific mortality in the unionized U.S. 
trucking industry. Environmental Health Perspect 115:1192-96.
    \2326\ Peters, A.; von Klot, S.; Heier, M.; Trentinaglia, I.; 
H[ouml]rmann, A.; Wichmann, H.E.; L[ouml]wel, H. (2004) Exposure to 
traffic and the onset of myocardial infarction. New England J Med 
351: 1721-30.
    \2327\ Zanobetti, A.; Stone, P.H.; Spelzer, F.E.; Schwartz, 
J.D.; Coull, B.A.; Suh, H.H.; Nearling, B.D.; Mittleman, M.A.; 
Verrier, R.L.; Gold, D.R. (2009) T-wave alternans, air pollution and 
traffic in high-risk subjects. Am J Cardiol 104: 665-670.
    \2328\ Dubowsky Adar, S.; Adamkiewicz, G.; Gold, D.R.; Schwartz, 
J.; Coull, B.A.; Suh, H. (2007) Ambient and microenvironmental 
particles and exhaled nitric oxide before and after a group bus 
trip. Environ Health Perspect 115: 507-512.

---------------------------------------------------------------------------

[[Page 24869]]

    Numerous reviews of this body of health literature have been 
published as well. In 2010, an expert panel of the Health Effects 
Institute (HEI) published a review of hundreds of exposure, 
epidemiology, and toxicology studies.\2329\ The panel rated how the 
evidence for each type of health outcome supported a conclusion of a 
causal association with traffic-associated air pollution as either 
``sufficient,'' ``suggestive but not sufficient,'' or ``inadequate and 
insufficient.'' The panel categorized evidence of a causal association 
for exacerbation of childhood asthma as ``sufficient.'' The panel 
categorized evidence of a causal association for new onset asthma as 
between ``sufficient'' and ``suggestive but not sufficient.'' 
``Suggestive of a causal association'' was how the panel categorized 
evidence linking traffic-associated air pollutants with exacerbation of 
adult respiratory symptoms and lung function decrement. It categorized 
as ``inadequate and insufficient'' evidence of a causal relationship 
between traffic-related air pollution and health care utilization for 
respiratory problems, new onset adult asthma, chronic obstructive 
pulmonary disease (COPD), nonasthmatic respiratory allergy, and cancer 
in adults and children. Other literature reviews have been published 
with conclusions generally similar to the HEI 
panel's.2330 2331 2332 2333 However, in 2014, researchers 
from the U.S. Centers for Disease Control and Prevention (CDC) 
published a systematic review and meta-analysis of studies evaluating 
the risk of childhood leukemia associated with traffic exposure and 
reported positive associations between ``postnatal'' proximity to 
traffic and leukemia risks, but no such association for ``prenatal'' 
exposures.\2334\
---------------------------------------------------------------------------

    \2329\ Health Effects Institute Panel on the Health Effects of 
Traffic-Related Air Pollution (2010). Traffic-related air pollution: 
a critical review of the literature on emissions, exposure, and 
health effects. HEI Special Report 17. Available at http://www.healtheffects.org.
    \2330\ Boothe, V.L.; Shendell, D.G. (2008). Potential health 
effects associated with residential proximity to freeways and 
primary roads: review of scientific literature, 1999-2006. J Environ 
Health 70: 33-41.
    \2331\ Salam, M.T.; Islam, T.; Gilliland, F.D. (2008). Recent 
evidence for adverse effects of residential proximity to traffic 
sources on asthma. Curr Opin Pulm Med 14: 3-8.
    \2332\ Sun, X.; Zhang, S.; Ma, X. (2014) No association between 
traffic density and risk of childhood leukemia: a meta-analysis. 
Asia Pac J Cancer Prev 15: 5229-32.
    \2333\ Raaschou-Nielsen, O.; Reynolds, P. (2006). Air pollution 
and childhood cancer: a review of the epidemiological literature. 
Int J Cancer 118: 2920-9.
    \2334\ Boothe, VL.; Boehmer, T.K.; Wendel, A.M.; Yip, F.Y. 
(2014) Residential traffic exposure and childhood leukemia: a 
systematic review and meta-analysis. Am J Prev Med 46: 413-422.
---------------------------------------------------------------------------

    Health outcomes with few publications suggest the possibility of 
other effects still lacking sufficient evidence to draw definitive 
conclusions. Among these outcomes with a small number of positive 
studies are neurological impacts (e.g., autism and reduced cognitive 
function) and reproductive outcomes (e.g., preterm birth, low birth 
weight).2335 2336 2337 2338.
---------------------------------------------------------------------------

    \2335\ Volk, H.E.; Hertz-Picciotto, I.; Delwiche, L.; et al. 
(2011). Residential proximity to freeways and autism in the CHARGE 
study. Environ Health Perspect 119: 873-77.
    \2336\ Franco-Suglia, S.; Gryparis, A.; Wright, R.O.; et al. 
(2007). Association of black carbon with cognition among children in 
a prospective birth cohort study. Am J Epidemiol. doi: 10.1093/aje/
kwm308. Available at http://dx.doi.org.
    \2337\ Power, M.C.; Weisskopf, M.G.; Alexeef, SE; et al. (2011). 
Traffic-related air pollution and cognitive function in a cohort of 
older men. Environ Health Perspect 2011: 682-687.
    \2338\ Wu, J.; Wilhelm, M.; Chung, J.; et al. (2011). Comparing 
exposure assessment methods for traffic-related air pollution in and 
adverse pregnancy outcome study. Environ Res 111: 685-6692.
---------------------------------------------------------------------------

    In addition to health outcomes, particularly cardiopulmonary 
effects, conclusions of numerous studies suggest mechanisms by which 
traffic-related air pollution affects health. Numerous studies indicate 
that near-roadway exposures may increase systemic inflammation, 
affecting organ systems, including blood vessels and 
lungs.2339 2340 2341 2342 Long-term exposures in near-road 
environments have been associated with inflammation-associated 
conditions, such as atherosclerosis and 
asthma.2343 2344 2345
---------------------------------------------------------------------------

    \2339\ Riediker, M. (2007). Cardiovascular effects of fine 
particulate matter components in highway patrol officers. Inhal 
Toxicol 19: 99-105. doi: 10.1080/08958370701495238 Available at 
http://dx.doi.org.
    \2340\ Alexeef, SE; Coull, B.A.; Gryparis, A.; et al. (2011). 
Medium-term exposure to traffic-related air pollution and markers of 
inflammation and endothelial function. Environ Health Perspect 119: 
481-486. doi:10.1289/ehp.1002560 Available at http://dx.doi.org.
    \2341\ Eckel. S.P.; Berhane, K.; Salam, M.T.; et al. (2011). 
Traffic-related pollution exposure and exhaled nitric oxide in the 
Children's Health Study. Environ Health Perspect (IN PRESS). 
doi:10.1289/ehp.1103516. Available at http://dx.doi.org.
    \2342\ Zhang, J.; McCreanor, J.E.; Cullinan, P.; et al. (2009). 
Health effects of real-world exposure diesel exhaust in persons with 
asthma. Res Rep Health Effects Inst 138. Available at http://www.healtheffects.org.
    \2343\ Adar, S.D.; Klein, R.; Klein, E.K.; et al. (2010). Air 
pollution and the microvasculatory: a cross-sectional assessment of 
in vivo retinal images in the population-based Multi-Ethnic Study of 
Atherosclerosis. PLoS Med 7(11): E1000372. doi:10.1371/
journal.pmed.1000372. Available at http://dx.doi.org.
    \2344\ Kan, H.; Heiss, G.; Rose, K.M.; et al. (2008). 
Prospective analysis of traffic exposure as a risk factor for 
incident coronary heart disease: the Atherosclerosis Risk in 
Communities (ARIC) study. Environ Health Perspect 116: 1463-1468. 
doi:10.1289/ehp.11290. Available at http://dx.doi.org.
    \2345\ McConnell, R.; Islam, T.; Shankardass, K.; et al. (2010). 
Childhood incident asthma and traffic-related air pollution at home 
and school. Environ Health Perspect 1021-26.
---------------------------------------------------------------------------

    Several studies suggest that some factors may increase 
susceptibility to the effects of traffic-associated air pollution. 
Several studies have found stronger respiratory associations in 
children experiencing chronic social stress, such as in violent 
neighborhoods or in homes with high family 
stress.2346 2347 2348
---------------------------------------------------------------------------

    \2346\ Islam, T.; Urban, R.; Gauderman, W.J.; et al. (2011). 
Parental stress increases the detrimental effect of traffic exposure 
on children's lung function. Am J Respir Crit Care Med (In press).
    \2347\ Clougherty, J.E.; Levy, J.I.; Kubzansky, L.D.; et al. 
(2007). Synergistic effects of traffic-related air pollution and 
exposure to violence on urban asthma etiology. Environ Health 
Perspect 115: 1140-46.
    \2348\ Chen, E.; Schrier, H.M.; Strunk, R.C.; et al. (2008). 
Chronic traffic-related air pollution and stress interact to predict 
biologic and clinical outcomes in asthma. Environ Health Perspect 
116: 970-5.
---------------------------------------------------------------------------

    The risks associated with residence, workplace, or schools near 
major roads are of potentially high public health significance due to 
the large population in such locations. According to the 2009 American 
Housing Survey, over 22 million homes (17.0 percent of all U.S. housing 
units) were located within 300 feet of an airport, railroad, or highway 
with four or more lanes. This corresponds to a population of more than 
50 million U.S. residents in close proximity to high-traffic roadways 
or other transportation sources. Based on 2010 Census data, a 2013 
publication estimated that 19 percent of the U.S. population (over 59 
million people) lived within 500 meters of roads with at least 25,000 
annual average daily traffic (AADT), while about 3.2 percent of the 
population lived within 100 meters (about 300 feet) of such 
roads.\2349\ Another 2013 study estimated that 3.7 percent of the U.S. 
population (about 11.3 million people) lived within 150 meters (about 
500 feet) of interstate highways or other freeways and 
expressways.\2350\ On average, populations near major roads have higher 
fractions of minority residents and lower socioeconomic status. 
Furthermore, on average, Americans spend more than an hour traveling 
each day, bringing nearly all residents into a

[[Page 24870]]

high-exposure microenvironment for part of the day.
---------------------------------------------------------------------------

    \2349\ Rowangould, G.M. (2013). A census of the U.S. near-
roadway population: public health and environmental justice 
considerations. Transportation Research Part D 25: 59-67.
    \2350\ Boehmer, T.K.; Foster, S.L.; Henry, J.R.; Woghiren-
Akinnifesi, E.L.; Yip, F.Y. (2013) Residential proximity to major 
highways--United States, 2010. Morbidity and Mortality Weekly Report 
62(3); 46-50.
---------------------------------------------------------------------------

    In light of these concerns, EPA has required through the NAAQS 
process that air quality monitors be placed near high-traffic roadways 
for determining concentrations of CO, NO2, and 
PM2.5 (in addition to those existing monitors located in 
neighborhoods and other locations farther away from pollution sources). 
Near-roadway monitors for NO2 began operation between 2014 
and 2017 in Core Based Statistical Areas (CBSAs) with population of at 
least 500,000. Monitors for CO and PM2.5 began operation 
between 2015 and 2017. These monitors will further the understanding of 
exposure in these locations.
    EPA and DOT continue to research near-road air quality, including 
the types of pollutants found in high concentrations near major roads 
and health problems associated with the mixture of pollutants near 
roads.
(8) Environmental Effects of Non-GHG Pollutants
(a) Visibility
    Visibility can be defined as the degree to which the atmosphere is 
transparent to visible light.\2351\ Visibility impairment is caused by 
light scattering and absorption by suspended particles and gases. 
Visibility is important because it has direct significance to people's 
enjoyment of daily activities in all parts of the country. Individuals 
value good visibility for the well-being it provides them directly, 
where they live and work, and in places where they enjoy recreational 
opportunities. Visibility is also highly valued in significant natural 
areas, such as national parks and wilderness areas, and special 
emphasis is given to protecting visibility in these areas. For more 
information on visibility see the final 2019 p.m. 
ISA.2352 2353
---------------------------------------------------------------------------

    \2351\ National Research Council, (1993). Protecting Visibility 
in National Parks and Wilderness Areas. National Academy of Sciences 
Committee on Haze in National Parks and Wilderness Areas. National 
Academy Press, Washington, DC. Available at http://www.nap.edu/books/0309048443/html/.
    \2352\ U.S. EPA. Integrated Science Assessment (ISA) for 
Particulate Matter (Final Report 2019). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
    \2353\ There is an ongoing review of the ISA for Oxides of 
Nitrogen Oxides of Sulfur, and Particulate Matter (Ecological 
Criteria), Available at https://wwwepa.gov/isa/integrated-science-assessment-isa-oxides-nitrogen-oxides-sulfur-andparticulate-matter.
---------------------------------------------------------------------------

    EPA is working to address visibility impairment. Reductions in air 
pollution from implementation of various programs associated with the 
Clean Air Act Amendments of 1990 (CAAA) provisions have resulted in 
substantial improvements in visibility and will continue to do so in 
the future. Because trends in haze are closely associated with trends 
in particulate sulfate and nitrate due to the relationship between 
their concentration and light extinction, visibility trends have 
improved as emissions of SO2 and NOX have 
decreased over time due to air pollution regulations such as the Acid 
Rain Program.\2354\
---------------------------------------------------------------------------

    \2354\ U.S. EPA (2009). Final Report: Integrated Science 
Assessment for Particulate Matter. U.S. Environmental Protection 
Agency, Washington, DC, EPA/600/R-08/139F, 2009.
---------------------------------------------------------------------------

    In the Clean Air Act Amendments of 1977, Congress recognized 
visibility's value to society by establishing a national goal to 
protect national parks and wilderness areas from visibility impairment 
caused by manmade pollution.\2355\ In 1999, EPA finalized the regional 
haze program to protect the visibility in Mandatory Class I Federal 
areas.\2356\ There are 156 national parks, forests and wilderness areas 
categorized as Mandatory Class I Federal areas.\2357\ These areas are 
defined in CAA Section 162 as those national parks exceeding 6,000 
acres, wilderness areas and memorial parks exceeding 5,000 acres, and 
all international parks which were in existence on August 7, 1977.
---------------------------------------------------------------------------

    \2355\ See Section 169(a) of the Clean Air Act.
    \2356\ 64 FR 35714 (July 1, 1999).
    \2357\ 62 FR 38680-81 (July 18, 1997).
---------------------------------------------------------------------------

    EPA has also concluded that PM2.5 causes adverse effects 
on visibility in other areas that are not targeted by the Regional Haze 
Rule, such as urban areas, depending on PM2.5 concentrations 
and other factors such as dry chemical composition and relative 
humidity (i.e., an indicator of the water composition of the 
particles). EPA revised the PM2.5 standards in December 2012 
and established a target level of protection that is expected to be met 
through attainment of the existing secondary standards for 
PM2.5.
(b) Plant and Ecosystem Effects of Ozone
    The welfare effects of ozone include effects on ecosystems, which 
can be observed across a variety of scales, i.e. subcellular, cellular, 
leaf, whole plant, population and ecosystem. Ozone can produce both 
acute and chronic injury in sensitive species depending on the 
concentration level and the duration of the exposure.\2358\ In those 
sensitive species,\2359\ effects from repeated exposure to ozone 
throughout the growing season of the plant can tend to accumulate, so 
that even relatively low concentrations experienced for a longer 
duration have the potential to create chronic stress on 
vegetation.\2360\ Ozone damage to sensitive species includes impaired 
photosynthesis and visible injury to leaves. The impairment of 
photosynthesis, the process by which the plant makes carbohydrates (its 
source of energy and food), can lead to reduced crop yields, timber 
production, and plant productivity and growth. Impaired photosynthesis 
can also lead to a reduction in root growth and carbohydrate storage 
below ground, resulting in other, more subtle plant and ecosystems 
impacts.\2361\ These latter impacts include increased susceptibility of 
plants to insect attack, disease, harsh weather, interspecies 
competition and overall decreased plant vigor. The adverse effects of 
ozone on areas with sensitive species could potentially lead to species 
shifts and loss from the affected ecosystems,\2362\ resulting in a loss 
or reduction in associated ecosystem goods and services. Additionally, 
visible ozone injury to leaves can result in a loss of aesthetic value 
in areas of special scenic significance like national parks and 
wilderness areas and reduced use of sensitive ornamentals in 
landscaping.\2363\
---------------------------------------------------------------------------

    \2358\ 73 FR 16486 (March 27, 2008).
    \2359\ 73 FR 16491 (March 27, 2008). Only a small percentage of 
all the plant species growing within the U.S. (over 43,000 species 
have been catalogued in the USDA PLANTS database) have been studied 
with respect to ozone sensitivity.
    \2360\ The concentration at which ozone levels overwhelm a 
plant's ability to detoxify or compensate for oxidant exposure 
varies. Thus, whether a plant is classified as sensitive or tolerant 
depends in part on the exposure levels being considered. Chapter 9, 
Section 9.3.4 of U.S. EPA, 2013 Integrated Science Assessment for 
Ozone and Related Photochemical Oxidants. Office of Research and 
Development/National Center for Environmental Assessment. U.S. 
Environmental Protection Agency. EPA 600/R-10/076F.
    \2361\ 73 FR 16492 (March 27, 2008).
    \2362\ 73 FR 16493-94 (March 27, 2008). Ozone impacts could be 
occurring in areas where plant species sensitive to ozone have not 
yet been studied or identified.
    \2363\ 73 FR 16490-97 (March 27, 2008).
---------------------------------------------------------------------------

    The most recent Integrated Science Assessment (ISA) for Ozone 
presents more detailed information on how ozone affects vegetation and 
ecosystems.2364 2365 The ISA concludes that ambient 
concentrations of ozone are associated with a number of adverse welfare 
effects and characterizes the

[[Page 24871]]

weight of evidence for different effects associated with ozone.\2366\ 
The ISA concludes that visible foliar injury effects on some 
vegetation, reduced vegetation growth, reduced productivity in 
terrestrial ecosystems, reduced yield and quality of some agricultural 
crops, and alteration of below-ground biogeochemical cycles are 
causally associated with exposure to ozone. It also concludes that 
reduced carbon sequestration in terrestrial ecosystems, alteration of 
terrestrial ecosystem water cycling, and alteration of terrestrial 
community composition are likely to be causally associated with 
exposure to ozone.
---------------------------------------------------------------------------

    \2364\ U.S. EPA. Integrated Science Assessment of Ozone and 
Related Photochemical Oxidants (Final Report). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA 
is available at http://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=247492#Download.
    \2365\ There is an ongoing review of the ozone NAAQS, EPA 
intends to finalize an updated Integrated Science Assessment in 
early 2020 Available at (https://www.epa.gov naaqs/ozone-o3-
standards-integrated-science-assessments-currentreview).
    \2366\ The Ozone ISA evaluates the evidence associated with 
different ozone related health and welfare effects, assigning one of 
five ``weight of evidence'' determinations: causal relationship, 
likely to be a causal relationship, suggestive of a causal 
relationship, inadequate to infer a causal relationship, and not 
likely to be a causal relationship. For more information on these 
levels of evidence, please refer to Table II of the ISA.
---------------------------------------------------------------------------

(c) Atmospheric Deposition
    Wet and dry deposition of ambient particulate matter delivers a 
complex mixture of metals (e.g., mercury, zinc, lead, nickel, aluminum, 
and cadmium), organic compounds (e.g., polycyclic organic matter, 
dioxins, and furans) and inorganic compounds (e.g., nitrate, sulfate) 
to terrestrial and aquatic ecosystems. The chemical form of the 
compounds deposited depends on a variety of factors including ambient 
conditions (e.g., temperature, humidity, oxidant levels) and the 
sources of the material. Chemical and physical transformations of the 
compounds occur in the atmosphere as well as the media onto which they 
deposit. These transformations in turn influence the fate, 
bioavailability and potential toxicity of these compounds.
    Adverse impacts to human health and the environment can occur when 
particulate matter is deposited to soils, water, and biota.\2367\ 
Deposition of heavy metals or other toxics may lead to the human 
ingestion of contaminated fish, impairment of drinking water, damage to 
terrestrial, freshwater and marine ecosystem components, and limits to 
recreational uses. Atmospheric deposition has been identified as a key 
component of the environmental and human health hazard posed by several 
pollutants including mercury, dioxin and PCBs.\2368\
---------------------------------------------------------------------------

    \2367\ U.S. EPA. Integrated Science Assessment for Particulate 
Matter (Final Report). U.S. Environmental Protection Agency, 
Washington, DC, EPA/600/R-08/139F, 2009.
    \2368\ U.S. EPA (2000). Deposition of Air Pollutants to the 
Great Waters: Third Report to Congress. Office of Air Quality 
Planning and Standards. EPA-453/R-00-0005.
---------------------------------------------------------------------------

    The ecological effects of acidifying deposition and nutrient 
enrichment are detailed in the Integrated Science Assessment for Oxides 
of Nitrogen and Sulfur-Ecological Criteria.2369 2370 
Atmospheric deposition of nitrogen and sulfur contributes to 
acidification, altering biogeochemistry and affecting animal and plant 
life in terrestrial and aquatic ecosystems across the United States. 
The sensitivity of terrestrial and aquatic ecosystems to acidification 
from nitrogen and sulfur deposition is predominantly governed by 
geology. Prolonged exposure to excess nitrogen and sulfur deposition in 
sensitive areas acidifies lakes, rivers and soils. Increased acidity in 
surface waters creates inhospitable conditions for biota and affects 
the abundance and biodiversity of fishes, zooplankton and 
macroinvertebrates and ecosystem function. Over time, acidifying 
deposition also removes essential nutrients from forest soils, 
depleting the capacity of soils to neutralize future acid loadings and 
negatively affecting forest sustainability. Major effects in forests 
include a decline in sensitive tree species, such as red spruce (Picea 
rubens) and sugar maple (Acer saccharum). In addition to the role 
nitrogen deposition plays in acidification, nitrogen deposition also 
leads to nutrient enrichment and altered biogeochemical cycling. In 
aquatic systems increased nitrogen can alter species assemblages and 
cause eutrophication. In terrestrial systems nitrogen loading can lead 
to loss of nitrogen-sensitive lichen species, decreased biodiversity of 
grasslands, meadows and other sensitive habitats, and increased 
potential for invasive species.
---------------------------------------------------------------------------

    \2369\ NOX and SOX secondary ISA2369 U.S. 
EPA. Integrated Science Assessment (ISA) for Oxides of Nitrogen and 
Sulfur Ecological Criteria (Final Report). U.S. Environmental 
Protection Agency, Washington, DC, EPA/600/R-08/082F, 2008.
    \2370\ There is an ongoing review of the ISA for Oxides and 
Nitrogen, Oxides of Sulfur, and Particulate Matter (Ecological 
Criteria), Available at https://www.epa.gov/isa/integrated-science-assessment-isa-oxides-nitrogen-oxides-sulfur-and-particulate-matter.
---------------------------------------------------------------------------

    Building materials including metals, stones, cements, and paints 
undergo natural weathering processes from exposure to environmental 
elements (e.g., wind, moisture, temperature fluctuations, sunlight, 
etc.). Pollution can worsen and accelerate these effects. Deposition of 
PM is associated with both physical damage (materials damage effects) 
and impaired aesthetic qualities (soiling effects). Wet and dry 
deposition of PM can physically affect materials, adding to the effects 
of natural weathering processes, by potentially promoting or 
accelerating the corrosion of metals, by degrading paints and by 
deteriorating building materials such as stone, concrete and 
marble.\2371\ The effects of PM are exacerbated by the presence of 
acidic gases and can be additive or synergistic due to the complex 
mixture of pollutants in the air and surface characteristics of the 
material. Acidic deposition has been shown to have an effect on 
materials including zinc/galvanized steel and other metal, carbonate 
stone (as monuments and building facings), and surface coatings 
(paints).\2372\ The effects on historic buildings and outdoor works of 
art are of particular concern because of the uniqueness and 
irreplaceability of many of these objects. In addition to aesthetic and 
functional effects on metals, stone and glass, altered energy 
efficiency of photovoltaic panels by PM deposition is also becoming an 
important consideration for impacts of air pollutants on materials.
---------------------------------------------------------------------------

    \2371\ U.S. EPA. Integrated Science Assessment (ISA) for 
Particulate Matter (Final Report, 2019). U.S Environmental 
Protection Agency, Washington, DC, EPA/600/R-l9/188, 2019.
    \2372\ Irving, P.M., e.d. 1991. Acid Deposition: State of 
Science and Technology, Volume III, Terrestrial, Materials, Health, 
and Visibility Effects, The U.S. National Acid Precipitation 
Assessment Program, Chapter 24, pp. 24-76.
---------------------------------------------------------------------------

(d) Environmental Effects of Air Toxics
    Emissions from producing, transporting and combusting fuel 
contribute to ambient levels of pollutants that contribute to adverse 
effects on vegetation. Volatile organic compounds, some of which are 
considered air toxics, have long been suspected to play a role in 
vegetation damage.\2373\ In laboratory experiments, a wide range of 
tolerance to VOCs has been observed.\2374\ Decreases in harvested seed 
pod weight have been reported for the more sensitive plants, and some 
studies have reported effects on seed germination, flowering and fruit 
ripening. Effects of individual VOCs or their role in conjunction with 
other stressors (e.g., acidification, drought, temperature extremes) 
have not been well studied. In a recent study of a mixture of VOCs 
including ethanol and toluene on herbaceous plants, significant effects 
on seed production, leaf water content and photosynthetic

[[Page 24872]]

efficiency were reported for some plant species.\2375\
---------------------------------------------------------------------------

    \2373\ U.S. EPA (1991). Effects of organic chemicals in the 
atmosphere on terrestrial plants. EPA/600/3-91/001.
    \2374\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M 
Skewes, DN Price AR Brown, AD Sharpe. (2003). Effects of VOCs on 
herbaceous plants in an open-top chamber experiment. Environ. 
Pollut. 124:341-343.
    \2375\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M 
Skewes, DN Price AR Brown, AD Sharpe. (2003). Effects of VOCs on 
herbaceous plants in an open-top chamber experiment. Environ. 
Pollut. 124:341-343.
---------------------------------------------------------------------------

    Research suggests an adverse impact of vehicle exhaust on plants, 
which has in some cases been attributed to aromatic compounds and in 
other cases to nitrogen oxides.2376 2377 2378 The impacts of 
VOCs on plant reproduction may have long-term implications for 
biodiversity and survival of native species near major roadways. Most 
of the studies of the impacts of VOCs on vegetation have focused on 
short-term exposure and few studies have focused on long-term effects 
of VOCs on vegetation and the potential for metabolites of these 
compounds to affect herbivores or insects.
---------------------------------------------------------------------------

    \2376\ Viskari E.-L. (2000). Epicuticular wax of Norway spruce 
needles as indicator of traffic pollutant deposition. Water, Air, 
and Soil Pollut. 121:327-337.
    \2377\ Ugrekhelidze D, F Korte, G Kvesitadze (1997). Uptake and 
transformation of benzene and toluene by plant leaves. Ecotox. 
Environ. Safety 37:24-29.
    \2378\ Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D 
Knoppik, B Hock. (1987). Toxic components of motor vehicle emissions 
for the spruce Picea abies. Environ. Pollut. 48: 235-43.
---------------------------------------------------------------------------

(c) How the Agencies Estimated Impacts on Emissions
    The rule implements an emissions inventory methodology for 
estimating impacts. Vehicle emissions inventories are often described 
as three-legged stools, comprised of activity (i.e., miles traveled, 
hours operated, or gallons of gasoline burned), population (or number 
of vehicles), and emission factors. An emissions factor is a 
representative value that attempts to relate the quantity of a 
pollutant released to the atmosphere with an activity associated with 
the release of that pollutant.\2379\ Depending on the vehicle activity 
available, emission factors may be on a distance-, time-, or fuel-
basis. For example, an emissions inventory for a light-duty fleet could 
simply be the vehicle miles traveled multiplied by the appropriate per-
mile emission factor for a chosen pollutant.
---------------------------------------------------------------------------

    \2379\ USEPA, Basics Information of Air Emissions Factors and 
Quantification, https://www.epa.gov/air-emissions-factors-and-quantification/basic-information-air-emissions-factors-and-quantification.
---------------------------------------------------------------------------

    As described in Section VI.A, Overview of Methods, the agencies 
used specific models to develop inputs to the CAFE model, such as fuel 
prices and emission factors. The CAFE model estimates how manufacturers 
might respond to a given regulatory scenario (CAFE/CO2 
standards) and fuel prices, and what impact that response will have on 
emissions. As mentioned above, the agencies have used DOT's CAFE model 
to estimate impacts of the CAFE and CO2 standards 
promulgated today. Details of the analysis are presented below and in 
the accompanying RIA, EIS, and model documentation. To estimate the 
response on emissions, several steps are involved. The estimation of 
emissions involves accounting for vehicular fuel type (e.g., gasoline, 
diesel, electric) and fuel economy (accounting for the estimated gap, 
discussed below, between ``laboratory'' and actual on-road fuel 
economy), vehicular turnover and travel demand, fuel properties (carbon 
content), and upstream process emissions. Like other models, the CAFE 
model includes procedures to estimate annual rates at which new 
vehicles are used and subsequently scrapped. Together, these procedures 
result in, for each vehicle model in each model year, estimates of the 
number remaining in service in each calendar year, as well as the 
annual mileage accumulation (i.e. VMT) in each calendar year. 
Quantities of emissions derive from this vehicle operation.
    For every vehicle model in the market file, the model estimates the 
VMT per vehicle (using the assumed VMT schedule, the vehicle fuel 
economy, fuel price, and the rebound assumption). Those miles are 
multiplied by the number off each vehicle model/configuration remaining 
in service in any given calendar year. Fuel consumption is the product 
of miles driven and fuel economy, which can be tracked by model year 
cohort in the model. Carbon dioxide emissions from vehicle tailpipes 
are the simple product of gallons consumed and the carbon content of 
each gallon. As discussed in the CAFE model overview, the simulated 
application of technology results in estimates of the cost, fuel type, 
fuel economy, and fuel share applicable to each vehicle model in each 
model year. Together with quantities of travel, and with estimates of 
the ``gap'' between ``laboratory'' and ``on-road'' fuel economy, these 
enable calculation of quantities of fuel consumed in each year during 
the useful life of each vehicle model produced in each model year. The 
model calculates emissions of CO2, CH4, and 
N2O, criteria pollutants, and air toxics, reporting 
emissions both from vehicle tailpipes and from upstream processes 
(e.g., petroleum refining) involving in producing and supplying fuels.
    In order to calculate calendar year fuel consumption, the model 
needs to account for the inherited on-road fleet in addition to the 
model year cohorts affected by this rule. Using the VMT of the average 
passenger car and light truck from each cohort, the model computes the 
fuel consumption of each model year class of vehicles for its age in a 
given CY. The sum across all ages (and thus, model year cohorts) in a 
given CY provides estimated CY fuel consumption.
    For this rule, vehicle tailpipe (downstream) and upstream emission 
inventories were developed separately. In addition to the tailpipe 
emissions of carbon dioxide, each gallon of gasoline produced for 
consumption by the on-road fleet has associated ``upstream'' emissions 
that occur in the extraction, transportation, refining, and 
distribution of the fuel. The tailpipe inventories apply per-mile 
emission factors from the Motor Vehicle Emission Simulator (MOVES) and 
the upstream inventories apply per-gallon of fuel consumed emission 
factors from the Argonne National Laboratory's Greenhouse gases, 
Regulated Emissions, and Energy use in Transportation (GREET) Model. 
The model accounts for upstream emissions and reports them accordingly. 
More detailed descriptions of emission data sources and calculations 
are provided in the following section.
    The agencies received several comments on estimation of criteria 
pollutant impacts in the NPRM. As discussed elsewhere in this preamble, 
EDF modified aspects of the CAFE model as part of their comments to the 
agencies. Specifically in regards to criteria pollutant emissions, EDF 
made several alternative assumptions, including assertions that 
criteria pollutant impacts were not as negligible as the agencies 
claimed, and that fatalities due to criteria pollutant emissions would 
be higher than the agencies showed in the NPRM. The agencies declined 
to adopt EDF's suggested changes to the model and inputs, but did make 
the changes discussed in this section that refined the agencies' 
accounting of criteria pollutant emissions and explicitly modeled 
criteria pollutant fatalities, as discussed below.
    Also discussed elsewhere in this preamble, some commenters 
expressed that the agencies' analysis (by implication, their modeling) 
should account for some States' mandates that manufacturers sell 
minimum quantities of ``Zero Emission Vehicles'' (ZEVs).\2380\ These 
commenters stressed the

[[Page 24873]]

importance of the ZEV mandate in relation to maintaining air quality 
requirements and reducing effects of climate change.
---------------------------------------------------------------------------

    \2380\ CBD et al., NHTSA-2018-0067-12123; States and Cities, 
NHTSA-2018-0067-11735; SCAQMD, NHTSA-2018-0067-11813.
---------------------------------------------------------------------------

    The reference case analysis for today's rule, like that for the 
proposal, does not simulate compliance with ZEV mandates,\2381\ because 
such mandates are subject to preemption under EPCA and are therefore 
not enforceable. As discussed in the One National Program Action, 
California and other states remain free to revise their overall average 
emissions standards to further reduce ozone forming emissions and seek 
a waiver of Clean Air Act preemption from EPA, as described above, 
while not violating NHTSA's preemption authority. These States and 
local governments would continue to be allowed to take other actions so 
long as those are not related to fuel economy and are consistent with 
any other relevant Federal law.
---------------------------------------------------------------------------

    \2381\ The NPRM version of the model included experimental 
capabilities to account for mandates and credits for the sale of 
ZEVs, but the agencies did not utilize those capabilities for the 
NPRM for the same reasons discussed above.
---------------------------------------------------------------------------

(1) Activity Levels
    As discussed in Section VI.A, for each vehicle model/configuration 
in each model year during 2017-2050, the CAFE model estimates and 
records the fuel type (e.g., gasoline, electricity), fuel economy, and 
number of units sold in the U.S. The model also makes use of an 
aggregated representation of vehicles sold in the U.S. during 1978-
2016. The model estimates the numbers of each cohort of vehicles 
remaining in service in each calendar year, and the amount of driving 
accumulated by each such cohort in each calendar year. The CAFE model 
estimates annual vehicle-miles of travel (VMT) for each individual car 
and light truck model produced in each model year at each age of their 
lifetimes, which extend for a maximum of 40 years. Since a vehicle's 
age is equal to the current calendar year minus the model year in which 
it was originally produced, the age span of each vehicle model's 
lifetime corresponds to a sequence of 40 calendar years beginning in 
the calendar year corresponding to the model year it was 
produced.\2382\ These estimates reflect the gradual decline in the 
fraction of each car and light truck model's original model year 
production volume that is expected to remain in service during each 
year of its lifetime, as well as the well-documented decline in their 
typical use as they age. Using this relationship, the CAFE model 
calculates total VMT for the entire fleet of cars and light trucks in 
service during each calendar year spanned by the agencies' analysis.
---------------------------------------------------------------------------

    \2382\ In practice, many vehicle models bearing a given model 
year designation become available for sale in the preceding calendar 
year, and their sales can extend through the following calendar year 
as well. However, the CAFE model does not attempt to distinguish 
between model years and calendar years; vehicles bearing a model 
year designation are assumed to be produced and sold in that same 
calendar year.
---------------------------------------------------------------------------

    Based on these estimates, the model also calculates quantities of 
each type of fuel or energy, including gasoline, diesel, and 
electricity, consumed in each calendar year. By combining these with 
estimates of each model's fuel or energy efficiency, the model also 
estimates the quantity and energy content of each type of fuel consumed 
by cars and light trucks at each age, or viewed another way, during 
each calendar year of their lifetimes. As with the accounting of VMT, 
these estimates of annual fuel or energy consumption for each vehicle 
model and model year combination are combined to calculate the total 
volume of each type of fuel or energy consumed during each calendar 
year, as well as its aggregate energy content.
    The procedures the CAFE model uses to estimate annual VMT for 
individual car and light truck models produced during each model year 
over their lifetimes and to combine these into estimates of annual 
fleet-wide travel during each future calendar year, together with the 
sources of its estimates of their survival rates and average use at 
each age, are described in detail in Section VI.D.1 of this final rule. 
The data and procedures it employs to convert these estimates of VMT to 
fuel and energy consumption by individual model, and to aggregate the 
results to calculate total consumption and energy content of each fuel 
type during future calendar years, are also described in detail in that 
same section.
    The model documentation accompanying today's notice describes these 
procedures in detail.\2383\ The quantities of travel and fuel 
consumption estimated for the cross section of model years and calendar 
years constitutes a set of ``activity levels'' based on which the model 
calculates emissions. The model does so by multiplying activity levels 
by emission factors. As indicated in the previous section, the 
resulting estimates of vehicle use (VMT), fuel consumption, and fuel 
energy content are combined with emission factors drawn from various 
sources to estimate emissions of GHGs, criteria air pollutant, and 
airborne toxic compound that occur throughout the fuel supply and 
distribution process, as well as during vehicle operation, storage, and 
refueling. Emission factors measure the mass of each GHG or criteria 
pollutant emitted per vehicle-mile of travel, gallon of fuel consumed, 
or unit of fuel energy content. The following section identifies the 
sources of these emission factors and explains in detail how the CAFE 
model applies them to its estimates of vehicle travel, fuel use, and 
fuel energy consumption to estimate total annual emissions of each GHG, 
criteria pollutant, and airborne toxic.
---------------------------------------------------------------------------

    \2383\ CAFE model documentation is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

(2) What emission factors did the agencies apply?
(a) Tailpipe (Downstream) Emission Factors
    In a full fuel cycle analysis, emissions that occur from the 
fueling pump to vehicle wheels are usually referred to as tailpipe or 
simply downstream emissions. Today's rule primarily impacts 
CO2 emissions. The agencies have calculated tailpipe 
CO2 emissions based on fuel consumption and fuel properties 
(i.e., fuel density and carbon content) that result in gram per gallon 
emission factors. For all other exhaust constituents (except sulfur 
dioxide, discussed below), the agencies have calculated emissions by 
applying per-mile emission factors to quantities of travel (i.e., VMT). 
This rulemaking's tailpipe emission factors are from EPA's Motor 
Vehicle Emission Simulator (MOVES), which serves as the federal 
regulatory model for mobile-source emission inventories, with a few 
notable exceptions. In particular, light-duty gasoline and diesel 
tailpipe emission factors for the following criteria pollutants, 
greenhouse gases (other than CO2), and air toxics are drawn 
from MOVES2014a: \2384\
---------------------------------------------------------------------------

    \2384\ For the emission factors informing the Final EIS, 
updating to MOVES 2014b would have produced values identical to 
those based on MOVES 2014a.

 Criteria pollutants
    [cir] Carbon monoxide (CO),
    [cir] Volatile organic compounds (VOC),
    [cir] Nitrogen oxides (NOX), and
    [cir] Fine particulate matter (PM2.5)
 Greenhouse gases
    [cir] Methane (CH4), and
    [cir] Nitrous oxide (N2O)
 Air toxics
    [cir] Acetaldehyde,
    [cir] Acrolein,
    [cir] Benzene,
    [cir] Butadiene,
    [cir] Formaldehyde,
    [cir] Diesel particulate matter (DPM10), and

[[Page 24874]]

    [cir] Methyl tert-butyl ether (MTBE)

    These MOVES-based emission factors are specified separately for 
gasoline and diesel vehicles, by model year (ranging from MY 1975 to 
2050), and by vehicle age (ranging from zero to 39 years old). The 
structure of criteria pollutant emission standards is such that these 
factors do not vary with fuel economy unless a change in fuel type 
(e.g., from gasoline to electricity) is involved.
    Since tailpipe sulfur dioxide (SO2) emissions are 
dependent on the sulfur content of the fuel, a single SO2 
emission factor in grams per million British thermal units (MMBTU) of 
fuel consumed is applied respectively for gasoline, diesel, and ethanol 
(E85) across all model years after MY 2017 based on a longitudinal 
analysis in MOVES.
    As previously mentioned, EDF submitted supplemental comments on 
SO2 emissions, stating that ``SO2 emissions 
should be proportional to fuel consumption'' and ``that the tailpipe 
SO2 emissions by calendar year from the Volpe Model do not 
change proportionally to the changes in fuel consumption across various 
CO2 control scenarios.'' \2385\ The version of the model 
supporting the 2012 final rule calculated tailpipe SO2 
emissions on a gram per gallon basis. Supporting the ensuing rulemaking 
regarding heavy-duty pickups and vans, and the 2016 draft TAR, EPA 
staff provided SO2 emission factors specified on a gram per 
mile basis. DOT modified the model in order to apply these 
SO2 emission factors as provided by EPA. The CAFE Model 
documentation released with the NPRM clearly describes how the agencies 
calculated emissions in the model. Although the version of model 
applied for the NPRM did not change this approach to calculating 
tailpipe SO2 emissions, the agencies agree that 
SO2 emissions should be proportional to fuel consumption, 
and DOT has revised the model accordingly. For SO2 
emissions, the inputs to the model include the number of grams of 
SO2 emitted by a vehicle per gallon of fuel consumed by the 
vehicle.
---------------------------------------------------------------------------

    \2385\ EDF, NHTSA-2018-0067-12363.
---------------------------------------------------------------------------

    The agencies also received comments on the use of MOVES. Most 
notably, the National Farmers Union stated ``Concerns have been raised 
regarding the models used by EPA to determine emissions from fuels. 
Third-party reviews have shown that MOVES2014 may be inadequate as a 
tool for estimating the exhaust emissions of gasoline blends containing 
more than 10 percent ethanol. The model's results for mid-level ethanol 
blends have been shown to be inconsistent with other results from the 
scientific literature for both exhaust emissions and evaporative 
emissions, including results from real-world emissions testing.'' 
\2386\ The agencies considered comments on the use of MOVES and ethanol 
blends and notes that MOVES may be unreliable for fuel blends over E10; 
however, MOVES is not designed to model mid-level ethanol blends. 
MOVES2014 is designed to model ethanol volumes up to 15 percent (E0 to 
E15), and it can also model E85 (ethanol volumes of 70 to 85 percent), 
but MOVES2014 is not designed to model intermediate fuel blends. 
Moreover, the agencies did not explicitly consider blends above E10 as 
part of the analysis, but rather ethanol blending is considered in 
relation to how to achieve a higher octane level and a higher anti-
known index.
---------------------------------------------------------------------------

    \2386\ National Farmers Union, NHTSA-2018-0067-11972.
---------------------------------------------------------------------------

    The Pennsylvania Department of Environmental Protection stated that 
there may be a significant State-specific rebound effect in 
Pennsylvania given Pennsylvania's regional role in natural gas and 
petroleum processing and refining. According to this commenter, the 
proposed rule does not adequately take into account significant local, 
State, and regional air quality impacts because it dilutes the 
emissions impact of the rule across the entire Nation. The Center for 
Biological Diversity, the Consumer Federation of America, and other 
commenters expressed concern that the proposed rule would increase 
criteria pollutants in areas with large minority populations, 
especially those in areas near oil refineries.
    Results of these tailpipe emissions calculations are summarized 
below in Section VII and in the FRIA accompanying today's notice, and 
presented in greater detail in the accompanying Final EIS.
(b) Upstream Emission Factors
    Fuel cycle emissions occurring between the extraction well and the 
fueling pump are often called upstream emissions. This rule has drawn 
upstream emission factors exclusively from the Greenhouse gases, 
Regulated Emissions, and Energy use in Transportation (GREET) model, 
developed by the U.S. Department of Energy's Argonne National 
Laboratory. The upstream gasoline, diesel, and electricity emission 
factors for criteria pollutants--namely, CO, VOC, NOX, 
PM2.5, and SO2--and greenhouse gases--namely, 
CO2, CH4, and N2O--have been updated 
with GREET 2018 data. The upstream emission factors for the air toxics 
mentioned above were unchanged from the proposal. For the final rule, 
upstream emission factors cover the following analysis years, 2017, 
2020, 2025, 2030, 2035, 2040, 2045, and 2050, and four distinct 
upstream processes:
     Petroleum Extraction,
     Petroleum Transportation,
     Petroleum Refining, and
     Fuel Transportation, Storage, and Distribution (TS&D).
    These upstream emission factors for each fuel type and analysis 
year were generated by a process using emission factor values found in 
the GREET 2018 spreadsheet tool and adjustment factors where 
appropriate. Emission factors for the petroleum extraction process are 
the aggregation of different crude feedstock--such as crude oil, oil 
sands, and shale oil--emission factors multiplied by their associated 
adjustments for transportation to refineries losses, storage losses, 
and energy share by crude feedstock. Emission factors for the petroleum 
transportation process are emissions by crude feedstock sources--such 
as crude oil fields, surface and in-situ mining, and shale reserves--
and multiplied the associated energy shares. Emission factors for the 
petroleum refining are the sum of the crude input, combustion, and non-
combustion products multiplied by the transportation of blended fuel 
loss factors. The refining emission factors applies a non-ethanol 
energy content adjustment for gasoline, blended at E10. Diesel does not 
have any such ethanol content adjustment. Emission factors for the Fuel 
TS&D process are based on the blended fuel transportation and 
distribution emissions as well as an energy content factor for both the 
petroleum and ethanol portions of the fuels. Again, diesel does not 
have an ethanol adjustment.
    The aggregated upstream emission factors used in the rule are 
aggregated across the four processes for each fuel type and analysis 
year. The aggregated upstream emission factor in the sum of the fuel 
TS&D emission factor, the petroleum refining emission factor multiplied 
by the share of fuel savings leading to reduced domestic refining, the 
pair of petroleum extraction and transportation emission factors 
multiplied by both the share of fuel savings and the share of reduced 
domestic refining from domestic crude. The upstream adjustments are 
replicated from the proposal.
    Finally, the upstream emission factors for electricity are also 
updated with GREET 2018 data. Upstream electricity emissions factors 
are derived from

[[Page 24875]]

electricity for transportation use feedstock and fuel emissions by 
analysis year. As the analysis supporting the proposal noted, there are 
three possible supply ``pathways'' for fuel consumed by the U.S. light-
duty vehicle fleet:
    1. Importing fuel that has been refined overseas into the U.S.
    2. Refining fuel in the U.S. from crude petroleum produced overseas 
and imported into the U.S.
    3. Refining fuel in the U.S. from crude petroleum produced in the 
U.S.\2387\
---------------------------------------------------------------------------

    \2387\ The proposal assumed that all fuel refined outside the 
U.S. and then imported into the U.S. would be refined from petroleum 
that was also produced outside the U.S. Although some of it could be 
refined from crude petroleum produced in the U.S. and exported, the 
analysis assumed that the fraction supplied via this pathway is 
negligible.
---------------------------------------------------------------------------

    The distribution of fuel consumed within the U.S. that is supplied 
via each of these pathways has important implications for domestic 
``upstream'' emissions, because each pathway produces domestic 
emissions arising from a different combination of activities that occur 
within the U.S. For example, pathway 1 involves domestic emissions that 
occur during crude petroleum extraction, transportation of crude oil 
from production or nearby temporary storage facilities to domestic 
refineries, refining of crude petroleum to produce transportation 
fuels, and storage and distribution of refined fuels.\2388\ In 
contrast, pathway 2 generates domestic emissions during transportation 
of crude petroleum from U.S. coastal ports to domestic refineries, as 
well as from fuel refining, storage, and distribution, while pathway 3 
produces domestic emissions only from storage and distribution of 
refined fuel.
---------------------------------------------------------------------------

    \2388\ By longstanding EPA convention, emissions that occur when 
vehicles are being refueled at retail stations or vehicle storage 
depots (such as buses) are ascribed to vehicle use, rather than to 
fuel supply.
---------------------------------------------------------------------------

    The analysis supporting the proposal made two central assumptions 
in estimating upstream emissions from fuel supply. First, 50 percent of 
any change in domestic fuel consumption by cars and light trucks 
operating on petroleum-based liquid fuels (gasoline and diesel) would 
be reflected in changes in imports of refined fuel, while the remaining 
50 percent would be reflected in changes in the volume of those fuels 
refined domestically. Second, 90 percent of any change in the volume of 
fuel refined domestically was assumed to be reflected in changes in the 
volume of crude petroleum imported into the U.S, with the remaining 10 
percent reflected in changes in the volume of crude petroleum produced 
within the U.S. The agencies developed these assumptions to analyze the 
environmental impacts of alternative CAFE and CO2 standards 
for model years 2012-2016, and have continued to rely in their analyses 
supporting subsequent rules.
    To illustrate the effect of these assumptions, for each increase in 
domestic fuel consumption of 100 gallons, 50 additional gallons would 
be supplied via pathway 1 (refined outside the U.S. and imported in 
already-refined form). Additional fuel supplied via pathway 2 (U.S. 
domestic refining of imported crude oil) would account for 90 percent 
of the remaining 50 gallons of increased consumption, or 45 gallons. 
Finally, the remaining 5 gallons of increased fuel consumed within the 
U.S. would be supplied via pathway 3 (domestic refining of crude oil 
produced within the U.S.). This same breakdown was applied to changes 
in fuel consumption estimated to occur throughout the analysis period 
used for the proposal, which extended from 2017 through 2050.
    The agencies estimated the resulting changes in upstream emissions 
of criteria air pollutants and airborne toxics occurring within the 
U.S. by applying emission factors for the appropriate stages of the 
fuel supply chain (petroleum extraction, petroleum transportation to 
refineries, fuel refining, and fuel storage and distribution) to the 
changes in the total energy content of fuel supplied by each pathway, 
and summed the results.\2389\ The energy content of fuel rather than 
its volume was used as the basis for estimating emissions, because 
emission factors are typically expressed in mass per unit of fuel 
energy supplied--for example, grams per million Btu--rather than per 
unit volume of fuel supplied.
---------------------------------------------------------------------------

    \2389\ Increases in upstream GHG emissions were calculated from 
the increase in U.S. domestic fuel consumption, without regard to 
whether they occurred within the U.S.
---------------------------------------------------------------------------

    In the proposal, the agencies made no explicit assumptions about 
the future mix of electric generating capacity that would be used to 
supply increased electricity consumed by BEVs and PHEVs. Instead, the 
agencies implicitly relied on the assumptions about future evolution of 
the nationwide mix of generation sources that were reflected in the 
U.S. average emission factors for electricity produced to power 
transportation vehicles, including cars and light trucks, which as 
described previously were drawn from the most recent version of Argonne 
National Laboratory's GREET model that was available at the time of the 
proposal. These assumptions were consistent with those made by EIA in 
its AEO 2017 Reference case analysis and publications.\2390\
---------------------------------------------------------------------------

    \2390\ https://greet.es.anl.gov/publication-greet-2017-summary.
---------------------------------------------------------------------------

    While the agencies' use of these assumptions to estimate upstream 
emissions did not prompt widespread comments on their analyses in 
support of previous CAFE rulemakings, the more recent proposal did draw 
a large number of comments focusing on those same assumptions. Most 
commenters asserted that the entirety of any increase in consumption of 
petroleum-based fuels by cars and light trucks resulting from the 
proposal would be met via increased domestic refining, primarily from 
crude petroleum produced in the U.S., and would thus generate 
additional upstream emissions within the U.S. throughout the fuel 
supply process. Even some commenters who argued elsewhere that the U.S. 
would continue to be a large-scale importer of petroleum asserted that 
the entire increase in fuel consumption resulting from the proposal 
would be refined from additional domestically-produced petroleum.\2391\
---------------------------------------------------------------------------

    \2391\ For example, IPI notes that AEO 2019 shows the U.S. will 
continue to import crude petroleum through 2050, and will remain a 
net importer as measured by the energy content rather than the 
volume of U.S. petroleum exports and imports; see IPI, NHTSA-2018-
0067-12213. Similarly, EDF argued that because U.S. petroleum 
imports have been declining and gasoline imports are currently low, 
the best assumption was that the entire increase in gasoline 
consumption resulting from the proposal would be supplied from 
increased domestic refining of U.S.-produced crude petroleum; see 
EDF, NHTSA-2018-0067-12108.
---------------------------------------------------------------------------

    As a consequence, most commenters argued that the agencies' 
analysis of the proposal significantly underestimated the increases in 
upstream emissions that were likely to result, with some also asserting 
that the increases in emissions of criteria air pollutants would cause 
potentially serious degradation of air quality in the areas surrounding 
U.S. refineries. For example, EDF stated, ``NHTSA assumed that 50% of 
all the gasoline saved by more stringent CAFE and CO2 
standards would have been imported (i.e., refined overseas). . . . It 
is difficult to see how this could be the case when the nation is 
producing enough crude oil to be a net exporter. It is also difficult 
to see how this could be the case when gasoline consumption is 
decreasing and sufficient domestic refining capacity exists to fulfill 
today's demand, let alone decreased demand in the future. . . . 
Assuming that 100% of the differences in gasoline consumption between 
control scenarios will be refined in the U.S. appears to be much more 
consistent with the available data. Likewise, it seems reasonable to 
assume

[[Page 24876]]

that differences in the crude oil requirements of the various scenarios 
will also affect domestic production more so than imports.'' \2392\
---------------------------------------------------------------------------

    \2392\ EDF, NHTSA-2018-0067-12108, p. 53. Others making similar 
assertions include IPI, NHTSA-2018-0067-12213, p. 5.
---------------------------------------------------------------------------

    However, one commenter did agree with the agencies' assessment of 
the proposal's likely impact on U.S. petroleum imports, noting that 
``Through 2050, there will only be a small increase in domestic oil 
production due to increased demand, well under 1%. . . . The vast 
majority (88% through 2050) of the additional petroleum that will be 
required to fuel light-duty vehicles in the proposed case will be 
imported. This assessment is not too far off of a single comment in the 
NPRM, `Using NEMS, it was estimated that 50% of increased gasoline 
consumption would be supplied by increased domestic refining and that 
90% of this additional refining would use imported crude petroleum.' '' 
\2393\
---------------------------------------------------------------------------

    \2393\ David Gohlke, EPA-HQ-OAR-2018-0283-5082, p. 1.
---------------------------------------------------------------------------

    The agencies note that there seems to be considerable confusion 
among commenters about the agencies' assumptions regarding import 
shares, and what they are attempting to measure. The agencies' 
assumptions are intended to measure the effects of changes in 
consumption of petroleum-derived transportation fuels by cars and light 
trucks that are attributable to this final rule on changes in U.S. 
production and imports of crude petroleum, in domestic refining of 
crude petroleum to produce transportation fuels, and in the volume of 
refined fuel distributed for domestic consumption. While recent data on 
U.S. fuel consumption, domestic production and imports of crude 
petroleum, and imports of refined petroleum products may be useful in 
estimating these desired measures, they are not themselves measures of 
the marginal impacts of changes in fuel consumption on the volumes of 
fuel supplied via each of the supply pathways described previously.
    Instead, the agencies rely on two types of information to estimate 
the current and likely future values of the desired measures. First, 
they examine recent changes in domestic consumption of petroleum-based 
motor fuels--particularly gasoline, since it is the primary fuel used 
by vehicles that are subject to CAFE and CO2 standards--and 
compare them to the accompanying changes in the three gasoline supply 
pathways, namely domestic petroleum production, U.S. imports of crude 
petroleum, and U.S. imports of refined gasoline (or components that are 
blended domestically to produce gasoline). Second, the agencies examine 
differences in forecasts of U.S. petroleum production, fuel refining, 
and imports of refined fuel under alternative future scenarios that 
were included in AEO 2018 whose projections of domestic fuel 
consumption differ in ways that include alternative CAFE standards. 
While this latter approach would ideally compare scenarios that differ 
only in their assumptions about the stringency of CAFE and 
CO2 standards but are otherwise strictly comparable, such 
idealized comparisons are rarely possible because other factors almost 
always differ as well between the alternative scenarios being compared.
(i) Assumptions Used To Analyze Impacts of the Final Rule on Petroleum 
Imports and Emissions
    In response to comments, the agencies conducted a detailed 
examination of recent changes in U.S. fuel consumption, domestic fuel 
refining, and U.S. imports and exports of crude petroleum as well as 
refined fuel (primarily gasoline). This included comparing changes in 
these variables at both the national aggregate level and for three 
separate regions of the U.S. In addition, they examined differences in 
the forecast values of these variables under alternative assumptions 
about fuel economy standards, although as indicated above these 
comparisons are complicated by the fact that factors other than CAFE 
and CO2 standards also differ between these alternative 
scenarios.
    The agencies also identified a fourth ``pathway'' to supply the 
increase in U.S. gasoline consumption anticipated to result from this 
final rule. The U.S. is now a net exporter of refined gasoline (and 
products that are blended to produce gasoline), and the volume of U.S 
gasoline exports is likely to increase for at least the next two 
decades. This introduces the possibility that some--and perhaps all--of 
the anticipated increase in domestic gasoline consumption will be met 
simply by redirecting U.S. gasoline exports to serve domestic 
consumption. This additional source of supply would result in no 
increase in domestic refining activity, and thus no increase in 
emissions from refining of petroleum-based transportation fuels.\2394\
---------------------------------------------------------------------------

    \2394\ Increased domestic emissions would only occur in this 
case to the extent that domestic distribution of gasoline entailed 
higher emissions than transporting it to U.S. coastal ports for 
export.
---------------------------------------------------------------------------

    Throughout most of the past half-century, the nation has been a 
large net importer of crude petroleum, taking its price as determined 
in world markets and importing the volumes necessary to meet the 
difference between U.S. demand for refined petroleum products and 
domestic supplies. Throughout this period, the U.S. has also been 
largely self-sufficient in refining, meaning that the gap between 
domestic demand for refined products and the volumes refined from crude 
petroleum extracted within the U.S. was primarily met by domestic 
refining of imported crude petroleum, with only marginal volumes of 
gasoline and other products imported or exported. U.S. refinery 
capacity and output generally increased over this period in proportion 
to growth in domestic consumption of fuel and other products refined 
from petroleum.
    In the past decade, however, this situation has changed 
dramatically. U.S. production of crude petroleum has more than doubled 
since 2008, making the nation one of the world's largest producers, 
while net imports of crude oil and refined products have declined by 
nearly 80 percent.\2395\ Domestic gasoline consumption declined by more 
than 6 percent between 2007 and 2012, and recovered to its 2007 levels 
only as recently as 2016, remaining near or slightly below its 2016 
level since then.\2396\ As a consequence, the U.S. shifted from being a 
net importer of refined petroleum products to a net exporter in 2011, 
and has become a net exporter of gasoline and ``blending stock'' since 
2016.\2397\
---------------------------------------------------------------------------

    \2395\ These and other petroleum statistics cited here were 
calculated from data available at EIA, Petroleum and Other Liquids, 
2019, https://www.eia.gov/petroleum/data.php. U.S. production of 
crude petroleum rose from 1.83 billion barrels in 2008 to 4.01 
billion barrels in 2018, or by 119%, During that same period, net 
U.S. imports of crude petroleum and refined products declined from 
4.07 billion to 0.85 billion barrels, or by 79%. Net U.S. imports 
are the difference between the nation's total (or gross) imports 
from elsewhere in the world and the volumes it exports to other 
nations.
    \2396\ U.S. gasoline consumption declined from 3.39 billion 
barrels in 2007 to 3.18 billion barrels in 2012, or by 6.2 percent, 
rose to 3.41 billion barrels in 2016, and remained near that level 
through 2018.
    \2397\ In 2010, U.S. net imports of refined petroleum products 
were 98 million barrels, but by 2011 U.S. net exports were 160 
million barrels. U.S. net exports of refined products then increased 
steadily through 2018, reaching 1.23 billion barrels in that year. 
In 2015, U.S. net imports of gasoline and blending components 
totaled 19 million barrels, but by 2016, U.S. net exports were 20 
million barrels, and grew to 93 million barrels in 2018. Another 
recent change in petroleum markets has been the increasing 
production and trade in gasoline blendstock in domestic and 
international petroleum trade. While in earlier periods refineries 
normally produced finished gasoline and shipped it to local storage 
terminals for distribution and retailing, in recent years, 
refineries have increasingly shifted to producing standardized 
gasoline blendstocks, such as Reformulated Blendstock for Oxygenate 
Blending (or ``RBOB''), which are then shipped and blended with 
ethanol or other additives to make finished gasoline that meets 
local regulatory requirements or customer specifications. Although 
this process has clear cost and operational advantages, particularly 
with extensive geographic and seasonal variation in gasoline 
formulations, it complicates the tabulation and comparison of 
petroleum statistics. In both EIA and most international trade 
statistics, finished gasoline and blendstocks are treated as 
separate products, and as reported in EIA statistics, large volumes 
of finished gasoline are now produced from blendstocks by local 
``blenders,'' rather than by more centralized ``refiners.'' In 
addition, the volume of refinery production of gasoline and 
blendstock is now systematically lower than consumption of finished 
gasoline, because up to 10 percent of the volume of gasoline sold at 
retail can be made up of ethanol that is blended into gasoline after 
it leaves the refinery.

---------------------------------------------------------------------------

[[Page 24877]]

    Over the past decade, increased availability of crude petroleum and 
other refinery feedstocks in combination with declining gasoline 
consumption has presented U.S. refiners with a choice between 
continuing to produce gasoline at or near their capacity while boosting 
exports, or cutting back on refinery output. U.S. refiners elected not 
to cut back on their production of gasoline; instead, they actually 
increased the volume they refined. U.S. production of finished gasoline 
increased by 9 percent between 2007 and 2018.
    The excess of gasoline production resulting from increased refinery 
capacity and stable consumption has partly displaced previous gasoline 
and blendstock imports, with the remainder taking the form of increased 
U.S. exports. Thus, as Figure VI-92 below shows, the nation now has a 
capacity to produce gasoline that considerably exceeds its current 
domestic consumption. This surplus of gasoline appears likely to 
increase in coming few years, as EIA's Annual Energy Outlook 2019 
reference case (EIA, 2019) anticipates that domestic gasoline 
consumption will continue to decline until nearly 2040. Therefore, the 
U.S. seems likely to remain a net exporter of gasoline through the next 
three decades.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.513

    Although EIA's Annual Energy Outlook does not include separate 
forecasts of gasoline exports and imports, that same agency's Short 
Term Energy Outlook projects that U.S. gasoline exports will continue 
to rise through 2020 (EIA, 2019).\2398\ Combined with EIA's reference 
case forecast in the AEO 2019, the forecasts of declining U.S. gasoline 
consumption and rising net exports of refined petroleum products 
suggest that the United States will remain a growing net exporter of 
refined petroleum products--including gasoline--through nearly 2040. In 
turn, this suggests that any increase in domestic gasoline consumption 
resulting from this final rule is likely to

[[Page 24878]]

low anticipated growth in U.S. exports, rather than prompting growth in 
domestic refining and associated upstream emissions.
---------------------------------------------------------------------------

    \2398\ AEO does not forecast gasoline refining, imports, or 
exports separately, instead reporting them as part of total refined 
petroleum products.
---------------------------------------------------------------------------

    Regional patterns of U.S. gasoline consumption, refining, and trade 
also suggests that redirecting U.S. gasoline exports to domestic 
markets is likely to be an important source of additional supply to 
meet any increase in U.S. consumption stemming from this final rule. 
The nation's East Coast (which comprises the Energy Information 
Administration's Production and Distribution District 1, or PADD 1) 
currently accounts for about 32 percent of U.S. gasoline consumption, 
but has historically produced significantly less than gasoline than it 
consumes. As Figure VI-93 below shows, the gap between consumption and 
local supply within PADD1 has recently narrowed, as gasoline production 
along the East Coast has increased rapidly in recent years, while 
shipments into the region from the remainder of the U.S. and foreign 
imports (which come mostly from Canada) declined. In June 2019, 
however, press reports suggested that that one of the largest East 
Coast refineries (Philadelphia Energy Solutions, which represents some 
28 percent of East Coast refining capacity) would be closed.\2399\ At 
the same time, construction of new refineries continues to be hindered 
by the density of population concentrations and commercial development 
along the nation's East Coast, casting doubt on the potential for 
continued increases in local gasoline refining and supply within PADD 
1.
---------------------------------------------------------------------------

    \2399\ Seba, E. (2019, July 5). Philadelphia refinery closing 
reverses two years of U.S. capacity gains. Retrieved September 19, 
2019, from Reuters: https://www.reuters.com/article/us-usa-refinery-blast-capacity/philadelphia-refinery-closing-reverses-two-years-of-u-s-capacity-gains-idUSKCN1U0283.
[GRAPHIC] [TIFF OMITTED] TR30AP20.514

BILLING CODE 4910-59-C
    As a consequence, it seems likely that at least in the near term, 
any increase in gasoline consumption along the Nation's East Coast in 
response to this rule would be supplied primarily by Gulf Coast 
refineries or increased foreign imports, rather than from increased 
production in East Coast refineries. Pipelines available to transport 
refined petroleum products from Gulf Coast refineries to the East Coast 
may also face capacity limitations, in which case most of any increase 
in gasoline consumption there would need to be met by increased imports 
from abroad. Over the longer term, however, it is possible that 
increases in East Coast gasoline consumption could be met partly by 
expanded refining activity within the region.
    The West Coast, which includes Nevada and Arizona (EIA's PADD 5), 
currently accounts for 168 percent of

[[Page 24879]]

U.S. gasoline consumption. Almost all of the gasoline consumed in that 
region is also refined within it, although small volumes are shipped 
into Arizona from neighboring PADDs by pipeline, and small volumes are 
also exported to Latin America by tanker. The West Coast is relatively 
isolated from other U.S. sources of refined gasoline by long 
transportation distances and limited pipeline capacity, while import 
terminals for crude petroleum are relatively numerous, and it therefore 
appears more likely that marginal increases in gasoline consumption 
from the rule will be met from increases in local (i.e., within-PADD) 
refining. Figure VI-94 shows that this has been the case in recent 
decades, as growth in gasoline production within PADD 5 throughout that 
period has closely paralleled growth in local consumption, while net 
exports have remained minimal.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.515

    The central region of the United States (PADDs 2-4) accounts for 
the remaining 52 percent of current U.S. gasoline consumption, while 
producing about three-quarters of the nation's gasoline and blendstock. 
Although as Figure VI-95 shows the central region was a minor net 
exporter of gasoline as recently as 2007, it now exports some 800,000 
barrels per day of gasoline and blendstock, and has accounted for 
virtually all of the recent growth in U.S. exports of these two 
categories of refined products. Recent press reports indicate that 
firms are currently making significant new investments to add refining 
capacity on the Gulf Coast to process the growing supply of U.S. shale 
oil (Douglas, 2019), and with the projected future decline in U.S. 
consumption, any additional gasoline refined there is likely to 
increase U.S. exports. Thus, future increases in gasoline consumption 
in the central region of the U.S. of the magnitude likely to result 
from adopting these final standards is expected to be met by diverting 
gasoline exports to domestic consumption, even in the absence of 
additional refinery investments.

[[Page 24880]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.516

BILLING CODE 4910-59-C
    Table VI-278 below compares recent changes in gasoline consumption 
and various sources of supply for these three U.S. regions during the 
recent period (2012-18) when gasoline consumption has generally 
increased. As it shows, recent increases in consumption along the U.S. 
East Coast have been supplied by increased production within the 
region. As noted previously, however, it appears likely that production 
capacity there will contract significantly in the near term, and that 
future increases in consumption will need to be met from foreign 
imports or shipments from other U.S. regions. As the table also shows, 
recent increases in gasoline production in the Midwest and Gulf Coast 
region have been adequate to supply increased consumption within the 
region as well as major increases in foreign exports and shipments to 
other U.S. regions. Finally, increased consumption on the Nation's West 
Coast appears to have been met via a combination of increased 
production within the region and drawdowns of previously accumulated 
inventories (not shown in the table).
    At the national level, where net shipments among regions 
necessarily cancel one another (resulting in the zero entry for Net 
Receipts from Other PADDS shown in the table), recent increases in 
production have been sufficient to meet increased domestic consumption, 
while simultaneously enabling a major increase in exports. This 
suggests that from the nationwide aggregate perspective, incremental 
increases in domestic gasoline consumption resulting from this rule 
could be met by a reduction in U.S. exports of domestically-refined 
gasoline to other nations, accompanied by increases in shipments from 
the Midwest and Gulf Coast regions to the nation's East and West 
Coasts.

[[Page 24881]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.517

    To summarize, based on changes in the various sources of supply 
that have accompanied recent changes in consumption within different 
regions of the U.S., the agencies anticipate that:
     Most of any marginal increases in U.S. gasoline 
consumption resulting from this rule that occur on the East Coast of 
the U.S. is likely to be met in the near term by increased transfers 
from other regions of the U.S. or higher foreign imports, and possibly 
by expanded refining activity in the longer term;
     Most of any marginal increases in U.S. gasoline 
consumption resulting from this rule that occur on the West Coast is 
likely to be supplied by increased gasoline refining within that 
region; and
     Most or all of any marginal increase in U.S. gasoline 
consumption resulting from this rule that occurs in the Central region 
is likely to be supplied by redirecting foreign exports to supply 
markets within that region.
    With these expectations and acknowledging the uncertainty 
surrounding them, the agencies have concluded that assuming 50 percent 
of any increase in U.S. gasoline consumption will lead to increased 
domestic refining activity--and thus to increases in domestic refinery 
emissions--continues to be reasonable, and perhaps even overstates the 
expected increase in domestic refinery emissions. In particular, the 
agencies find that assuming 50 percent is more reasonable than assuming 
that either none or 100 percent of any change in gasoline consumption 
will be translated into changes in domestic gasoline refining. Thus, 
the agencies have elected to continue to employ the 50 percent 
assumption in their central analysis, and to examine the sensitivity of 
its results to varying this fraction over the entire possible range, 
from zero to 100 percent.
(ii) Changes in Crude Oil Supply to Domestic Refineries
    The agencies also re-evaluated their assumption that 90 percent of 
the increase in crude petroleum refined in the U.S. to produce 
additional gasoline consumed as a result of this rule would be imported 
from abroad (thus resulting in increased emissions for its storage at 
import terminals, and transportation to domestic refineries), while the 
remaining 10 percent would be produced domestically (thus resulting in 
emissions from its extraction, local storage, and transportation to 
U.S. refineries). As discussed in more detail below, the agencies 
conclude that domestic petroleum production responds primarily to 
technological innovations, investments in exploration and development 
of new domestic sources of oil, and variation in the world price of 
petroleum, rather than to U.S. demand for refined products such as 
gasoline. As a consequence, they conclude that any increase in gasoline 
consumption attributable to this final rule is unlikely by itself to 
have a significant effect on domestic petroleum production, and that 
their previous assumption continues to be reasonable.
    U.S. oil production is primarily a function of development 
opportunities identified during prior exploration programs, innovations 
in the technological for drilling and extracting crude petroleum, 
producer's expectations regarding future world petroleum prices, and 
the U.S. tax and regulatory situations surrounding petroleum 
exploration and production. Crude oil is a fungible, non-perishable 
commodity, and can usually be transported among local oil markets 
around the globe at some cost. As a consequence, the price of oil in a 
U.S. domestic market such as Texas is highly correlated with its price 
in markets located in Northern Europe, the Far East, and the Middle 
East.
    In contrast, U.S. gasoline consumption depends on a broad array of 
factors that overlap only partially with the determinants of U.S. crude 
petroleum production. These include domestic economic growth and its 
consequences for transportation demand, current and future vehicle fuel 
economy, gasoline prices, excise and sales taxes levied on gasoline, 
technological and cultural changes, vehicle prices, and the evolution 
of transportation systems and the built environment.
    As a consequence, changes in U.S. consumption and supply of 
petroleum products will primarily be reflected in changes the 
destinations of domestically produced and imported crude petroleum, 
rather than in changes in their production volumes. To the extent that 
changes in U.S. gasoline demand for lead to changes in the volume 
refined domestically (the subject of the previous analysis), increased 
refining activity is thus likely to be reflected in a shift in U.S. 
imports or exports of crude oil, rather than in a change in U.S. 
production of crude oil. Instead, any effect of this rule on U.S. crude 
oil production would arise primarily from the impact of increased 
domestic gasoline demand on global oil prices, which will be limited by 
the fact that U.S. gasoline demand accounts for a relatively small 
share of total global demand for petroleum products, and by

[[Page 24882]]

the response of global supply to any upward pressure on prices. Thus, 
any effect of this rule on U.S. petroleum production is likely to be 
extremely modest.\2400\
---------------------------------------------------------------------------

    \2400\ U.S. gasoline consumption currently accounts for about 9% 
of total global demand for refined petroleum products, and the AEO 
2019 reference case projects that this will decline to 6% by the 
year 2035, and remain at that level through 2050. These figures are 
calculated from AEO 2019 Reference Case, Tables 11 and 21, available 
at https://www.eia.gov/outlooks/aeo/tables_ref.php.
---------------------------------------------------------------------------

    Localized and temporary changes in domestic production might arise 
in response to capacity limitations or transportation bottlenecks 
associated with particular regions or refineries, which could 
temporarily create markets for higher-priced crude oil. However, these 
situations would normally be localized and prevail for only a limited 
time. At the same time, the effects of any change in domestic petroleum 
consumption on world oil prices would be attenuated, because as 
indicated previously the impact of increased domestic consumption would 
be felt on prices and volumes supplied in the much larger global 
petroleum market, rather than confined to the smaller U.S. market. Any 
resulting changes in global oil prices and petroleum production would 
inevitably be small when viewed on a world scale, and likely to prompt 
only minimal responses in U.S. petroleum supply.
    As one indication of the likely minimal impacts of higher U.S. 
gasoline consumption on U.S. production of crude petroleum, EIA's 
Annual Energy Outlook 2018 included a side case called ``No New 
Efficiency Requirements,'' which included a freeze on U.S. fuel economy 
standards beginning in 2020. Although this scenario does not correspond 
exactly to either the agencies' earlier proposal or this final rule, 
comparing its results to those from the AEO 2018 reference case 
illustrates the insensitivity of domestic crude oil production to 
increases in gasoline consumption, as represented in EIA's National 
Energy Modeling System (NEMS).
    Figure VI-96 below presents such a comparison, showing historical 
trends is U.S. gasoline consumption and petroleum production, and 
comparing their projected future trends in the AEO 2018 Reference Case 
and No New Efficiency Requirements alternative. As the figure 
illustrates, the large increase in U.S. gasoline consumption under the 
latter scenario relative to the Reference Case is accompanied by an 
almost indiscernible change in U.S. crude petroleum production, for 
exactly the reasons described above.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.518


[[Page 24883]]


BILLING CODE 4910-59-C
    The agencies conclude that in the context of the current global 
petroleum market, increases in U.S. gasoline demand on the scale likely 
to result from this final rule are unlikely to produce changes in the 
market that prompt a significant increase in domestic petroleum 
production. Instead, they are likely to affect mainly the destinations 
and uses of crude petroleum--including refining gasoline within the 
U.S.--that is already being supplied to the global market. As a 
consequence, the agencies have elected to retain our previous 
assumption that any increase in domestic gasoline refining that occurs 
as a consequence of adopting this final rule is unlikely by itself to 
lead to a significant increase in domestic crude oil production or in 
the associated upstream emissions. Specifically, the agencies continue 
to assume that 10 percent of any increase in domestic gasoline refining 
would utilize increased U.S. production of crude petroleum.
    The agencies chose to model upstream emissions in order to generate 
full fuel cycle emissions--using GREET for the upstream component and 
MOVES for the downstream component--because each alternative has 
varying levels of fuel consumption, and the specific gallons of 
gasoline, diesel, E85, and other fuels evaluated in today's rule will 
lead to different tailpipe and upstream emission outcomes.
    While it may be fair to characterize MOVES and GREET as partial 
equilibrium models rather than general equilibrium models, the agencies 
did not make any modifications to the MOVES or GREET emission factors 
themselves. Changes in emission results were initiated through changes 
in fleet composition or activity, especially changes in vehicle miles 
travelled as well as vehicle sales and population. Other changes were 
made to average vehicle mass and road load coefficients such as 
aerodynamic drag and rolling resistance corresponding to the various 
regulatory alternatives. Each alternative consists of a package of 
technology changes, so a particular technology change was not modeled 
alone and would need to be evaluated separately to quantify incremental 
changes. Please consult the FRIA for quantified impacts for the 
technology packages laid out by alternative.
d) How Did the Agencies Estimate and Value Health Impacts From Changes 
in Air Quality
    The agencies' analyses estimates changes in the population-wide 
incidence of selected health impacts, as well as changes in the 
aggregate monetary value of those health impacts that may occur from 
the changes in emissions of criteria air pollutants projected to result 
from this final rule and the alternative that were considered. As with 
other estimated impacts of the final rule and alternatives, these 
changes are measured from a baseline that is represented by the 
adoption of the augural CAFE standards and the extension of EPA's 
updated CO2 estimates, providing a more precise accounting 
of physical impacts and costs and benefits of the standards, and also 
directly responds to comments, as discussed below.\2401\
---------------------------------------------------------------------------

    \2401\ See EPA, Office of Air and Radiation, Office of Air 
Quality Planning and Standards, Technical Support Document, 
Estimating the Benefit per Ton of Reducing PM2.5 
Precursors from 17 Sectors, February 2018, available at https://www.epa.gov/sites/production/files/2018-02/documents/sourceapportionmentbpttsd_2018.pdf.
---------------------------------------------------------------------------

    Many commenters expressed concern over the health impacts from 
increased GHG emissions and criteria pollutants. The American Lung 
Association et al. stated ``Today, nearly 40 percent of Americans--more 
than 124 million--live in communities in nonattainment for ozone and 
particulate matter, with many residents impacted more severely by local 
pollution sources, including near-road pollution. . . . Near-road 
pollution has been found to increase asthma attacks in children, 
cardiovascular health impacts, impaired lung function and premature 
death. . . . Reducing VOC emissions will help reduce the burden of 
these carcinogens on many communities, especially those living or 
working near these roadways.'' \2402\ As discussed in this Section, the 
agencies agree with these statements and have considered health effects 
as part of the analysis for today's rule. The Institute for Policy 
Integrity stated ``the agencies fixate on alleged on-road fatality 
effects while arbitrarily ignoring the mortalities, morbidities, and 
other welfare effects associated with emissions.'' \2403\ As described 
in this Section, in the analysis for this rule, the agencies estimate 
both air quality-related fatalities and their costs, in addition to the 
agencies' analysis on vehicle-related fatalities. Many public 
commenters also expressed concern for health issues associated with 
increased pollutants and emissions over what was anticipated by the 
agencies' 2012 analysis. The agencies carefully considered these 
comments and provided additional analysis to consider health impacts, 
as described below.
---------------------------------------------------------------------------

    \2402\ American Lung Association et al., NHTSA-2018-0067-11765.
    \2403\ Institute for Policy Integrity, NHTSA-2018-0067-12213.
---------------------------------------------------------------------------

    The estimated health impacts reflect the nationwide baseline level 
of emissions of each pollutant, an assumed geographic distribution of 
increased emissions, the resulting changes in concentrations of 
criteria pollutants at various locations nationwide (some of which 
reflect accumulations of emissions, while others are chemical by-
products formed in atmospheric reactions), increased exposure of the 
U.S. population to unhealthful concentrations of each pollutant, and 
the consequences of increased exposure for the aggregate frequency of 
each health impact. The agencies' analysis assumes that the increases 
in upstream and vehicle emissions are distributed in proportion to 
current emissions associated with fuel supply and vehicle use. This is 
consistent with the way EPA estimates health impacts and health damage 
costs for the refining and on-road mobile sources sectors, since those 
are estimated by assuming an increase in emissions from those sectors 
that is distributed in proportion to current emissions from each one, 
and estimating the resulting changes in accumulations of air 
pollutants, population exposure, health impacts, and associated 
monetary value. The accompanying estimates of per-ton damage costs 
apply unit values to the increased frequency of each health effect, 
representing the dollar costs or estimated willingness-to-pay to avoid 
its occurrence, and combine the results to estimate total damage costs.
    EPA analysts utilize a large volume of underlying data, a number of 
intermediate calculations, and many simplifying assumptions to develop 
these estimates of health impacts and health damage costs per ton of 
additional emissions, and discussing these in detail is well beyond the 
scope of this rule. These underlying data, assumptions, and 
calculations are described in detail in the document that reports the 
values used for the agencies' analysis.\2404\ EPA quantifies health 
impacts and damage costs for emissions from 17 separate sectors of U.S. 
economic activity, and reports values for increases in premature 
mortality and the combined costs of damages from premature mortality 
and various other health impacts per ton of PM2.5, nitrate,

[[Page 24884]]

and sulfate emissions.\2405\ These values include high and low 
estimates of both premature mortality and health damage costs, which 
primarily reflect alternative published estimates of the premature 
mortality impact of PM2.5 emissions.\2406\ Alternative 
values are also reported for 3 percent and 7 percent discount rates; 
discounting affects the values because of the delay (or ``latency 
period'') between exposure to air pollution and the development of some 
health impacts, most notably premature deaths.
---------------------------------------------------------------------------

    \2404\ See EPA, Office of Air and Radiation, Office of Air 
Quality Planning and Standards, Technical Support Document, 
Estimating the Benefit per Ton of Reducing PM2.5 
Precursors from 17 Sectors, February 2018, available at https://www.epa.gov/sites/production/files/2018-02/documents/sourceapportionmentbpttsd_2018.pdf.
    \2405\ Premature mortality includes deaths that are estimated to 
occur before the normally expected life span of persons with 
specified demographic characteristics.
    \2406\ Estimated willingness to pay to avoid premature death 
accounts for 98% of the total health damage costs included in these 
estimates; see EPA, p. 10.
---------------------------------------------------------------------------

    The agencies' analysis uses those values for the petroleum refining 
sector (sector 15) to represent impacts resulting from emissions that 
occur during the fuel production and distribution process (upstream 
emissions), and those for the on-road mobile source sector (sector 13) 
to represent the impacts of emissions resulting from car and light 
truck use. The agencies apply EPA's estimates of per-ton increases in 
premature mortality and health damage costs for these sectors to their 
estimates of changes in nationwide total emissions of PM2.5, 
nitrogen oxides (NOx), and sulfur dioxide (SO2) from the 
fuel supply process and from car and light truck use.
    Table VI-279 and Table VI-280 below report values the agencies used 
in the estimates of premature mortality impacts and total health damage 
costs per ton of emissions to analyze the consequences of this final 
rule. The results for this analysis are provided in Section VII of this 
rule. The dollar values reported in the tables below differ slightly 
from those reported in the underlying source, because they have been 
adjusted from the 2015$ used in that source to the 2018 dollars used 
throughout this analysis. Values for intervening years were 
interpolated from those shown in the tables, and values for the year 
2030 shown in the tables were assumed to prevail for years beyond 2030. 
The agencies' central analysis of the rule uses averages of the low and 
high values shown in each table, while the low and high values 
themselves are used in the sensitivity analyses described in Section 
VII of this rule.
BILLING CODE 4910-59-P

[[Page 24885]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.519


[[Page 24886]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.520

BILLING CODE 4910-59-C
    The valuation of premature mortality effects rely on the results of 
``benefits per ton'' approach (BPT). This approach is a reduced form 
approach, which is

[[Page 24887]]

less complex than full-scale air quality modeling, requiring less 
agency resources and time. Based on EPA's work to examine reduced form 
approach, the BPT may yield estimates of PM2.5--benefits for 
the mobile sector that are as much as 10 percent greater than those 
estimated when using full air quality modeling.
    The EPA is currently working on a systematic comparison of results 
from its BPT technique and other reduced-form techniques with results 
from full-form photochemical modelling. While this analysis employed 
photochemical modeling simulations, we acknowledge that the Agency has 
elsewhere applied reduced-form techniques. The summary report from the 
``Reduced Form Tool Evaluation Project'', which has not yet been peer 
reviewed, is available on EPA's website at https://www.epa.gov/benmap/reduced-form-evaluation-project-report. Under the scenarios examined in 
that report, EPA's BPT approach in the 2012 rule (which was based off a 
2005 inventory) may yield estimates of PM2.5--benefits for 
the mobile sector that are as much as 10 percent greater than those 
estimated when using full air quality modeling. The estimate increases 
to 30 percent greater for the electricity sector. The EPA continues to 
work to develop refined reduced-form approaches for estimating 
PM2.5 benefits.
    In addition, considerable uncertainty surrounds many of the 
assumptions and other inputs used in the agencies' analysis of economic 
and environmental impacts likely to result from adopting the final 
standards, rather than ratifying the augural standards. Perhaps most 
notably, because fuel prices are inherently volatile and forecasts of 
their future level depend critically on developments in the often 
unstable and politicized global oil market, those forecasts are 
inherently uncertain, as evidenced by the fact that actual gasoline 
prices are well below those the agencies relied on in their 2012 
analysis of CAFE and CO2 standards for model years 2017-25. 
While the agencies' current analysis updates those projections to 
reflect EIA's 2019 Annual Energy Outlook, which now anticipates that 
future prices will remain well below those the agencies projected in 
their 2012 analysis, it remains possible that EIA's current forecast 
will continue to overestimate actual future prices (of course, EIA's 
current forecast could also prove to be too low, although the recent 
record suggests a larger risk that the opposite will be the case). 
Further, gasoline prices are only one of a number of assumptions about 
which the agencies have reason to be uncertain; others include the fuel 
economy and other features of car and light truck models that 
manufacturers will offer during future model years, how buyers will 
respond to changes in the features of competing models in the face of 
future fuel prices and economic conditions, and how much they (and 
subsequent owners) will ultimately drive the models they purchase over 
their lifetimes. Uncertainty about all of these factors is reflected in 
similar risks that the agencies' projections of changes in vehicle use 
and fuel consumption under the final standards will prove to be in 
error. Finally, uncertainty about the agencies' companion projections 
of those standards' impacts on PM emissions and premature mortality is 
compounded by the currently unknown effects of future control 
technologies and regulations on actual refinery and vehicle emissions, 
as well as by the sources of potential error in estimating the effects 
of changes in emissions on premature mortality discussed above. 
Although it may seem that the agencies' estimates of increases in 
premature mortality resulting from the final standards are more likely 
to be too high than too low, it is extremely difficult to anticipate 
whether this is actually the case.
    Separately, the DEIS and FEIS accompanying this rule describe that 
the BPT estimates are subject to several assumptions and uncertainties 
that make it difficult to draw conclusions about the estimated monetary 
values.\2407\ Non-exhaustively, these reasons include that estimates do 
not reflect local variability in population density, meteorology, 
exposure, baseline health incidence rates, or other local factors that 
might lead to an overestimate or underestimate of the actual benefits 
of controlling fine particulates, and that the health impact studies 
include several sources of uncertainties, including: Within-study 
variability (the precision with which a given study estimates the 
relationship between air quality changes and health impacts), across-
study variation (different published studies of the same pollutant/
health effect relationship typically do not report identical findings, 
and in some cases the differences are substantial), the application of 
concentration-response functions nationwide (does not account for any 
relationship between region and health impact to the extent that there 
is such a relationship), and extrapolation of impact functions across 
population (the agencies assumed that certain health impact functions 
applied to age ranges broader than those considered in the original 
epidemiological study).
---------------------------------------------------------------------------

    \2407\ See DEIS and FEIS at Chapter 4, Air Quality--Health 
Impacts.
---------------------------------------------------------------------------

    Full-scale photochemical modeling provides the needed spatial and 
temporal detail to more precisely estimate changes in ambient levels of 
these pollutants and their associated impacts on human health and 
welfare. This modeling provides insight into the uncertainties 
associated with the use of benefit-per-ton estimates. The agencies 
conducted a photochemical modeling analysis for the Final EIS using the 
same methods as in the previous CAFE Final EISs 2408 2409 
and the HD Fuel Efficiency Standards Phases 1 and 2 Final 
EISs.2410 2411 The air quality modeling and health effects 
analysis focused on ozone and fine particulate matter equal to or less 
than 2.5 microns in diameter (PM2.5). As indicated in the 
Draft EIS, the agencies performed photochemical air quality modeling 
based on the inputs and emissions forecasts used in the Draft EIS. 
Consistent with prior rulemakings and as described in the scoping 
notice, to accommodate the substantial time required to complete the 
air quality modeling analysis, NHTSA proposed to initiate air quality 
modeling before the inputs and emissions forecasts for the Final EIS 
were finalized.\2412\ NHTSA received no public comments in response to 
the scoping notice addressing this analytical approach, and the agency 
proceeded accordingly. Therefore, NHTSA used the inputs and emissions 
forecasts for the Proposed Action and alternatives as stated in the 
Draft EIS for the analysis in this final rulemaking. For additional

[[Page 24888]]

information on the scoping notice and comments received, see Section X.
---------------------------------------------------------------------------

    \2408\ NHTSA (2010). Final Environmental Impact Statement, 
Corporate Average Fuel Economy Standards, Passenger Cars and Light 
Trucks, Model Years 2012-2016. Washington, DC, National Highway 
Traffic Safety Administration.
    \2409\ NHTSA (2012). Final Environmental Impact Statement, 
Corporate Average Fuel Economy Standards Passenger Cars and Light 
Trucks, Model Years 2017-2025, Docket No. NHTSA-2011-0056. July 
2012. Available at: https://one.nhtsa.gov/Laws-&-Regulations/CAFE-%E2%80%93-Fuel-Economy/Environmental-Impact-Statement-for-CAFE-Standards,-2017%E2%80%93202.
    \2410\ NHTSA (2011). Final Environmental Impact Statement, 
Medium and Heavy-Duty Fuel Efficiency Improvement Program. 
Washington, DC, National Highway Traffic Safety Administration.
    \2411\ NHTSA (2016). Phase 2 Fuel Efficiency Standards for 
Medium- and Heavy-Duty Engines and Vehicles. Final Environmental 
Impact Statement. Available at: https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/mdhd2-final-eis.pdf.
    \2412\ NHTSA, ``Notice of Intent to Prepare an Environmental 
Impact Statement for Model Year 2022-2025 Corporate Average Fuel 
Economy Standards,'' 82 FR 34740, 34743 fn. 15 (Jul. 26, 2017).
---------------------------------------------------------------------------

    Some stakeholders submitted comments about the agencies' use of 
underlying NPRM modeling to conduct the photochemical modeling; for 
example, NCDEQ recognized the agencies statement that there was not 
sufficient time to collect the modeling, but stated that they 
``strongly believe that the inputs and results should be readily 
available for public comment before the EIS and rulemaking are 
finalized.'' \2413\ Those comments are addressed in Section X and in 
the FEIS accompanying this rule. As part of EDF's alternative 
examination of the CAFE model and inputs, EDF utilized the same EPA 
benefit-per-ton method the agencies utilized for the final rule 
(discussed further below) to estimate health effects due to criteria 
pollutant emissions, concluding that the proposal would increase 
premature mortality due to increases in particulate matter emissions. 
EDF stated that these results indicated that the potential impacts of 
the rule are large, and accordingly, ``NHTSA and EPA must conduct 
detailed and thorough emission, photochemical and health effects 
modeling to quantify the effect of this or any other proposal to relax 
the CAFE and CO2 standards and increase upstream 
emissions.'' \2414\
---------------------------------------------------------------------------

    \2413\ North Carolina Department of Environmental Quality, 
NHTSA-2018-0067-12025.
    \2414\ Environmental Defense Fund, NHTSA-2018-0067-12108.
---------------------------------------------------------------------------

    The agencies estimated air quality changes and health-related 
benefits at the national scale based on a detailed analysis of air 
quality and health effects throughout the contiguous 48 states. 
Different regions of the country could experience either a net increase 
or a net decrease in emissions because of the rule, depending on the 
relative magnitude of the changes in emissions from decreased fuel 
economy, decreased vehicle use, and increased fuel production and 
distribution under each alternative. The EIS air quality analysis 
addresses regional differences using grid-based air quality modeling 
and analysis techniques, which account for local and regional 
differences in emissions and many of the other factors (such as 
meteorology and atmospheric processes) that affect air quality and the 
resulting health effects at any given location. This air quality 
modeling analysis is intended as a screening application of both the 
Community Multiscale Air Quality (CMAQ) model and the Environmental 
Benefits Mapping and Analysis Program (BenMAP) tool for the purposes of 
quantifying and comparing the air quality and health-related benefits.
    To examine and quantify the air quality and health-related benefits 
associated with implementing the final CAFE standards for MY 2021-2026 
light-duty vehicles, the agencies performed a national-scale 
photochemical air quality modeling and health benefit assessment with 
the following key steps:
     Preparing emission inventories.
     Modeling air quality.
     Assessing air quality-related health impacts.
    The following widely used tools were used for the air quality and 
health effects assessment:
     Sparse-Matrix Operator Kernel Emissions (SMOKE) processing 
tool (version 3.7) to prepare model-ready emissions.
     Community Multiscale Air Quality (CMAQ) model (version 
5.2.1) to quantify air quality changes for the different fuel economy 
alternatives.
     Environmental Benefits Mapping and Analysis Program--
Community Edition (BenMAP-CE) tool (version 1.4) to assess the health-
related impacts of the simulated changes in air quality.
    The national-scale modeling analysis employed the standard CMAQ 
continental modeling domain. The horizontal resolution of the grid for 
this modeling domain is 36 kilometers (22.4 miles). Air quality and 
health-related impacts were calculated for each grid cell in the entire 
contiguous United States (48 states). Although the modeling domain does 
not include all 50 states, nearly all of the affected emissions and 
population are included in the domain; therefore, the results are 
expected to represent those for a national-scale analysis. The agencies 
applied the CMAQ model for an annual simulation period using 
meteorological inputs for a base year of 2011.
    The agencies performed modeling for 2035 (although the emission 
inputs represented a variety of different projection years, including 
2030, 2035, and 2040, based on best available data). As in the Draft 
EIS, the agencies chose 2035 for analysis of the various fuel economy 
alternatives because a large proportion of vehicles in operation are 
expected to meet the level of the standards set forth by 2035. EPA 
provided up-to-date, projected, national-scale emissions data for 2040 
for motor vehicles and for 2030 for all other sources. The emissions 
were processed for the 36-kilometer (22.4-mile) resolution modeling 
domain using SMOKE. The resulting model-ready inventories contain 
emissions for all criteria pollutants (as required for photochemical 
modeling) for multiple source categories (sectors), including on-road 
mobile sources, non-road mobile sources (e.g., construction equipment, 
locomotives, ships, and aircraft), electric generating unit (EGU) point 
sources, non-EGU point sources, area sources, and biogenic sources.
    Following preparation of baseline emissions inventories, the 
baseline emissions for the light-duty vehicle portion of the on-road 
mobile emissions and the relevant upstream categories were replaced 
with data reflecting the alternatives analyzed in the Draft EIS. As 
discussed above, NHTSA calculated national estimates of on-road 
emissions for these vehicle classes for 2035, including both downstream 
and upstream emissions.
    The agencies then applied CMAQ, using the emissions specific to 
each alternative. The simulated difference in air quality between the 
Draft EIS No Action Alternative and each action alternative represents 
the change in air quality associated with that alternative. Following 
the application of CMAQ, the agencies processed the CMAQ outputs for 
input to the BenMAP-CE health effects analysis tool, and used BenMAP-CE 
to estimate the health impacts and monetized health-related benefits 
associated with the changes in air quality simulated by CMAQ for each 
of the action alternatives. The BenMAP-CE tool includes health impact 
functions, which relate a change in the concentration of a pollutant 
with a change in the incidence of a health endpoint. BenMAP-CE also 
calculates the economic value of health impacts. For this study, the 
health effects analysis considered the effects of ozone and 
PM2.5. The PM2.5 analysis includes sulfate and 
nitrate particulates (secondary PM2.5) formed from emissions 
of SO2 (sulfur dioxide) and NOX, respectively. 
BenMAP-CE does not estimate health impacts associated with changes in 
directly emitted sulfur dioxide (SO2), carbon monoxide (CO), 
and other emissions. Health effects were calculated at the 36-kilometer 
scale (grid cell size) and aggregated nationally to determine overall 
impact.
    Figure VI-97 shows the components of the air quality modeling and 
health-related benefits analysis. Note that both the emissions and 
meteorological inputs are used by SMOKE.

[[Page 24889]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.521

    Discussion of the photochemical modeling results is presented in 
the FEIS accompanying this final rule.

E. Compliance Example Walk-Through

    To illustrate the CAFE model's simulation of a manufacturer's 
potential response to fuel prices and new standards, the NPRM provided 
an example of how the preliminary version of the model showed, on a 
year-by-year basis, how GM could potentially respond to a set of CAFE 
standards, starting from MY 2016 (the latest year for which the 
agencies were able to develop a full and detailed characterization of 
the fleet of vehicles produced for sale in the U.S. at the time of 
publishing the NPRM). Although no analysis that does not rely heavily 
on a manufacturer's confidential product planning information can, with 
high fidelity, predict what that manufacturer will do, the CAFE model, 
by realistically reflecting product planning considerations in a 
detailed year-by-year context, can describe a course that manufacturer 
could realistically take. Indeed, when manufacturers provide 
information to the agencies, they often emphasize year-by-year plans. 
Although such information is typically considered confidential business 
information (CBI), public comments by the Alliance illustrate the 
concept for a hypothetical manufacturer. Although the illustration 
includes credit carry-back (aka borrowing) that most manufacturers have 
a history of avoiding, the illustration clearly demonstrates that the 
Alliance views product planning as a year-by-year exercise:
BILLING CODE 4910-59-P

[[Page 24890]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.522

BILLING CODE 4910-59-C
    Like the peer reviewers who examined the model's simulation of 
technology application and compliance, automakers have been widely 
supportive of the CAFE model's approach of year-by-year analysis 
informed by product planning realities. For example, Toyota commented, 
``The preamble correctly notes that manufacturers try to keep costs 
down by applying most major changes mainly during vehicle redesigns and 
more modest changes during product refresh, and that redesign cycles 
for vehicle models can range from six to ten years, and eight to ten-
years for powertrains. . . This appreciation for standard business 
practice enables the modeling to capture more accurately the way 
vehicles share engines, transmissions, and platforms. There are now 
more realistic limits placed on the number of engines and transmissions 
in a powertrain portfolio which better recognizes manufacturers must 
manage limited engineering resources and control supplier, production, 
and service costs.'' \2416\
---------------------------------------------------------------------------

    \2415\ NHTSA-2018-0067-12073, at 28.
    \2416\ NHTSA-2018-0067-12098, at 6.
---------------------------------------------------------------------------

    The CAFE model's year-by-year approach to estimating manufacturers' 
potential responses to standards and fuel prices is consistent with 
EPCA/EISA's requirement that CAFE standards be set at the maximum 
feasible levels for each fleet (passenger car and light truck) in each 
model year. Some commenters correctly observe that the CAA (which 
provides no direction regarding tailpipe CO2 emissions 
standards) does not require such a year-by-year determination, but 
suggest, further, that EPA should refrain from making use of year-by-
year analysis. In particular, CBD et. al. commented as follows:

    Furthermore, the Volpe model and association [sic] tools are not 
designed in accordance with EPA's independent statutory authority 
under Clean Air Act Section 202. The Volpe and OMEGA models have an 
overarching difference in their architecture--one where the Volpe 
modeling approach is designed to match NHTSA's statutory authority, 
but not EPA's. The EPCA requirements drive the design of the Volpe 
model, in that it performs a year-by-year analysis in order to 
demonstrate that NHTSA is meeting its EPCA obligations. As a result, 
the Volpe model attempts to simulate for each manufacturer, by year, 
their refresh and redesign cadence across their vehicle platforms 
and then predict a manufacturer's technology deployment decision-
making process for each platform. But under the Clean Air Act, EPA 
is not required to demonstrate that standards are set at the maximum 
feasible level year-by-year, as EPCA explicitly requires for 
NHTSA.\2417\
---------------------------------------------------------------------------

    \2417\ NHTSA-2018-0067-12000, Appendix A, at 24-25.

    Although CBD is correct that the CAA does not require a year-by-
year determination or year-by-year analysis, CBD wrongly claims that 
the CAFE model's modeling approach is not ``in accordance'' with the 
CAA. CBD's claim is analogous to saying ``just say you want to drive 
across the country; don't bother looking at a map.'' As the NPRM 
demonstrated, the CAFE model can be used to simulate compliance with 
CO2 standards. That the model follows a year-by-year 
approach to doing so simply means that it takes greater pains to 
describe realistic pathways forward from a known model year. 
Manufacturers are by no means the only stakeholders to recognize that 
product planning is actually a year-by-year process. Supporting its 
comments on the agencies' proposal, CARB provided a study by Roush 
Industries, focusing on a potential design pathway for the Toyota 
RAV4.\2418\ While this report, which was cited by CARB in its comments, 
asserted the agencies' modeling underestimated fuel consumption 
benefits and overestimated costs, Roush, like the Alliance, clearly 
interpreted the question of realism as a

[[Page 24891]]

year-by-year question, as illustrated by the following chart in Roush's 
report:
---------------------------------------------------------------------------

    \2418\ Rogers, G., ``Technical Review of: The Safer Affordable 
Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-2026 
Passenger Cars and Light Trucks, Final Report.'' Roush Industries. 
October 25, 2018. See CARB, NHTSA-2018-0067-11984.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.523

BILLING CODE 4910-59-C
    While a year-by-year representation is essential to the estimation 
of pathways that individual manufacturers could realistically take to 
apply technologies to specific vehicle models, the CAFE model also 
accounts for a range of other important engineering and product 
planning considerations. For example, among specific vehicle models, 
engines and transmissions are often shared, and a given vehicle design 
platform may encompass a range of different specific vehicle models. 
This means not every configuration of every vehicle model can be as 
optimized for fuel economy as if each could be considered in isolation. 
This isn't to say that such optimization is technologically impossible, 
but rather to say that the resources involved in such optimization 
would be financially impracticable. Moreover, CAFE and CO2 
standards apply to fleets, not specific products. This means, for 
example, that if a given engine is shared among both passenger cars and 
light trucks, changes made to that engine in response to one fleet's 
standard will impact products in the other fleet. Consistent with the 
fact that CAFE and CO2 compliance applies to fleets on a 
year-by-year basis, the CAFE model explicitly accounts for sharing 
among specific model/configurations when simulating year-by-year 
compliance. The Roush report's authors ``have not performed a complete 
fleet-compliance simulation.'' \2420\ Therefore, even notwithstanding 
differences in estimates of redesign schedules and technology efficacy 
and costs, Roush's analysis of the RAV4 is highly idealized. As 
discussed below, together with inputs based on Toyota's actual MY 2017 
production, the CAFE model represents the RAV4 as encompassing multiple 
configurations, spanning both the passenger car and light truck 
regulatory classes, all on a common vehicle platform that includes 
several other vehicle models, and some RAV4s sharing engines with some 
Camrys. Compared to estimating the potential to apply technology to a 
handful of specific model/configurations in isolation, analysis that 
accounts for manufacturers' actual production considerations produces 
more realistic results.
---------------------------------------------------------------------------

    \2419\ Rogers, G., ``Technical Review of: The Safer Affordable 
Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-2026 
Passenger Cars and Light Trucks, Final Report,'' at 26. Roush 
Industries. October 25, 2018. See CARB, NHTSA-2018-0067-11984.
    \2420\ Ibid. at 6.
---------------------------------------------------------------------------

    Nothing about the CAA discourages realism in regulatory analysis, 
and even if the CAA did so, the CAFE model can easily be run for 
isolated model years, or run in a manner that otherwise ignores 
practical limits on development and manufacturing complexity.\2421\ EPA 
elected to use the CAFE model as designed because doing so produces a 
more realistic basis to estimate regulatory impacts. EPA considers its 
use of the CAFE model entirely consistent with all CAA and other 
statutory and other requirements governing the agency's development of

[[Page 24892]]

motor vehicle CO2 emissions standards which, unlike criteria 
pollutant standards, are specified on a year-by-year basis, and 
inherently involve the entirety of manufacturers' vehicles and fleets.
---------------------------------------------------------------------------

    \2421\ Idealized simulation of compliance with a hypothetically 
isolated model year could be accomplished by, when running the 
model, setting the various ``start'' and ``end'' years to the same 
value. Sharing of engines and transmission among different model/
configurations could be ignored by, in the CAFE model's ``market'' 
input file, assigning each engine, transmission, and vehicle 
platform to a single model/configuration (e.g., such that each of 
the six versions of the RAV4 is on its own vehicle platform, and 
uses a dedicated engine and transmission).
---------------------------------------------------------------------------

    Of course, like any other model, the CAFE model used for the NPRM 
had room for improvement. As discussed above, the agencies have 
responded to public comments by making changes to some aspects of the 
CAFE model itself. Only a few such changes, all of which are discussed 
above in greater detail, impact the CAFE model's simulation of 
manufacturers' application of fuel-saving technologies. Among these, 
three are especially important: First, the model now uses a more 
``open'' application of its technology ``decision trees.'' While the 
primary objective of this change is to make the model's cost accounting 
more transparent (by recasting costs as absolute rather than 
incremental), it also makes the model somewhat more likely to identify 
and apply any highly cost-effective yet comparatively ``advanced'' 
combinations of technology. Second, the model introduces a ``cost per 
credit'' metric for comparing available opportunities to add specific 
technologies to specific vehicles.\2422\ As discussed above and in the 
summary of the sensitivity analysis conducted for today's notice, 
changing from the NPRM's ``effective cost'' metric to this new ``cost 
per credit'' metric leads the model to, at least for the combination of 
inputs in today's central analysis, more frequently select less costly 
technology pathways than more costly pathways, at least when simulating 
compliance with CO2 standards. Third, the CAFE model can now 
extend its explicit simulation of manufacturers' technology application 
well into the future. Today's analysis extends this explicit simulation 
through model year 2050. Because today's reference case input estimates 
include continued increases in fuel prices alongside continued 
(``learning''-related) reductions in technology costs, extending the 
explicit simulation shows manufacturers making significant voluntary 
improvement in the longer term (e.g., after MY 2035), even if CAFE and 
CO2 remain unchanged.
---------------------------------------------------------------------------

    \2422\ Notable comments on this metric appear at NHTSA-2018-
0067-12039, Appendix, pp. 28-34, and at NHTSA-2018-0067-12108, 
Appendix B, pp. 66-70.
---------------------------------------------------------------------------

    The agencies have also revised most of the inputs to the CAFE 
model, both to respond to comments and to better reflect an ever-
changing world. Sections appearing above discuss changes to model 
inputs, such as the analysis fleet, technology-related inputs, and fuel 
prices. Many of these changes are important to the model's simulated 
application of fuel-saving technology. Updating the analysis fleet from 
a MY 2016 to a MY 2017 basis ensures that fuel economy and 
CO2 improvements manufacturers actually realized by adding 
technologies between those model years is accounted for, and ensures 
that changes in product offerings and production volumes between those 
model years are also accounted for. With this update, the agencies also 
more fully accounted for compliance credits accumulated prior to the 
MYs represented explicitly in today's analysis. Some manufacturers have 
accumulated large volumes of such credits, and are able to apply those 
credits well past MY 2016, and to trade them to other manufacturers. 
Updated vehicle simulations correct errors and make use of additional 
engine performance estimates (i.e., engine efficiency ``maps''), and 
cost estimates for some technologies reflect additional data and 
consideration of comments. Also, fuel prices in the forecast used for 
today's analysis are somewhat higher than those used for the NPRM; by 
itself, this change makes the model tend to show larger and more 
widespread voluntary fuel economy increases and accompanying 
CO2 emissions reductions, although this increased tendency 
is countered by the impact of changing to the ``cost per credit'' 
metric.
    The following example will illustrate the model's behavior when 
simulating compliance with CO2 standards. While the example 
focuses on the baseline CO2 standards and on a specific 
manufacturer (Toyota), and highlights a specific vehicle model (the 
Toyota RAV4), results for other scenarios, manufacturers, and vehicle 
models reflect application of the same logic. Because this example 
begins with the MY 2017 fleet, and does not make use of manufacturers' 
product plans (which the agencies have historically treated as 
confidential business information, today's analysis cannot and does not 
fully reflect manufacturers' actual product design decisions, even in 
the short term. Nevertheless, the analysis yields a realistic and 
detailed characterization of a path each manufacturer could take in 
response to a given set of standards and other input estimates (e.g., 
of technology costs and fuel prices).
    As discussed above, the model considers all models and model/
configurations produced for sale in the U.S. by a given manufacturer. 
The Toyota Camry and Tundra are examples of specific Toyota passenger 
car and light truck models, Toyota produces a range of configurations 
(e.g., with different engines) of each of these vehicle models, and 
inputs to the CAFE model ensure that each such configuration is 
accounted for. CAFE model output files show the progressive application 
of technology to each model/configuration over time under each 
regulatory alternative. Here, focusing on different versions of one 
model, the RAV4, illustrates the process and results.
    The RAV4 is one of the vehicle models included in a vehicle 
platform that also includes the Camry, Corolla, Prius, Lexus CT 200h, 
Lexus NX 200t, and Lexus NX 300h. As mentioned above, the CAFE model 
reflects the agencies' assumption that significant changes to vehicle 
structures or materials will most practicably be applied throughout a 
vehicle platform as models within the platform are redesigned. Within 
this platform, the CAFE model identifies the Corolla LE, at more than 
180,000 units produced in MY 2017, as the most likely ``leader'' for 
such changes. Inputs to today's analysis also show that most of the 
RAV4s produced for the U.S. in MY 2017 shared a 2.5L naturally 
aspirated 4-cylinder gasoline engine with many Camrys. The CAFE model 
identifies the Camry as the leader for new versions of that engine. The 
same inputs show many RAV4s shared a 6-speed automatic transmission 
with a range of other vehicle models, including the Avalon, Camry, 
Lexus ES 350, Highlander, Lexus NX 200t, and the CAFE model identifies 
the Camry as the most likely leader for changes to this transmission. 
Model inputs also show other RAV4s shared a different 6-speed automatic 
transmission with the Lexus NX 200t, and the CAFE model identifies the 
RAV4 as the most likely leader for changes to this transmission. 
Finally, the MY2017 RAV4 also included two ``strong'' (power split) 
hybrid-electric versions (SE and XLE). Although these shared an engine 
with other Toyota hybrids (Avalon, Camry, Lexus ES 300h and NX 300h), 
the CAFE model reflects the agencies' assumption that it could be 
practicable to ``split off'' plug-in (or fuel cell) configurations 
rather than necessarily replace all strong hybrids sharing an engine 
with PHEVs, BEVs, or FCVs.
    Inputs for today's analysis have Toyota redesigning the RAV4 every 
five years, starting with MY 2019, and freshening the model 2-3 years 
after each redesign. Given this design cycle, and all the other inputs 
to today's analysis, the CAFE model shows that under the baseline 
CO2 standards,

[[Page 24893]]

Toyota could potentially make changes to the RAV4 summarized in the 
table that follows. The first changes occur in 2019, with Toyota 
improving aerodynamics of the hybrid RAV4s, and with the conventional 
RAV4s inheriting a new high compression ratio (HCR) engine introduced 
with the MY 2018 redesign of the Camry, and also adding 8-speed 
automatic (A8) transmissions,\2423\ improved accessories (IACC), and 
tires with reduced rolling resistance (ROLL20). With the MY 2024 
redesign, all versions of the RAV4 receive further aerodynamic 
improvements (AERO20) and ``Level 1'' mass reduction, engine friction 
reduction (EFR) is applied to the HCR engine the non-hybrid versions 
share with the Camry, and secondary axle disconnect (SAX) is applied to 
the non-hybrid versions of the RAV4. With the MY 2027 freshening, 
Toyota applies low-drag brakes to all the RAV4s. The MY 2029 redesign 
does not make any powertrain changes, but applies more significant mass 
reduction (MR3) to all RAV4s. In MY 2039, Toyota replaces the hybrid 
RAV4 SE and XLE with 200-mile (BEV200) and 300-mile (BEV300) electric 
vehicle, respectively.
---------------------------------------------------------------------------

    \2423\ While it is not necessary for the compliance simulation 
to produce real predictions of manufacturer product designs, only 
plausible ones, these changes to the RAV4 did in fact occur during 
the 2019 redesign.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.524


[[Page 24894]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.525

    This progressive application of technology to the RAV4 produces a 
series of emission reductions shown in the following table (and, though 
not shown, corresponding fuel economy improvements). The table also 
shows the progression of CO2 targets for these vehicles, 
reflecting the fact that targets are higher for the hybrid and 
conventional AWD versions of the RAV4, classified as light trucks, than 
for the FWD RAV4s classified as passenger cars. Also notably, the 
conventional RAV4s never achieve their respective CO2 
emissions targets. This merely reflects the fact that credits for 
reducing A/C refrigerant leakage apply at the fleet level rather than 
on a per-vehicle basis and, in any event, Toyota can respond by 
improving CO2 levels enough among enough other vehicle 
models that Toyota's overall average CO2 levels comply with 
Toyota's overall requirements, taking into account the potential 
application of compliance credits.

[[Page 24895]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.526

    These CO2 values could be converted to equivalent fuel 
economy levels by multiplying their reciprocals by 8887 grams per 
gallon (e.g., 8887 g/gal x 1/(144 g/mi) = 62 mpg), differences in 
compliance provisions are such that results would be offset from actual 
fuel economy levels under CAFE standards. When simulating compliance 
with CAFE or CO2 standards, the CAFE model reports both fuel 
economy and CO2 targets and achieved levels, even when the 
model is ``enforcing'' compliance with only one of these sets of 
standards. When simulating

[[Page 24896]]

compliance with baseline CO2 standards, results for the 
example discussed here show the following fuel economy targets and 
achieved levels for the RAV4.
[GRAPHIC] [TIFF OMITTED] TR30AP20.527


[[Page 24897]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.528

    The progressive application of technology also produces increases 
(and some eventual decreases) in costs. For each RAV4 configuration, 
the following table shows costs beyond MY 2017 technology, in 2018 
dollars. The conventional RAV4s incur a significant cost increase in MY 
2019, primarily for the new HCR engine inherited from the Camry. Costs 
continue to increase through MY 2029 as additional technology 
accumulates, with another significant increase for MR4 in MY 2029. 
After MY 2029, technology costs for conventional RAV4s gradually 
decline through MY 2050, in response to ongoing learning. In MY 2039, 
the BEV200 RAV4 is less expensive than the HEV RAV4 it replaces, 
leading this version's cost to drop by about $500 between MY 2033 and 
MY 2034, and with learning, to fall quickly well below this version's 
MY 2017 cost. Conversely, the BEV300 RAV4 introduced in MY 2039 is 
about $950 more expensive than the MY 2038 hybrid RAV4 it replaces, and 
even with learning, the BEV300 remains more expensive through MY 2050 
than the hybrid RAV4. These BEVs are not needed for compliance; the 
model shows Toyota could introduce them because, if battery costs 
continue to decline while gasoline prices continue to increase, BEVs 
could eventually become attractive on an economic basis.

[[Page 24898]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.529


[[Page 24899]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.530

BILLING CODE 4910-59-C
    As mentioned above, by making sufficient improvements to other 
vehicle models, Toyota could refrain from making the conventional RAV4s 
meet their CO2 emissions targets. More broadly, Toyota can 
also use compliance credits to cover compliance gaps. The CAFE model 
accounts for the potential to transfer compliance credits between the 
passenger car (PC) and light truck (LT) fleets. The model also accounts 
for the potential to apply credits from prior model years (i.e., 
credits that have been ``banked'' or, equivalently, ``carried 
forward''), including compliance credits earned prior to MY 2017. These 
aspects of the model interact with the model's accounting for multiyear 
planning--that is, the potential that a manufacturer, depending on its 
product design cadence and on the progression of standards, might apply 
``extra'' technology in some model years in order to facilitate 
compliance in later model years. For example, if a manufacturer is only 
redesigning 15% of its fleet volume in MY 2025, that manufacturer might 
be best off--even setting aside credit banking--applying some ``extra'' 
technology (i.e., technology that leads to overcompliance) as part of 
vehicle redesigns planned for MYs 2018-2024, and carrying that 
technology forward into MY 2025 when there are fewer opportunities 
available to reduce CO2 emissions in new models. As shown in 
Figure VI-100, in Toyota's case, the model shows that Toyota could 
offset its light truck compliance gaps during MY 2017-2019 by applying 
compliance credits earned for light trucks prior to MY 2017. The graph 
also shows Toyota applying extra technology to its passenger car fleet 
during MYs 2018-2024 in order to comply with the MY 2025 passenger car 
standard, but also to carry forward compliance credits and use those 
credits to offset large compliance gaps for Toyota's light truck fleet 
during MYs 2023-2027. After MY 2025, the model shows the effects of 
some technology continuing to be inherited (especially during MYs 2026-
2030) from prior MYs, of Toyota continuing to make voluntary 
improvements where economically attractive (like the MY 2039 RAV4 EV 
mentioned above), and of Toyota continuing to transfer compliance 
credits from the passenger car to the light truck fleet.\2424\
---------------------------------------------------------------------------

    \2424\ While the fleets (PC and LT) are shown separately for 
compliance purposes in this example, the ability to utilize credits 
from either fleet toward total model year compliance (in the current 
year, without caps or limits) means that the fleets for a 
manufacturer comply jointly in each model year.

---------------------------------------------------------------------------

[[Page 24900]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.531

    As the above figure shows, credit banking and transfers play an 
important role in Toyota's simulated response to the standards. If 
exercised in a manner that sets aside credit banking, the CAFE model 
shows Toyota increasing its application of fuel-saving technologies 
through MY 2025, and carrying those improvements forward, such that 
Toyota's overall average CO2 emission rate is 16 g/mi lower 
in MY 2025 when credit banking is not accounted for, as illustrated by 
the next chart appearing below. Though not shown here, accounting for 
credit banking also impacts the simulation other OEMs' compliance 
pathways, because inputs to today's analysis assume that Toyota would 
likely not need to use all of its pre-2017 compliance credits before 
these credits expire in 2021, and that Toyota could therefore sell 
those older credits other manufacturers (e.g., FCA, VW). By accounting 
for credit banking, the CAFE model thereby avoids considerable 
potential understatement of future CO2 emissions from light-
duty vehicles.

[[Page 24901]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.532

    As indicated by the following chart, a failure to account for 
credit banking would also increase Toyota's modeled per-vehicle costs 
by nearly $1,000 in MY 2025. By accounting for credit banking, the CAFE 
model thus avoids considerable potential overstatement of compliance 
costs. Though not shown here, accounting for credit banking while also 
applying inputs that reflect Toyota's ability to sell older credits to 
some other OEMs also enables the CAFE model to avoid overstatement of 
compliance costs for those OEMs.

[[Page 24902]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.533

    While the model's simulation of manufacturers' potential responses 
to CAFE standards applies the same inputs and analytical methods, it 
does so accounting for several important statutory and regulatory 
differences between CO2 standards and CAFE standards, and 
for specific statutory direction regarding how CAFE standards are to be 
considered for purposes of setting standards at the maximum feasible 
levels in each model year. EPCA places specific limits on the amount of 
credit that can be transferred between fleets, and requires that 
domestic passenger cars meet minimum standards without applying 
credits. EPCA also requires that the determination of maximum feasible 
stringency set aside the potential to apply compliance credits or 
introduce new alternative fuel vehicles (include BEVs and FCVs, but not 
including plug-in HEVs) during the model years under consideration. 
Especially with standards that continue to become more stringent, 
applying these statutory constraints to the analysis leads the model to 
tend to show greater overcompliance with standards in earlier model 
years, because even setting aside the potential to carry forward or 
transfer credits, Toyota is likely to find it more practicable to apply 
some ``extra'' technology when redesigning vehicles during MYs 2017-
2024 than to attempt to address MY 2025 standards by working with only 
vehicles scheduled to be redesigned in MY 2025. The model also tends to 
show greater overcompliance in later model years, because some of that 
extra technology from years leading up to the last year of stringency 
increases takes time to carry forward to ensuing model years. These 
aspects of the CAFE ``standard setting'' analysis are evident in the 
model's solution for Toyota, shown in the following figure. With the 
use of credits set aside after MY 2020, Toyota overcomplies with light 
truck standards during MYs 2018-2023 in order to carry technology 
forward into MY 2025. Although Toyota only marginally overcomplies with 
MY 2025 standards, the inheritance of technology during MYs 2026-2029 
contributes to increased overcompliance (which is to be expected given 
the degree of platform and powertrain sharing between the fleets). 
Continued increases in overcompliance after 2030 arise due to cost 
learning effects (especially for batteries) and increased fuel prices.

[[Page 24903]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.534

VII. What Does the Analysis Show, and What Does It Mean?

A. Impacts of the Standards--Final and Alternatives

    New CAFE and CO2 standards will have a range of impacts. 
EPCA/EISA and NEPA require DOT to consider such impacts when making 
decisions about new CAFE standards, and the CAA requires EPA to do so 
when making decisions about new emissions standards. Like past 
rulemakings, today's announcement is supported by the analysis of many 
potential impacts of new standards. Today's rulemaking finalizes new 
standards through model year 2026. While the CAFE model explicitly 
estimates manufacturers responses to standards through model year 2050 
and the associated impacts through calendar year 2089, today's 
rulemaking presents estimates of impacts on model years through MY 
2029, including impacts through these vehicles' full useful lives 
(i.e., for MY 2029 vehicles, through 2068). Today's rulemaking also 
presents estimates of overall impacts in each calendar year through 
2050, accounting for all model years through 2050. The agencies of 
course do not know today what will actually come to pass decades from 
now under the new final standards or under any of alternatives under 
consideration. The analysis is intended less as a forecast, than as an 
assessment--reflecting the best judgments regarding many different 
factors--of impacts that could occur.\2425\ As discussed below, the 
analysis was conducted using several defined alternatives to explore 
the sensitivity of this assessment to a variety of potential changes in 
key analytical inputs (e.g., fuel prices).
---------------------------------------------------------------------------

    \2425\ ``Prediction is very difficult, especially if it's about 
the future.'' Attributed to Niels Bohr, Nobel Laureate in Physics.
---------------------------------------------------------------------------

    This section summarizes various impacts of the final standards and 
other regulatory alternatives defined above. The no-action alternative 
provides the baseline relative to which all impacts are shown. Because 
the final standards (and the other alternatives considered), being of a 
``deregulatory'' nature, are less stringent than the no-action 
alternative, all impacts are directionally opposite to impacts reported 
in recent CAFE and CO2 rulemakings. For example, while past 
rulemakings reported positive values for fuel consumption avoided under 
new standards, today's rulemaking reports negative values, as fuel 
consumption is expected be somewhat greater under today's new final 
standards than under standards defining the baseline no-action 
alternative. Reported negative values for avoided fuel consumption 
could also be properly interpreted as simply ``additional fuel 
consumption.'' Similarly, reported negative values for costs could be 
properly interpreted as ``avoided costs'' or ``benefits,'' and reported 
negative values for benefits could be properly interpreted as ``forgone 
benefits'' or ``costs.'' However, today's rulemaking retains reporting 
conventions consistent with past rulemakings, anticipating that, 
compared to other options, doing so will facilitate review by most 
stakeholders.
    Today's analysis presents results for individual model years in two 
different ways. The first way is similar to past rulemakings and shows 
how

[[Page 24904]]

manufacturers could respond in each model year under the new final 
standards and each alternative covering MYs 2021/2022-2026. The second, 
expanding on the information provided in past rulemakings, evaluates 
incremental impacts of new standards for each model year, in turn. In 
past rulemaking analyses, NHTSA modeled year-by-year impacts under the 
aggregation of standards applied in all model years, and EPA modeled 
manufacturers' hypothetical compliance with a single model years' 
standards in that model year. Especially considering multiyear planning 
effects, neither approach provides a clear basis to attribute impacts 
to specific standards first introduced in each of a series of model 
years. For example, of the technology manufacturers applied in MY 2017, 
some would have been applied even under the MY 2014 standards, and some 
were likely applied to position manufacturers toward compliance with 
(including credit banking to be used toward) MY 2018 standards. 
Therefore, of the impacts attributable to the model year 2017 fleet, 
only a portion can be properly attributed to the MY 2017 standards, and 
the impacts of the MY 2017 standards involve fleets leading up and 
extending well beyond MY 2016. Considering this, the final standards 
were examined on an incremental basis, modeling each new model year's 
standards over the entire span of included model years, using those 
results as a baseline relative to which to measure impacts attributable 
to the next model year's standards. For example, incremental costs 
attributable to the new standards for MY 2023 are calculated as 
follows:

COSTNew final,MY 2023 = (COSTNew final\through\MY 2023-COSTNo-
Action\through\MY 2023)-(COSTNew final\through\MY 2022-COSTNo-
Action\through\MY 2022)

where

COSTNew final,MY 2023: Incremental technology cost during MYs 2018-
2029 and attributable to the new final standards for MY 2023.
COSTNew final\through\MY 2022: Technology cost for MYs 2018-2029 
under new final standards through MY 2022.
COSTNew final\through\MY 2023: Technology cost for MYs 2018-2029 
under new final standards through MY 2023.
COSTNo-Action\through\MY 2022: Technology cost for MYs 2018-2029 
under no-action alternative standards through MY 2022.
COSTNo-Action\through\MY 2023: Technology cost for MYs 2018-2029 
under no-action alternative standards through MY 2023.

    Furthermore, today's analysis includes impacts on new vehicle sales 
volumes and the use (i.e., survival) of vehicles of all model years, 
such that standards introduced in a model year produce impacts 
attributable to vehicles having been in operation for some time. For 
example, as modeled here, standards for MY 2021 will impact the prices 
of new vehicles starting in MY 2017, and those price impacts will 
affect the survival of all vehicles still in operation in calendar 
years 2018 and beyond (e.g., MY 2021 standards impact the operation of 
MY 2007 vehicles in calendar year 2027). Therefore, while past 
rulemaking analyses focused largely on impacts over the useful lives of 
the explicitly modeled fleets, much of today's analysis considers all 
model years through 2029, as operated over their entire useful lives. 
For some impacts, such as on technology penetration rates, average 
vehicle prices, and average vehicle ownership costs, the focus was on 
the useful life of the MY 2029 fleet, as the simulation of 
manufacturers' technology application and credit use (when included in 
the analysis) continues to evolve after model year 2026, stabilizing by 
model year 2029.
    Responding to comments recommending that the agencies present 
impacts on a calendar year basis, today's rulemaking does so, with the 
presented results extending through calendar year 2050, the last 
calendar year that includes an on-road fleet with all vehicle vintages 
represented.
    Effects were evaluated from four perspectives: The social 
perspective, the manufacturer perspective, the private perspective, and 
the physical perspective. The social perspective focuses on economic 
benefits and costs, setting aside economic transfers such as fuel taxes 
but including economic externalities such as the social cost of 
CO2 emissions. The manufacturer perspective focuses on 
average requirements and levels of performance (i.e., average fuel 
economy level and CO2 emission rates), compliance costs, and 
degrees of technology application. The private perspective focuses on 
costs of vehicle purchase and ownership, including outlays for fuel 
(and fuel taxes). The physical perspective focuses on national-scale 
highway travel, fuel consumption, highway fatalities, and carbon 
dioxide and criteria pollutant emissions.
    This analysis does not explicitly identify ``co-benefits,'' as such 
a concept would include all benefits other than cost savings to vehicle 
buyers. Instead, it distinguishes between private benefits--which 
include economic impacts on vehicle manufacturers, buyers of new cars 
and light trucks, and owners (or users) of used cars and light trucks--
and external benefits, which represent indirect benefits (or costs) to 
the remainder of the U.S. economy that stem from the final rule's 
effects on the behavior of vehicle manufacturers, buyers, and users. In 
this accounting framework, changes in fuel use and safety impacts 
resulting from the final rule's effects on the number of used vehicles 
in use represent an important component of its private benefits and 
costs, despite the fact that previous analyses have failed to recognize 
these effects. The agencies' presentation of private costs and benefits 
clearly distinguishes between those that would be experienced by owners 
and users of cars and light trucks produced during previous model years 
and those that would be experienced by buyers and users of new cars and 
light trucks subject to the final standards. Moreover, it clearly 
separates these into benefits related to fuel consumption and those 
related to safety consequences of vehicle use. This is more meaningful 
and informative than simply identifying all impacts other than changes 
in fuel savings to buyers of new vehicles as ``co-benefits.''
    For the social perspective, the following effects for model years 
through 2029 as operated through calendar year 2068 are summarized:
     Technology Costs: Incremental cost, as expected to be paid 
by vehicle purchasers, of fuel-saving technology beyond that added 
under the no-action alternative.
     Hybrid Vehicle Welfare Loss: Loss of value to vehicle 
owners resulting from incremental increases in the numbers of strong 
and plug-in hybrid electric vehicles (strong HEVs or SHEVs, and PHEVs) 
and/or battery electric vehicles (BEVs), beyond increases occurring 
under the no-action alternative.\2426\ The loss of value is a function 
of the factors that lead to different valuations for conventional and 
electric versions of similar-size vehicles (e.g., differences in: 
Travel range, recharging time versus refueling time, performance, and 
comfort).
---------------------------------------------------------------------------

    \2426\ Through MY 2029, the ``standard setting'' analysis of 
CAFE standards sets aside the potential that manufacturers might by 
introduce new BEV (or FCV) vehicle models, but allows that the 
numbers of such vehicles produced might increase or decrease along 
with overall U.S. sales of new passenger cars and light trucks, and 
allows that additional BEV or FCV vehicle models might be intruded 
after MY 2029.
---------------------------------------------------------------------------

     Pre-tax Fuel Savings: Incremental savings, beyond those 
achieved under the no-action alternative, in outlays for fuel 
purchases, setting aside fuel taxes.
     Mobility Benefit: Value of incremental travel, beyond that

[[Page 24905]]

occurring under the no-action alternative.
     Lost New Vehicle Consumer Surplus: Value of incremental 
savings to new vehicle buyers due to cheaper vehicle prices.
     Implicit Opportunity Cost: \2427\ Value of other vehicle 
attributes forwent to apply technology to meet the standards.
---------------------------------------------------------------------------

    \2427\ This value is set to ``0'' for the central analysis.
---------------------------------------------------------------------------

     Refueling Benefit: Value of incremental reduction, 
compared to the no-action alternative, of time spent refueling 
vehicles.
     Non-Rebound Fatality Costs: Social value of additional 
fatalities, beyond those occurring under the no-action alternative, 
setting aside any additional travel attributable to the rebound effect.
     Rebound Fatality Costs: Social value of additional 
fatalities attributable to the rebound effect, beyond those occurring 
under the no-action alternative.
     Benefits Offsetting Rebound Fatality Costs: Assumed 
further value, offsetting rebound fatality costs internalized by 
drivers, of additional travel attributed to the rebound effect.
     Non-Rebound Non-Fatal Crash Costs: Social value of 
additional crash-related losses (other than fatalities), beyond those 
occurring under the no-action alternative, setting aside any additional 
travel attributable to the rebound effect.
     Rebound Non-Fatal Crash Costs: Social value of additional 
crash-related losses (other than fatalities) attributable to the 
rebound effect, beyond those occurring under the no-action alternative.
     Benefits Offsetting Rebound Non-Fatal Crash Costs: Assumed 
further value, offsetting rebound non-fatal crash costs internalized by 
drivers, of additional travel attributed to the rebound effect.
     Additional Congestion and Noise (Costs): Value of 
additional congestion and noise resulting from incremental travel, 
beyond that occurring under the no-action alternative.
     Energy Security Benefit: Value of avoided economic 
exposure to petroleum price ``shocks,'' the avoided exposure resulting 
from incremental reduction of fuel consumption beyond that occurring 
under the no-action alternative.
     Avoided CO2 Damages (Benefits): Social value of 
incremental reduction of CO2 emissions, compared to 
emissions occurring under the no-action alternative.
     Other Avoided Pollutant Damages (Benefits): Social value 
of incremental reduction of criteria pollutant emissions, compared to 
emissions occurring under the no-action alternative.
     Total Costs: Sum of incremental technology costs, hybrid 
vehicle welfare loss, fatality costs, non-fatal crash costs, and 
additional congestion and noise costs.
     Total Benefits: Sum of pretax fuel savings, mobility 
benefits, refueling benefits, Benefits Offsetting Rebound Fatality 
Costs, Benefits Offsetting Rebound Non-Fatal Crash Costs, energy 
security benefits, and benefits from reducing emissions of 
CO2, the CO2 equivalent of other associated 
gases, and criteria pollutants.
     Net Benefits: Total benefits minus total costs.
     Retrievable Electrification Costs: The portion of HEV, 
PHEV, and BEV technology costs which can be passed onto consumers, 
using the willingness to pay analysis described above.
     Electrification Tax Credits: Estimates of the portion of 
HEV, PHEV, and BEV technology costs which are covered by Federal or 
State tax incentives.
     Irretrievable Electrification Costs: The portion of HEV, 
PHEV, and BEV technology costs OEM's must either absorb as a profit 
loss, or cross-subsidize with the prices of internal combustion engine 
(ICE) vehicles.
     Total Electrification Costs: Total incremental technology 
costs attributable to HEV, PHEV, or BEV vehicles.
    For the manufacturer perspective, the following effects for the 
aggregation of model years 2017-2029 are summarized:
     Average Required Fuel Economy: Average of manufacturers' 
CAFE requirements for indicated fleet(s) and model year(s).
     Percent Change in Stringency from Baseline: Percentage 
difference between averages of fuel economy requirements under no-
action and indicated alternatives.
     Average Required Fuel Economy: Industry-wide average of 
fuel economy levels achieved by indicated fleet(s) in indicated model 
year(s).
     Percent Change in Stringency from Baseline: Percentage 
difference between averages of fuel economy levels achieved under no-
action and indicated alternatives.
     Total Technology Costs ($b): Cost of fuel-saving 
technology beyond that applied under no-action alternative.
     Total Civil Penalties ($b): Cost of civil penalties (for 
the CAFE program) beyond those levied under no-action alternative.
     Total Regulatory Costs ($b): Sum of technology costs and 
civil penalties.
     Sales Change (millions): Change in number of vehicles 
produced for sale in U.S., relative to the number estimated to be 
produced under the no-action alternative.
     Revenue Change ($b): Change in total revenues from vehicle 
sales, relative to total revenues occurring under the no-action 
alternative.
     Curb Weight Reduction: Reduction of average curb weight, 
relative to MY 2017.
     Technology Penetration Rates: MY 2030 average technology 
penetration rate for indicated ten technologies (three engine 
technologies, advanced transmissions, and six degrees of 
electrification).
     Average Required CO2: Average of manufacturers' 
CO2 requirements for indicated fleet(s) and model year(s).
     Percent Change in Stringency from Baseline: Percentage 
difference between averages of CO2 requirements under no-
action and indicated alternatives.
     Average Achieved CO2: Average of manufacturers' 
CO2 emission rates for indicated fleet(s) and model year(s).
    For the private perspective, the following effects for the MY 2030 
fleet are summarized:
     Average Price Increase: Average increase in vehicle price, 
relative to the average occurring under the no-action alternative.
     Implicit Opportunity Cost: The lost benefit of vehicle 
attributes that consumers prefer, which are sacrificed by manufacturers 
to comply with the standards.
     Hybrid Vehicle Welfare Loss (Costs): Average loss of value 
to vehicle owners resulting from incremental increases in the numbers 
of strong HEVs, PHEVs) and/or BEVs, beyond increases occurring under 
the no-action alternative. The loss of value is a function of the 
factors that lead to different valuations for conventional and electric 
versions of similar-size vehicles (e.g., differences in: Travel range, 
recharging time versus refueling time, performance, and comfort).
     Ownership Costs: Average increase in some other costs of 
vehicle ownership (taxes, fees, financing), beyond increase occurring 
under the no-action alternative.
     Lost Consumer Surplus: Value of incremental savings to new 
vehicle buyers due to cheaper vehicle prices.
     Fuel Savings: Average of fuel outlays (including taxes) 
avoided over a vehicle's expected useful lives, compared to outlays 
occurring under the no-action alternative.

[[Page 24906]]

     Mobility Benefit: Average incremental value of additional 
travel over average vehicles' useful lives, compared to travel 
occurring under the no-action alternative.
     Refueling Benefit: Average incremental value of avoided 
time spent refueling over average vehicles' useful lives, compared to 
time spent refueling under the no-action alternative.
     Total Costs: Sum of average price increase, welfare loss, 
and ownership costs.
     Total Benefits: Sum of fuel savings, the mobility benefit, 
and the refueling benefit.
     Net Benefits: Total benefits minus total costs.
    For the physical perspective, the following effects for model years 
through 2029 as operated through calendar year 2068 are summarized:
     Fuel Consumption, with rebound (billion gallons): 
Reduction of fuel consumption, relative to the no-action alternative, 
and including the rebound effect.
     Fuel Consumption, without rebound (billion gallons): 
Reduction of fuel consumption, relative to the no-action alternative, 
and excluding the rebound effect.
     Greenhouse Gases: Includes carbon dioxide 
(CO2), methane (CH4), and nitrous oxide 
(N2O), and values are reported separately for vehicles 
(tailpipe) and upstream processes (combining fuel production, 
distribution, and delivery) and shown as reductions in carbon dioxide 
or its equivalent relative to the no-action alternative.
     Criteria Pollutants: Includes carbon monoxide (CO), 
volatile organic compounds (VOC), nitrogen oxides (NOX), 
sulfur dioxide (SO2) and particulate matter (PM), and values 
are shown as reductions relative to the no-action alternative.
     Fuel Consumption: Aggregates all fuels, with electricity, 
hydrogen, and compressed natural gas (CNG) included on a gasoline-
equivalent-gallon (GEG) basis, and values are shown as reductions 
relative to the no-action alternative.
     VMT, with rebound (billion miles): Increase in highway 
travel (as vehicle miles traveled), relative to the no-action 
alternative, and including the rebound effect.
     VMT, without rebound (billion miles): Increase in highway 
travel (as vehicle miles traveled), relative to the no-action 
alternative, and excluding the rebound effect.
     Fatalities, with rebound: Increase in highway fatalities, 
relative to the no-action alternative, and including the rebound 
effect.
     Fatalities, without rebound: Increase in highway 
fatalities, relative to the no-action alternative, and excluding the 
rebound effect.
     Health Effects: Increase in the occurrence of a variety of 
health effects of criteria pollutant emissions, relative to the no-
action alternative, and reported separately for tailpipe and upstream 
emissions.
    Below, this section tabulates results for each of these four 
perspectives and does so separately for the new final CAFE and 
CO2 standards. More detailed results are presented in the 
FRIA accompanying today's rulemaking, and additional and more detailed 
analysis of environmental impacts for CAFE regulatory alternatives is 
provided in the corresponding Final Environmental Impact Statement 
(FEIS). Underlying CAFE model output files are available (along with 
input files, model, source code, and documentation) on NHTSA's 
website.\2428\ Summarizing and tabulating results for presentation here 
involved considerable ``off model'' calculations (e.g., to combine 
results for selected model years and calendar years, and to combine 
various components of social and private costs and benefits); tools 
Volpe Center staff used to perform these calculations are also 
available on NHTSA's website.\2429\
---------------------------------------------------------------------------

    \2428\ Compliance and Effects Modeling System, National Highway 
Traffic Safety Administration, https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
    \2429\ These tools, available at the same location, are scripts 
executed using R, a free software environment for statistical 
computing. R is available through https://www.r-project.org/.
---------------------------------------------------------------------------

    While the National Environmental Policy Act (NEPA) requires NHTSA 
to prepare an EIS documenting estimating environmental impacts of the 
regulatory alternatives under consideration in CAFE rulemakings, NEPA 
does not require EPA to do so for EPA rulemakings. With CO2 
standards for each regulatory alternative being harmonized as practical 
with corresponding CAFE standards, environmental impacts of 
CO2 standards should be directionally identical and similar 
in magnitude to those of CAFE standards. Nevertheless, in this section, 
following the series of tables below, today's announcement provides a 
more detailed analysis of estimated impacts of the new final CAFE and 
CO2 standards. Results presented herein for the CAFE 
standards differ slightly from those presented in the FEIS; while, as 
discussed above, EPCA/EISA requires that the Secretary determine the 
maximum feasible levels of CAFE standards in manner that, as presented 
here, sets aside the potential use of CAFE credits or application of 
alternative fuels toward compliance with new standards, NEPA does not 
impose such constraints on any analysis presented in corresponding 
FEISs, and the FEIS presents results of an ``unconstrained'' analysis 
that considers manufacturers' potential application of alternative 
fuels and use of CAFE credits.
    In terms of all estimated impacts, including estimated costs and 
benefits, the results of today's analysis are different for CAFE and 
CO2 standards. Differences arise because, even when the 
mathematical functions defining fuel economy and CO2 targets 
are ``harmonized,'' surrounding regulatory provisions may not be. For 
example, while both CAFE and CO2 standards allow credits to 
be transferred between fleets and traded between manufacturers, EPCA/
EISA places explicit and specific limits on the use of such credits, 
such as by requiring that each domestic passenger car fleet meet a 
minimum CAFE standard (as discussed above). The CAA provides no 
specific direction regarding CO2 standards, and while EPA 
has adopted many regulatory provisions harmonized with specific EPCA/
EISA provisions (e.g., separate standards for passenger cars and light 
trucks), EPA has not adopted all such provisions. For example, EPA has 
not adopted the EPCA/EISA provisions limiting transfers between 
regulated fleet or requiring separate compliance by domestic and 
imported passenger car fleets. Such differences introduce variance 
between impacts estimated under CAFE standards and under CO2 
standards. Also, as mentioned above, Congress has required that new 
CAFE standards be considered in a manner that sets aside the potential 
use of CAFE credits and the potential additional application of 
alternative fuel vehicles (such as electric vehicles) during the model 
years under consideration. Congress has provided no corresponding 
direction regarding the analysis of potential CO2 standards, 
and today's analysis does consider these potential responses to 
CO2 standards.
    Tables in the remaining section summarize these estimated impacts 
for each alternative, considering the same measures as shown above for 
the final standards. For the final standards, social costs and 
benefits, private costs and benefits, and environmental and energy 
impacts were evaluated, and were done so separately for CAFE and 
CO2 standards defining each regulatory alternative. Also, 
for the final standards, the compliance-related private costs and

[[Page 24907]]

benefits were evaluated separately for domestic and imported passenger 
cars under CAFE standards but not under CO2 standards 
because EPCA/EISA's requirement for separate compliance applies only to 
CAFE standards.
    Both the final standards and, all other alternatives involve 
standards less stringent than the no-action alternative. Therefore, as 
discussed above, incremental benefits and costs for each alternative 
are negative--in other words, each alternative involves forgone 
benefits and avoided costs. Environmental and energy impacts are 
correspondingly negative, involving forgone avoided CO2 
emissions and forgone avoided fuel consumption. For consistency with 
past rulemakings, these are reported as negative values rather than as 
additional CO2 emissions and additional fuel consumption.
    Like the NPRM and PRIA (and past rulemakings), today's rulemaking 
and FRIA emphasize a ``model year'' perspective when reporting impacts. 
That is, for enough model years (here, through MY 2029) to extend 
beyond those when the estimated use of ``banked'' credits is reasonably 
likely to be sufficient to show the average manufacturer not achieving 
required CAFE or CO2 levels, the presentation of results 
mainly considers the lifetime impacts attributable to vehicles produced 
in these model years. Because standards are actually enforced on a 
model year basis, this perspective aligns well with the consideration 
of impacts on manufacturers and new vehicle buyers. However, impacts on 
national energy consumption and the natural environment will involve 
all vehicles on the road in future years, including those produced 
after MY 2029. Therefore, similar to the approach followed in recent 
and past EISs (and today's FEIS), today's rulemaking also presents 
impacts on a ``calendar year'' basis--that is, summarizing overall 
impacts (i.e., including those attributable to vehicles produced after 
MY 2029) in each calendar year through 2050. As discussed in below, the 
model year and calendar year perspectives draw on the same CAFE model 
outputs, but differ in the scope of those outputs included in 
summarized information.
    As discussed above, more detailed results are available in the FRIA 
and FEIS accompanying today's rulemaking, as well as in underlying 
model output files posted on NHTSA's website.
1. Average Required Fuel Economy and CO2 Standard for PCs, 
LTs, and Combined
    The model fully represents the required CAFE and CO2 
levels for every manufacturer and every fleet. The standard for each 
manufacturer is based on the harmonic average of footprint targets (by 
volume) within a fleet, just as the standards prescribe. Unlike earlier 
versions of the CAFE model, the current version further disaggregates 
passenger cars into domestic and imported classes (which manufacturers 
report to NHTSA and EPA as part of their CAFE compliance submissions). 
This allows the CAFE model to more accurately estimate the requirement 
on the two passenger car fleets, represent the domestic passenger car 
floor (which must be exceeded by every manufacturer's domestic fleet, 
without the use of credits, but with the possibility of civil penalty 
payment), and allows it to enforce the transfer cap limit that exists 
between domestic and imported passenger cars, all for purposes of the 
CAFE program.
    In calculating the achieved CAFE level, the model uses the 
prescribed harmonic average of fuel economy ratings within a vehicle 
fleet. Under an ``unconstrained'' analysis, or in a model year for 
which standards are already final, it is possible for a manufacturer's 
CAFE to fall below its required level without generating penalties 
because the model will apply expiring or transferred credits to 
deficits if it is strategically appropriate to do so. Consistent with 
current EPA regulations, the model applies simple (not harmonic) 
production-weighted averaging to calculate average CO2 
levels.
    While the CAFE and CO2 standards themselves are, as 
discussed in Section VI, inputs to the agencies' analysis, because the 
standards are attribute-based standards specified separately for 
passenger car and light truck fleets and applicable to average fuel 
economy and CO2 levels, average requirements under these 
standards are analytical results, not analytical inputs. Also, because 
EPCA requires NHTSA to determine in advance minimum requirements that 
will be applicable to manufacturers' fleets of domestic passenger cars, 
these, too, are analytical results. The remainder of this section 
presents these results.
a) Passenger Car Requirements
    As discussed in Section V, the final standards are different from 
the preferred alternative identified in the proposal.
    We do not know yet with certainty what CAFE and CO2 
levels will ultimately be required of individual manufacturers, because 
those levels will depend on the mix of vehicles that each manufacturer 
produces for sale in future model years. Based on the market forecast 
of future sales used to examine the final standards, the agencies 
currently estimate that the target functions shown above would result 
in the following average required fuel economy and CO2 
emissions levels for all manufacturers during MYs 2021-2026:

[[Page 24908]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.535

    We emphasize again that the values in these tables are estimates, 
and not necessarily the ultimate levels with which each of these 
manufacturers will have to comply, for the reasons described 
above.\2430\
---------------------------------------------------------------------------

    \2430\ MY2017 values reflect the agencies' analysis, which uses 
an analysis fleet developed using MY2017 compliance data as of 
summer 2019. The analysis does not reflect subsequent updates and 
corrections to manufacturers' MY2017 compliance data.
---------------------------------------------------------------------------

b) Light Truck Requirements
    Again, while the agencies do not know yet with certainty what CAFE 
and CO2 levels will ultimately be required of individual 
manufacturers, because those levels will depend on the mix of vehicles 
that each manufacturer produces for sale in future model years, based 
on the market forecast of future sales used to examine today's proposed 
standards, the agencies currently estimate that the target functions 
shown above would result in the following average required fuel economy 
and CO2 emissions levels for individual manufacturers during 
MYs 2021-2026.
[GRAPHIC] [TIFF OMITTED] TR30AP20.536


[[Page 24909]]


    We emphasize again the values in these tables are estimates and not 
necessarily the ultimate levels with which each of these manufacturers 
will have to comply for reasons described above.\2431\
---------------------------------------------------------------------------

    \2431\ MY2017 values reflect the agencies' analysis, which uses 
an analysis fleet developed using MY2017 compliance data as of 
summer, 2019. The analysis does not reflect subsequent updates and 
corrections to manufacturers' MY2017 compliance data.
---------------------------------------------------------------------------

c) Average of PassengerCcar and Light Truck Requirements
    Overall average requirements will depend, further, on the relative 
shares of passenger cars and light trucks in the new vehicle fleet. The 
agencies' analysis estimates future shifts in these shares as vehicles' 
average prices and fuel economy levels change, and as fuel prices also 
change. Resultant estimates of overall average requirements are as 
follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.537

(d) Estimated Average Requirements for Specific Manufacturers
    Overall average requirements (e.g., reflecting both passenger car 
and light truck fleets) applicable to each manufacturer will depend on 
the mix (i.e., footprint distribution) of vehicles produced in each 
model year, and relative production shares of passenger cars and light 
trucks. Tables appearing below summarize estimated requirements through 
model year 2029. Estimates for specific fleets (e.g., domestic 
passenger cars, imported passenger cars, light trucks) are available in 
CAFE model output files accompanying today's rulemaking, as are 
estimates for each MYs 2030-2050.\2432\
---------------------------------------------------------------------------

    \2432\ The model and all inputs and outputs supporting today's 
notice are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 24910]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.538


[[Page 24911]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.539


[[Page 24912]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.540


[[Page 24913]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.541


[[Page 24914]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.542


[[Page 24915]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.543


[[Page 24916]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.544


[[Page 24917]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.545


[[Page 24918]]


BILLING CODE 4910-59-C
2. Impacts on Vehicle Manufacturers
    As mentioned above, impacts are presented from two different 
perspectives for today's final rule. From either perspective, overall 
impacts are the same. The first perspective, taken above in VII.A, 
examines overall impacts of the standards--i.e., the entire series of 
year-by-year standards--on each model year. The second perspective, 
presented here, provides a clearer characterization of the incremental 
impacts attributable to standards introduced in each successive model 
year. For example, the new final standards for MY 2023 are likely to 
impact manufacturers' application of technology in model years prior to 
MY 2023, as well as model years after MY 2023. By conducting analysis 
that successively introduces standards for each MY, in turn, isolates 
the incremental impacts attributable to new standards introduced in 
each MY, considering the entire span of MYs 1975-2029 and calendar 
years 2016-2069 included in the analysis that only considers the full 
series of successive MYs' standards. Tables appearing below summarize 
results as aggregated across these model and calendar years. Underlying 
model output files \2433\ report physical impacts and specific 
monetized costs and benefits attributable to each model year in each 
calendar (thus providing information needed to, for example, 
differentiate between impacts attributable to the MY 1975-2017 and MY 
2018-2029 cohorts). The FRIA presents costs and benefits for individual 
model years (with MY's 1975-2017 in a single bucket) for the final 
standards.
---------------------------------------------------------------------------

    \2433\ Available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

a) Industry Average Technology Penetration Rates
    The CAFE model tracks and reports technology application and 
penetration rates for each manufacturer, regulatory class, and model 
year, calculated as the volume of vehicles with a given technology 
divided by the total volume. The ``application rate'' accounts only for 
those technologies applied by the model during the compliance 
simulation, while the ``penetration rate'' accounts for the total 
percentage of a technology present in a given fleet, whether applied by 
the CAFE model or already present at the start of the simulation.
    In addition to the aggregate representation of technology 
penetration, the model also tracks each individual vehicle model on 
which it has operated. Accordingly, the CAFE model produces a record 
for every model year and every alternative that identifies with which 
technologies the vehicle started the simulation and which technologies 
the same vehicle had at the conclusion of each model year. Interested 
parties may use these outputs to assess how the compliance simulation 
modified any vehicle that was offered for sale in MY 2017 in response 
to a given regulatory alternative.
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.546


[[Page 24919]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.547

[GRAPHIC] [TIFF OMITTED] TR30AP20.548


[[Page 24920]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.549

[GRAPHIC] [TIFF OMITTED] TR30AP20.550


[[Page 24921]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.551

[GRAPHIC] [TIFF OMITTED] TR30AP20.552


[[Page 24922]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.553

[GRAPHIC] [TIFF OMITTED] TR30AP20.554


[[Page 24923]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.555

[GRAPHIC] [TIFF OMITTED] TR30AP20.556

[GRAPHIC] [TIFF OMITTED] TR30AP20.557


[[Page 24924]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.558

[GRAPHIC] [TIFF OMITTED] TR30AP20.559


[[Page 24925]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.560


[[Page 24926]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.561


[[Page 24927]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.562


[[Page 24928]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.563


[[Page 24929]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.564


[[Page 24930]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.565


[[Page 24931]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.566


[[Page 24932]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.567


[[Page 24933]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.568


[[Page 24934]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.569


[[Page 24935]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.570


[[Page 24936]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.571


[[Page 24937]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.572


[[Page 24938]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.573


[[Page 24939]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.574


[[Page 24940]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.575


[[Page 24941]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.576


[[Page 24942]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.577


[[Page 24943]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.578


[[Page 24944]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.579


[[Page 24945]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.580


[[Page 24946]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.581


[[Page 24947]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.582


[[Page 24948]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.583


[[Page 24949]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.584


[[Page 24950]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.585


[[Page 24951]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.586


[[Page 24952]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.587


[[Page 24953]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.588


[[Page 24954]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.589


[[Page 24955]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.590


[[Page 24956]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.591


[[Page 24957]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.592


[[Page 24958]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.593


[[Page 24959]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.594


[[Page 24960]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.585


[[Page 24961]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.596


[[Page 24962]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.597


[[Page 24963]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.598


[[Page 24964]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.599


[[Page 24965]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.600


[[Page 24966]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.601


[[Page 24967]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.602


[[Page 24968]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.603


[[Page 24969]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.604


[[Page 24970]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.605


[[Page 24971]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.606


[[Page 24972]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.607


[[Page 24973]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.608


[[Page 24974]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.609


[[Page 24975]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.610


[[Page 24976]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.611


[[Page 24977]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.612


[[Page 24978]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.613


[[Page 24979]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.614


[[Page 24980]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.615


[[Page 24981]]


BILLING CODE 4910-59-C
b) Technology Costs
    For each technology that the model adds to a given vehicle, it 
accumulates cost. The technology costs are defined incrementally and 
vary both over time and by technology class, where the same technology 
may cost more to apply to larger vehicles as it involves more raw 
materials or requires different specifications to preserve some 
performance attributes. While learning-by-doing can bring down cost, 
and should reasonably be implemented in the CAFE model as a rate of 
cost reduction that is applied to the cumulative volume of a given 
technology produced by either a single manufacturer or the industry as 
a whole, in practice this notion is implemented as a function of time, 
rather than production volume. Thus, depending upon where a given 
technology starts along its learning curve, it may appear to be cost-
effective in later years where it was not in earlier years. As the 
model carries forward technologies that it has already applied to 
future model years, it similarly adjusts the costs of those 
technologies based on their individual learning rates.
BILLING CODE 4910-59-P

[[Page 24982]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.616


[[Page 24983]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.617


[[Page 24984]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.618


[[Page 24985]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.619


[[Page 24986]]


BILLING CODE 4910-59-C
c) Civil Penalties
    The other costs that manufacturers incur as a result of CAFE 
standards are civil penalties resulting from non-compliance with CAFE 
standards. The CAFE model accumulates costs of $5.50 per 1/10-MPG under 
the standard, multiplied by the number of vehicles produced in that 
fleet, in that model year. The model reports as the full ``regulatory 
cost,'' the sum of total technology cost and total fines by the 
manufacturer, fleet, and model year. As mentioned above, the relevant 
EPCA/EISA provisions do not also appear in the CAA, so this option and 
these costs apply only to simulated compliance with CAFE standards.
d) Average Prices, Sales, and Revenue Changes
    In all previous versions of the CAFE model, the total number of 
vehicles sold in any model year, in fact the number of each individual 
vehicle model sold in each year, has been a static input that did not 
vary in response to price increases induced by CAFE standards, nor 
changes in fuel prices, or any other input to the model. The only way 
to alter sales, was to update the entire forecast in the market input 
file. However, in the 2012 final rule, the agencies included a dynamic 
fleet share model that was based on a module in the Energy Information 
Administration's NEMS model. This fleet share model did not change the 
size of the new vehicle fleet in any year, but it did change the share 
of new vehicles that were classified as passenger cars (or light 
trucks). That capability was not included in the central analysis but 
was included in the uncertainty analysis, which looked at the baseline 
and final standards in the context of thousands of possible future 
states of the world. As some of those futures contained extreme cases 
of fuel prices, it was important to ensure consistent modeling 
responses within that context. For example, at a gasoline price of $7/
gallon, it would be unrealistic to expect the new vehicle market's 
light truck share to be the same as the future where gasoline cost $2/
gallon. The current model has slightly modified, and fully integrated, 
the dynamic fleet share model. Every regulatory alternative and 
sensitivity case considered for this analysis reflects a dynamically 
responsive fleet mix in the new vehicle market.
    While the dynamic fleet share model adjusts unit sales across body 
styles (cars, SUVs, and trucks), it does not modify the total number of 
new vehicles sold in a given year. The CAFE model now includes a 
separate function to account for changes in the total number of new 
vehicles sold in a given year (regardless of regulatory class or body 
style), in response to certain macroeconomic inputs and changes in the 
average new vehicle price. The price impact is modest relative to the 
influence of the macroeconomic factors in the model. The combination of 
these two models modify the total number of new vehicles, the share of 
passenger cars and light trucks, and, as a consequence, the number of 
each given model sold by a given manufacturer. However, these two 
factors are insufficient to cause large changes to the composition of 
any of a manufacturer's fleets. In order to change significantly the 
mix of models produced within a given fleet, the CAFE model would 
require a way to trade off the production of one vehicle versus another 
both within a manufacturer's fleet and across the industry. While the 
agencies have experimented with fully-integrated consumer choice 
models, their performance has yet to satisfy the requirements of a 
rulemaking analysis.
    Above, Section VI discusses at length the sales model the agencies 
have applied in the analysis supporting today's rulemaking.
BILLING CODE 4910-59-P

[[Page 24987]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.620


[[Page 24988]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.621


[[Page 24989]]


BILLING CODE 4910-59-C
e) Labor
    As discussed in Section VI the analysis includes estimates of 
impacts on U.S. auto industry labor, considering the combined impact of 
changes in sales volumes and changes in outlays for additional fuel-
saving technology. Note: This analysis does not consider the 
possibility that potential new jobs and plants attributable to 
increased stringency will not be located in the United States, or that 
increased stringency will not lead to the relocation of current jobs or 
plants to foreign countries. Compared to the no-action alternative 
(i.e., the baseline standards), the new final standards (alternative 1) 
and other regulatory alternatives under consideration all involve 
reduced regulatory costs expected to lead to reduced average vehicle 
prices and, in turn, increased sales. While the increased sales 
slightly increase estimated U.S. auto sector labor hours, because 
producing and selling more vehicles uses additional U.S. labor, the 
reduced outlays for fuel-saving technology slightly reduce estimated 
U.S. auto sector labor hours, because manufacturing, integrating, and 
selling less technology means using less labor to do so. Of course, 
this is technology that may not otherwise be produced or deployed were 
it not for regulatory mandate, and the additional costs of this 
technology would be borne by a reduced number of consumers given 
reduction in sales in response to increased prices. Today's analysis 
shows the negative impact of reduced mandatory technology outlays 
outweighing the positive impact of increased sales. However, both of 
these underlying factors are subject to uncertainty. For example, if 
fuel-saving technology that would have been applied under the baseline 
standards is more likely to have come from foreign suppliers than 
estimated here, less of the forgone labor to manufacture that 
technology would have been U.S. labor. Also, if sales would be more 
positively impacted by reduced vehicle prices than estimated here, 
correspondingly positive impacts on U.S. auto sector labor could be 
magnified. Alternatively, if manufacturers are able to deploy 
technology to improve vehicle attributes that new car buyers prefer to 
fuel economy improvements, both technology spending and vehicle sales 
would correspondingly increase.
    The labor utilization analysis was focused on automotive labor 
because adjacent labor utilization factors and consumer spending 
factors for other goods and services are uncertain and difficult to 
predict. How direct labor changes may affect the macro economy and 
possibly change employment in adjacent industries were not considered. 
For instance, possible labor changes in vehicle maintenance and repair 
were not considered, nor were changes in labor at retail gas stations 
considered. Possible labor changes due to raw material production, such 
as production of aluminum, steel, copper, and lithium were not 
considered, nor were possible labor impacts due to changes in 
production of oil and gas, ethanol, and electricity considered. Effects 
of how consumers could spend money saved due to improved fuel economy 
were not analyzed, nor were effects of how consumers would pay for more 
expensive fuel savings technologies at the time of purchase analyzed; 
either could affect consumption of other goods and services, and hence 
affect labor in other industries. The effects of increased usage of 
car-sharing, ride-sharing, and automated vehicles were not analyzed. 
How changes in labor from any industry could affect gross domestic 
product and possibly affect other industries as a result were not 
estimated.
    Also, no assumptions were made about full-employment or not full-
employment and the availability of human resources to fill positions. 
When the economy is at full employment, a fuel economy regulation is 
unlikely to have much impact on net overall U.S. labor utilization; 
instead, labor would primarily be shifted from one sector to another. 
These shifts in employment impose an opportunity cost on society, 
approximated by the wages of the employees, as regulation diverts 
workers from other activities in the economy. In this situation, any 
effects on net employment are likely to be transitory as workers change 
jobs (e.g., some workers may need to be retrained or require time to 
search for new jobs, while shortages in some sectors or regions could 
bid up wages to attract workers). On the other hand, if a regulation 
comes into effect during a period of high unemployment, a change in 
labor demand due to regulation may affect net overall U.S. employment 
because the labor market is not in equilibrium. Schmalansee and Stavins 
point out that net positive employment effects are possible in the near 
term when the economy is at less than full employment due to the 
potential hiring of idle labor resources by the regulated sector to 
meet new requirements (e.g., to install new equipment) and new economic 
activity in sectors related to the regulated sector. In the longer run, 
the net effect on employment is more difficult to predict and will 
depend on the way in which the related industries respond to the 
regulatory requirements. For that reason, this analysis does not 
include multiplier effects but instead focuses on labor impacts in the 
most directly affected industries. Those sectors are likely to face the 
most concentrated labor impacts.
    The tables presented below summarize these results for the final 
standards and other regulatory alternatives considered. While values 
are reported as thousands of person-years, changes in labor utilization 
would not necessarily involve the same number of changes in actual 
jobs, as auto industry employers may use a range of strategies (e.g., 
shift changes, overtime) beyond simply adding or eliminating jobs.
(1) CAFE Standards

[[Page 24990]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.622

[GRAPHIC] [TIFF OMITTED] TR30AP20.623

(2) CO2 Standards

[[Page 24991]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.624

3. Impacts to Vehicle Buyers
BILLING CODE 4910-59-P

[[Page 24992]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.625


[[Page 24993]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.626


[[Page 24994]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.627


[[Page 24995]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.628


[[Page 24996]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.629


[[Page 24997]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.630


[[Page 24998]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.631


[[Page 24999]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.632


[[Page 25000]]


BILLING CODE 4910-59-C
a) Average Price Increase
BILLING CODE 4910-59-P

[[Page 25001]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.633


[[Page 25002]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.634


[[Page 25003]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.635


[[Page 25004]]


BILLING CODE 4910-59-C
4. Impacts to Society
    As the CAFE model simulates manufacturer compliance with regulatory 
alternatives, it estimates and tracks a number of consequences that 
generate social costs. The most obvious cost associated with the 
program is the cost of additional fuel economy improving/CO2 
emissions reducing technology that is added to new vehicles as a result 
of the rule. However, the model does not inherently draw a distinction 
between costs and benefits. For example, the model tracks fuel 
consumption and the dollar value of fuel consumed. This is the cost of 
travel under a given alternative (including the baseline). The ``cost'' 
or ``benefit'' associated with the value of fuel consumed is determined 
by the reference point against which each alternative is considered. 
The CAFE model reports absolute values for the amount of money spent on 
fuel in the baseline, then reports the amount spent on fuel in the 
alternatives relative to the baseline. If the baseline standard were 
fixed at the current level, and an alternative achieved significantly 
greater mpg by 2025, the total expenditures on fuel in the alternative 
would be lower, creating a fuel savings ``benefit.'' This analysis uses 
a baseline that is more stringent than each alternative considered, so 
the incremental fuel expenditures are greater for the alternatives than 
for the baseline.
    Other social costs and benefits emerge as the result of physical 
phenomena, like tailpipe emissions or highway fatalities, which are the 
result of changes in the composition and use of the on-road fleet. The 
social costs associated with those quantities represent an economic 
estimate of the social damages associated with the changes in each 
quantity. The model tracks and reports each of these quantities by: 
Model year and vehicle age (the combination of which can be used to 
produce calendar year totals), regulatory class, fuel type, and social 
discount rate.
    The full list of potential costs and benefits is presented in Table 
VII-90 as well as the population of vehicles that determines the size 
of the factor (either new vehicles or all registered vehicles) and the 
mechanism that determines the size of the effect (whether driven by the 
number of miles driven, the number of gallons consumed, or the number 
of vehicles produced).
BILLING CODE 4910-59-P

[[Page 25005]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.636


[[Page 25006]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.637


[[Page 25007]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.638


[[Page 25008]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.639


[[Page 25009]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.640


[[Page 25010]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.641


[[Page 25011]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.642


[[Page 25012]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.643


[[Page 25013]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.644


[[Page 25014]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.645


[[Page 25015]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.646


[[Page 25016]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.647


[[Page 25017]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.648


[[Page 25018]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.649


[[Page 25019]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.650


[[Page 25020]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.651


[[Page 25021]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.652


[[Page 25022]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.653


[[Page 25023]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.654

    The above tables summarizing estimated benefits and costs of the 
regulatory alternatives considered here exclude results of the implicit 
opportunity cost calculations discussed above and in Section 
VI.D.1.b)(8)

[[Page 25024]]

Implicit Opportunity Cost. The following four tables show corresponding 
benefits and costs when results of these calculations are included:

[[Page 25025]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.655


[[Page 25026]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.656


[[Page 25027]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.657


[[Page 25028]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.658


[[Page 25029]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.659


[[Page 25030]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.660


[[Page 25031]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.661


[[Page 25032]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.662

a) Impacts on Total Fleet Size, Usage, and Safety

[[Page 25033]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.663


[[Page 25034]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.664


[[Page 25035]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.665


[[Page 25036]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.666


[[Page 25037]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.667


[[Page 25038]]


BILLING CODE 4910-59-C
(1) Total Fleet Size and VMT
    The CAFE model carries a complete representation of the registered 
vehicle population in each calendar year, starting with an aggregated 
version of the most recent available data about the registered 
population for the first year of the simulation. In this analysis, the 
first model year considered is MY 2017, and the registered vehicle 
population enters the model as it appeared at the end of calendar year 
2016. The initial vehicle population is stratified by age (or model 
year cohort) and regulatory class--to which the CAFE model assigns 
average fuel economies based on the reported regulatory class industry 
average compliance value in each model year (and class). Once the 
simulation begins, new vehicles are added to the population from the 
market data file and age throughout their useful lives during the 
simulation, with some fraction of them being retired (or scrapped) 
along the way. For example, in calendar year 2018, the new vehicles 
(age zero) are MY 2018 vehicles (added by the CAFE model simulation and 
represented at the same level of detail used to simulate compliance), 
the age one vehicles are MY 2017 vehicles (added by the CAFE model 
simulation), and the age two vehicles are MY 2016 vehicles (inherited 
from the registered vehicle population and carried through the analysis 
with less granularity). This national registered fleet is used to 
calculate annual fuel consumption, vehicle miles traveled (VMT), 
pollutant emissions, and safety impacts under each regulatory 
alternative.
    In support of prior CAFE rulemakings, the CAFE model accounted for 
new travel that results from fuel economy improvements that reduce the 
cost of driving. The magnitude of the increase in travel demand is 
determined by the rebound effect. In both previous versions and the 
current version of the CAFE model, the amount of travel demanded by the 
existing fleet of vehicles is also responsive to the rebound effect 
(representing the price elasticity of demand for travel)--increasing 
when fuel prices decrease relative to the fuel price when the VMT on 
which our mileage accumulation schedules were built was observed. Since 
the fuel economy of those vehicles is already fixed, only the fuel 
price influences their travel demand relative to the mileage 
accumulation schedule and so is identical for all regulatory 
alternatives.
    While the average mileage accumulation per vehicle by age is not 
influenced by the rebound effect in a way that differs by regulatory 
alternative, three other factors influence total VMT in the model in a 
way that produces different total mileage accumulation by regulatory 
alternative. The first factor is the total industry sales response: New 
vehicles are both driven more than older vehicles and are more fuel 
efficient (thus producing more rebound miles). To the extent that more 
(or fewer) of these new models enter the vehicle fleet in each model 
year, total VMT will increase (or decrease) as a result. The second 
factor is the dynamic fleet share model. The fleet share influences not 
only the fuel economy distribution of the fleet, as light trucks are 
less efficient than passenger cars on average, but the total miles are 
influenced by fact that light trucks are driven more than passenger 
cars as well. Both of the first two factors can magnify the influence 
of the rebound effect on vehicles that go through the compliance 
simulation (MY 2017-2050) in the manner discussed above. The third 
factor influencing total annual VMT is the scrappage model. By 
modifying the retirement rates of on-road vehicles under each 
regulatory alternative, the scrappage model either increases or 
decreases the lifetime miles that accrue to vehicles in a given model 
year cohort.
    In addition to dynamically modifying the total number of new 
vehicles sold, a dynamic model of vehicle retirement, or scrappage, has 
also been implemented. The model implements the scrappage response by 
defining the instantaneous scrappage rate at any age using two 
functions. For ages less than 30, instantaneous scrappage is defined as 
a function of vehicle age, new vehicle price, fuel prices, cost per 
mile of driving (the ratio of fuel price and fuel economy), and GDP 
growth rate. For ages greater than 30, the instantaneous scrappage rate 
is a simple exponential function of age. While the scrappage response 
does not affect manufacturer compliance calculations, it impacts the 
lifetime mileage accumulation (and thus fuel savings) of all vehicles. 
Previous CAFE analyses have focused exclusively on new vehicles, 
tracing the fuel consumption and social costs of these vehicles 
throughout their useful lives; the scrappage effect also impacts the 
registered vehicle fleet that exists when a set of standards is 
implemented.
    For a given calendar year, the retirement rates of the registered 
vehicle population are governed by the scrappage model. To the extent 
that a given set of CAFE or CO2 standards accelerates or 
decelerates the retirement of vehicles, fuel consumption and social 
costs may change. The CAFE model accounts for those costs and benefits, 
as well as tracking all of the standard benefits and costs associated 
with the lifetimes of new vehicles produced under the rule. For more 
detail about the derivation of the scrappage functions, see Section VI.
(2) Fuel Consumption
    For every vehicle model in the market file, the model estimates the 
VMT per vehicle (using the assumed VMT schedule, the vehicle fuel 
economy, fuel price, and the rebound assumption). Those miles are 
multiplied by the volume for each vehicle. Fuel consumption is the 
product of miles driven and fuel economy, which can be tracked by model 
year cohort in the model. Carbon dioxide emissions from vehicle 
tailpipes are the simple product of gallons consumed and the carbon 
content of each gallon.
    In order to calculate calendar year fuel consumption, the model 
needs to account for the inherited on-road fleet in addition to the 
model year cohorts affected by this new final rule. Using the VMT of 
the average passenger car and light truck from each cohort, the model 
computes the fuel consumption of each model year class of vehicles for 
its age in a given CY. The sum across all ages (and thus, model year 
cohorts) in a given CY provides estimated CY fuel consumption.
    Because the model produces an estimate of the aggregate number of 
gallons sold in each CY, it is possible to calculate both the total 
expenditures on motor fuel and the total contribution to the Highway 
Trust Fund (HTF) that result from that fuel consumption. The Federal 
fuel excise tax is levied on every gallon of gasoline and diesel sold 
in the U.S., with diesel facing a higher per-gallon tax rate. The model 
uses a national perspective, where the State taxes present in the input 
files represent an estimated average fuel tax across all U.S. States. 
Accordingly, while the CAFE model cannot reasonably estimate potential 
losses to State fuel tax revenue from increasingly the fuel economy of 
new vehicles, it can do so for the HTF.
    In addition to the tailpipe emissions of carbon dioxide, each 
gallon of gasoline produced for consumption by the on-road fleet has 
associated ``upstream'' emissions that occur in the extraction, 
transportation, refining, and distribution of the fuel. The model 
accounts for these emissions as well (on a per-gallon basis) and 
reports them accordingly.
(3) Safety
    Earlier versions of the CAFE model accounted for the safety impacts 
associated with reducing vehicle mass

[[Page 25039]]

in order to improve fuel economy. In particular, NHTSA's safety 
analysis estimated the additional fatalities that would occur as a 
result of new vehicles getting lighter, then interacting with the on-
road vehicle population. In general, taking mass out of the heaviest 
new vehicles improved safety outcomes, while taking mass from the 
lightest new vehicles resulted in a greater number of expected highway 
fatalities. However, the change in fatalities did not adequately 
account for changes in exposure that occur as a result of increased 
demand for travel as vehicles become cheaper to operate. The current 
version of the model resolves that limitation and addresses additional 
sources of fatalities that can result from the implementation of CAFE 
or CO2 standards. These are discussed in greater detail in 
Section VI.
    The agencies have observed that older vehicles in the population 
are responsible for a disproportionate number of fatalities, both by 
number of registrations and by number of miles driven. Accordingly, any 
factor that causes the population of vehicles to turn over more slowly 
will induce additional fatalities--as those older vehicles continue to 
be driven, rather than being retired and replaced with newer (even if 
not brand new) vehicle models. The scrappage effect, which delays (or 
accelerates) the retirement of registered vehicles, impacts the number 
of fatalities through this mechanism--importantly affecting not just 
new vehicles sold from model years 2017-2050 but existing vehicles that 
are already part of the on-road fleet. Similarly, to the extent that a 
CAFE or CO2 alternative reduces new vehicle sales, it can 
slow the transition from older vehicles to newer vehicles, reducing the 
share of total vehicle miles that are driven by newer, more 
technologically advanced vehicles. Furthermore, newer vehicles are 
equipped with technologies that make driving safer not only safer for 
the occupants of newer vehicles, but also pedestrians, cyclists, and 
even occupants of other vehicles. Accounting for the change in vehicle 
miles traveled that occurs when vehicles become cheaper to operate 
leads to a number of fatalities that can be attributed to the rebound 
effect, independent of any changes to new vehicle mass, price, or 
longevity.
    The CAFE model estimates fatalities by combining the effects 
discussed above. In particular, the model estimates the fatality rate 
per billion miles VMT for each model year vehicle in the population 
(the newest of which are the new vehicles produced that model year). 
This estimate is independent of regulatory class and varies only by 
year (and not vehicle age). The estimated fatality rate is then 
multiplied by the estimated VMT (in billions of miles) for each vehicle 
in the population and the product of the change in curb weight and the 
relevant safety coefficient, as in the equation below.
[GRAPHIC] [TIFF OMITTED] TR30AP20.668

    For the vehicles in the historical fleet, meaning all those 
vehicles that are already part of the registered vehicle population in 
CY 2017, only the model year effect that determines the 
``FatalityEstimate'' is relevant. However, each vehicle that is 
simulated explicitly by the CAFE model, and is eligible to receive mass 
reduction technologies, must also consider the change between its curb 
weight and the threshold weights that are used to define safety 
classes. For vehicles above the threshold, reducing vehicle mass can 
have a smaller negative impact on fatalities (or even reduce 
fatalities, in the case of the heaviest light trucks). The 
``ChangePer100Lbs'' depends upon this difference. The sum of all 
estimated fatalities for each model year vehicle in the on-road fleet 
determines the reported fatalities, which can be summarized by either 
model year or calendar year.
b) Environmental Impacts
    Today's final rule directly involves the fuel economy and average 
CO2 emissions of light-duty vehicles, and the final rule is 
expected directly and significantly to impact national fuel consumption 
and CO2 emissions. Fuel economy and CO2 emissions 
are closely related, so that it is expected the impacts on national 
fuel consumption and national CO2 emissions will track in 
virtual lockstep with each other.
    Today's final rule does not directly involve pollutants such as 
carbon monoxide, smog-forming pollutants (nitrogen oxides and unburned 
hydrocarbons), fine particulate matter, or ``air toxics'' (e.g., 
formaldehyde, acetaldehyde, benzene). While today's final rule is 
expected to impact such emissions indirectly (by reducing travel demand 
and accelerating fleet turnover to newer and cleaner vehicles on one 
hand while, on the other, increasing activity at refineries and in the 
fuel distribution system), it is expected that these impacts will be 
much smaller than impacts on fuel use and CO2 emissions 
because standards for these other pollutants are independent of those 
for CO2 emissions.
    Following decades of successful regulation of criteria pollutants 
and air toxics, modern vehicles are already vastly cleaner than in the 
past, and it is expected that new vehicles will continue to improve. 
For example, the following chart shows trends in new vehicles' emission 
rates \2434\ for volatile organic compounds (VOCs) and nitrogen oxides 
(NOX)--the two motor vehicle criteria pollutants that 
contribute to the formation of smog.
---------------------------------------------------------------------------

    \2434\ The emission rate is the rate at which a vehicle emits a 
given pollutant into the atmosphere. Tailpipe emission rates are 
expressed on a gram per mile basis. For example, driving 15,000 
miles in a year, a vehicle with a 0.4 g/mi NOX emission 
rate would emit 6,000 grams of NOX.

---------------------------------------------------------------------------

[[Page 25040]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.669

    Because new vehicles are so much cleaner than older models, it is 
expected that under any of the alternatives considered here for fuel 
economy and CO2 standards, emissions of smog-forming 
pollutants would continue to decline nearly identically over the next 
two decades. The following chart shows estimated total fuel 
consumption, CO2 emissions, and smog-forming emissions under 
the baseline and new final standards (CAFE standards--trends for 
CO2 standards would be very similar), normalized to 2017 
levels in order to allow the three to be shown together on a single 
chart:

[[Page 25041]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.670

    The following table summarizes relative differences between the 
baseline/augural and final standards:
[GRAPHIC] [TIFF OMITTED] TR30AP20.671

    As indicated, the agencies' analysis indicates that through 2050, 
increases in annual light-duty fuel consumption and CO2 
emissions would remain below 10 percent, and increases in annual light-
duty emissions of smog-forming pollutants would remain below 2.5 
percent.
    As the analysis affirms, while fuel economy and CO2 
emissions are two sides (or, arguably, the same side) of the same coin, 
fuel economy and CO2 are only incidentally related to 
pollutants such as smog, and any positive or negative impacts of 
today's rulemaking on these other air quality problems would most 
likely be far too small to observe.

[[Page 25042]]

    The remainder of this section summarizes the impacts on fuel 
consumption and emissions for both the new final CAFE standards and the 
new final CO2 standards.
(1) Understanding Energy and Environmental Impacts
    Today's rulemaking and accompanying FRIA and FEIS all examine a 
range of physical impacts. These impacts reflect the combined effect of 
a range of different factors, some of which are independent of one 
another, and some of which interact. The scope and nature of this set 
of factors is such that, even among knowledgeable experts, intuition is 
often uninformative or even misleading.
    On one hand, it is reasonable to be confident that the more CAFE 
and CO2 standards are relaxed, the more national-scale fuel 
consumption and CO2 emissions will increase, because the 
standards apply directly to the average rates at which new vehicle 
consume fuel and, in turn, emit CO2. While other factors--
including some that work against this expectation--are involved, these 
other factors are insufficient to belie this basic expectation that 
less stringent standards will lead to increased fuel consumption and 
CO2 emissions.
    On the other hand, while it is intuitive to expect that the 
increased fuel consumption should lead to some additional emissions to 
produce and distribute fuel, those processes are expected to become 
cleaner over time, and refineries may respond by reducing exports of 
petroleum products rather than increasing overall activity. Although 
many believe that more fuel-efficient vehicles are, by definition, 
``cleaner,'' most pollutants impacting air quality are regulated on an 
average per-mile basis, such that vehicles' ``cleanliness'' is 
effectively independent from vehicles' fuel economy.\2435\ However, 
because emissions standards relevant to air quality are so much more 
stringent than in the past, and because some emission control 
technologies (e.g., catalytic converters) tend to deteriorate as 
vehicles age, average emission rates of vehicles are very dependent on 
when those vehicles were produced and how old they are. This means that 
total vehicular emissions of pollutants impacting air quality depend 
not directly on fuel economy, but rather on the amount of highway 
travel (since emissions are regulated on a per-mile basis) and on how 
that travel is distributed among older and newer vehicles. The agencies 
estimate that relaxing CAFE and CO2 standards will, by 
decreasing the price and fuel economy levels of vehicles produced after 
MY 2017, lead to changes in the quantities of new vehicles produced and 
sold in the U.S., as well as changes in fleet mix (i.e., the relative 
shares of passenger cars and light trucks, which are subject to 
different emissions standards), and changes in the rates at which older 
vehicles are removed from service (i.e., scrapped). Is it reasonable to 
expect that less stringent standards will necessarily accelerate the 
turnover to newer, cleaner vehicles? Does that depend on fuel prices? 
Yet another factor involves the prevalence of electric vehicles, which 
emit no air pollutants directly, but do use electricity. How might that 
electricity be generated in the future? Also, does it necessarily 
follow that less stringent CAFE and CO2 standards will 
reduce the sale of battery electric vehicles (BEVs) in the long term? 
Could less stringent standards increase long-term BEV sales if 
manufacturers are able to make early investments in BEV research and 
development, or wait for the costs of BEV systems to decline, rather 
than making larger nearer-term commitments to, say, very advanced 
engine technologies? With air quality depending on how emissions of 
various pollutants are impacted (and sometimes in different ways) by 
these factors, there is scant basis for a priori expectations regarding 
the direction, much less the magnitude of air quality impacts under the 
various regulatory alternatives.
---------------------------------------------------------------------------

    \2435\ For example, in 42 U.S.C. 7521(g), the 1990 Clean Air Act 
Amendments defined specific numerical standards for passenger car 
and light truck CO, NMHC (i.e., VOC), and NOx emission rates, and 
defined them on a gram per mile basis, such that the 3-cylinder 1993 
Geo Metro and the 12-cylinder 1993 Ferrari 512 were both regulated 
to 0.4 grams per mile of NOx, even though the Metro's average fuel 
economy rating, at 47 mpg, was more than four times greater than the 
Ferrari's 11 mpg rating.
---------------------------------------------------------------------------

    Although, like any other model, the CAFE model involves many 
uncertainties and does not account for every possible factor or 
interaction, the model does enable the agencies to estimate emissions 
impacts accounting for the factors mentioned above, and specific 
results can be understood through careful examination of model inputs, 
outputs, and methods. To illustrate this, the agencies consider 
estimated emissions of nitrogen oxides (NOX), a class of 
pollutants that contribute to the formation of ground-level ozone 
(i.e., smog) that is harmful to public health and welfare. The agencies 
apply the same ``unconstrained'' modeling approach as underlies the 
FEIS. Graphing estimated annual tailpipe, upstream, and combined total 
NOX emissions from passenger cars and light trucks shows 
emissions declining significantly over time, with results from the 
various action alternatives (focusing here on the least stringent, 
preferred, and most stringent alternatives, and applying the same 
vertical scale to all three charts) being virtually indistinguishable 
from the no-action alternative:
BILLING CODE 4910-59-P

[[Page 25043]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.672


[[Page 25044]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.673


[[Page 25045]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.674

BILLING CODE 4910-59-C
    Closer examination, though, reveals that although differences are 
very small on a relative scale, they do exhibit definitive trends. 
Reducing stringency causes total annual tailpipe NOX 
emissions to decline initially, as scrappage of older higher-emitting 
vehicles is accelerated and sales of new vehicles increase slightly 
relative to augural standards. Over time, both of these trends are 
impacted by steadily increasing fuel prices, but more important, 
reducing stringency causes the market to shift somewhat more slowly to 
electric vehicles than under the augural standards. Because electric 
vehicles emit no NOX directly, the impact on NOX 
emissions of this dampening of electric vehicle sales eventually 
outweighs the other impacts, such that by approximately 2035, less 
stringent standards begin increasing annual tailpipe NOx emissions 
rather than decreasing these emissions (relative to the augural 
standards):

[[Page 25046]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.675

    On the other hand, at least through 2050, less stringent standards 
show increased upstream NOX emissions. These increases 
continue to build through the late 2030s, as total fuel consumption 
under the less stringent standards continues to increase relative to 
levels under the augural standards. However, by 2040, these increases 
are steadily shrinking, due to the same delayed shift to electric 
vehicles:

[[Page 25047]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.676

    Model outputs indicate that on a per-mile basis, upstream 
NOX emissions beyond 2030 are 2-24 percent greater for 
electricity than for gasoline, varying over time and between regulatory 
alternatives. (Although the agencies have applied the same upstream 
emission factors to all regulatory alternatives, comparative per-mile 
upstream emissions also depend on comparative vehicle efficiency.) This 
means that, although a shift to electrification reduces tailpipe 
emissions, it also tends to increase net upstream emissions.
    Taken together, these changes in tailpipe emissions produce very 
slight decreases in overall annual NOX emissions through 
about 2026 under each regulatory alternative. Beyond 2026, the 
regulatory action alternatives all produce increased overall annual 
NOX emissions relative to the augural standards, although 
for the most stringent regulatory alternative considered here, these 
increases plateau after about 2040:

[[Page 25048]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.677

    Still, although trends and differences between regulatory 
alternatives are clear on the scale of the last three of the above 
charts, the preceding three charts place these emissions changes in 
context, and show that they are barely discernable. For example, the 
largest increase shown in the last of the above charts is about 0.015 
million tons, in 2050, when total emissions are 0.33-0.35 million tons, 
down from about 1.5 million tons in 2017. In other words, the largest 
increase in overall annual NOX emissions is only about 1 
percent of recent annual NOX emissions attributable to 
passenger cars and light trucks.
    The FEIS accompanying today's rulemaking presents tailpipe, 
upstream, and total emissions for a range of pollutants, and presents 
results of photochemical modeling to estimate corresponding changes in 
air quality, as well as results of calculations to estimate resultant 
health impacts. As indicated by the following chart, at least for the 
final standards, VOC and PM emissions follow overall trends broadly 
similar to those followed by NOX emissions, although, 
relative to recent (2017) total emissions attributable to passenger 
cars and light trucks, changes in VOC and PM emissions are not as small 
as changes in NOX emissions. Under the final standards, 
combined tailpipe and upstream CO emissions are very slightly lower 
than under the augural standards through the early 2030s, after which 
these emissions changes begin increasing at rates similar to those for 
VOC, NOX, and PM. CO2 emissions changes exhibit 
the expected trend mentioned above, with combined tailpipe and upstream 
emissions steadily increasing under the final standards. However, the 
final standards lead combined tailpipe and upstream SO2 
emissions to decrease relative to the augural standards, and as a share 
of 2017 emissions, these decreases grow from about 2 percent in 2035 to 
about 10 percent in 2050:

[[Page 25049]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.678

    As indicated by the following chart, changes in tailpipe 
SO2 emission follow trends nearly identical to those 
followed by changes in CO2 emissions, because both result 
directly from the quantity and composition (sulfur and carbon per 
gallon, respectively) of fuel consumed:

[[Page 25050]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.679

    This means that the decreases in overall SO2 emissions 
must be attributable to decreases in upstream SO2 emissions. 
The following chart shows SO2 emissions decreases becoming 
steadily larger after the mid-2030s, suggesting that, as discussed 
above, delaying the shift to electric vehicles leads to delays in 
emissions from electricity generation, and for some pollutants (notably 
below, SO2 and CO2), these emissions from 
electricity generation are large enough to reverse trends in overall 
emissions changes. For SO2, this reflects, among other 
things, the fact that, in order to enable catalytic converters to 
operate more efficiently, gasoline in sulfur is now limited to an 
average of 10 parts per million.\2436\
---------------------------------------------------------------------------

    \2436\ See https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-motor-vehicles-tier-3.

---------------------------------------------------------------------------

[[Page 25051]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.680

    Again, the FEIS accompanying today's rulemaking further explores 
changes in emissions; the purpose of this discussion is not to 
duplicate material appearing in the FEIS, but rather to discuss some of 
the underlying factors and how they can lead to some of the trends 
reported in the FEIS.
    Unlike the FEIS, today's rulemaking and accompanying FRIA largely 
examine impacts on a ``model year basis.'' As discussed below, while a 
calendar year basis involves considering impacts in one or a series of 
calendar years, a model year basis involves considering impacts over 
the useful lives of vehicles produced in one or over a series of model 
years. A calendar year approach answers the question ``what do we 
estimate will happen in, for example, 2035?,'' and a model year 
approach answers the question ``what impacts do we estimate will be 
attributable to vehicles produced in 2025?'' The calendar approach does 
not extend beyond 2050, the last year in which the analysis includes a 
complete on-road fleet. On the other hand, while it accounts for model 
year 2050 vehicles' fuel consumption and emissions through 2089, the 
model year approach as implemented here does not extend beyond model 
year 2029.
    These are differences in temporal perspective that, for some types 
of impacts, lead to differences in reported trends. For example, 
returning to tailpipe NOX emissions, Figure VII-6 (using the 
calendar year perspective) shows that relaxing the stringency of CAFE 
standards leads annual tailpipe NOX emissions to increase 
starting around 2035, but leads these emissions to decrease in the 
nearer term. As discussed above, this shift can be attributed to the 
less stringent standards leading to a delayed shift toward electric 
vehicles. Because the model year perspective as implemented here 
extends through 2029, it largely sets aside this shift to electric 
vehicles, even for the ``unconstrained'' modeling underlying the FEIS 
(modeling which, unlike the ``standard setting'' type of analysis 
required by EPCA, considers that, even during 2018-2029, additional 
electric vehicles might be produced in response to standards). 
Consequently, unlike the calendar year perspective as applied beyond 
2035, the model year perspective that extends through MY 2029 always 
shows tailpipe NOX emissions decreasing as the stringency of 
CAFE standards is relaxed relative to the augural standards.
    In addition to this difference in temporal perspective, the FEIS, 
relative to the rulemaking and FRIA, applies a perspective that is 
different in terms of how manufacturers could respond to standards. The 
``unconstrained'' modeling underlying the FEIS allows for the potential 
that manufacturers might apply CAFE compliance credits or introduce 
additional electric vehicles in any model year. This is intended to 
reflect how manufacturers might respond to standards in the real world. 
However, EPCA requires that, for purposes of determining the maximum 
feasible standards, NHTSA set aside the potential that manufacturers 
might apply credits or increase electric vehicle offerings in the model 
years under consideration. Therefore, for CAFE, the preamble and FRIA 
use modeling that sets aside the potential use of credits

[[Page 25052]]

and the potential introduction of new electric vehicles through 2029 
(although, since standards prior to MY 2021 are not subject to 
reconsideration, this modeling does consider the potential use of 
credits through MY 2020). As indicated by the following chart, 
especially prior to model year 2030, this leads to significant 
differences in EV market penetration between the two types of analyses:
[GRAPHIC] [TIFF OMITTED] TR30AP20.681

    Over time, these differences in EV sales lead to significant 
differences in the steadily accumulating share of overall highway 
travel powered with electricity:

[[Page 25053]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.682

    For most pollutants, the fact that EVs do not emit air pollutants 
outweighs the fact that combustion-based power plants do. As discussed 
above, sulfur content in gasoline is so low that the opposite is the 
case for net SO2 emissions.
    A complete quantitative analysis of differences between calendar 
year-based emissions trends shown in the FEIS and model year-based 
emissions trends shown in the rulemaking and FRIA would involve 
examination of all of the factors mentioned above. However, considering 
the temporal difference in perspective between the two types of 
analyses, and considering the differences in the timing and pace of the 
estimated transition to electric vehicles, differences in emissions 
trends are inevitable.
(2) CO2 Damages
    Section V discusses, among other things, the need of the Nation to 
conserve energy, providing context for the estimated impacts on 
national-scale fuel consumption summarized below. Corresponding to 
these changes in fuel consumption, the agencies estimate that today's 
final rule will impact CO2 emissions. CO2 is one 
of several gases that absorb infrared radiation, thereby trapping heat 
and potentially making the planet warmer. The most important such gases 
directly emitted by human activities include carbon dioxide 
(CO2), methane (CH4), nitrous oxide 
(N2O), and several fluorine-containing halogenated 
substances. Although CO2, CH4, and N2O 
occur naturally in the atmosphere, human activities have changed their 
atmospheric concentrations. From the pre-industrial era (i.e., ending 
about 1750) to 2016, concentrations of these gases have increased 
globally by 44, 163, and 22%, respectively.\2437\ The FEIS accompanying 
today's rulemaking discusses the potential impacts of the emission of 
such gases at greater length, and also summaries analysis quantifying 
some of these impacts (e.g., average temperatures) for each of the 
considered regulatory alternatives.
---------------------------------------------------------------------------

    \2437\ Impacts and U.S. emissions of CO2 are 
discussed at greater length in EPA's 2018 ``Inventory of U.S. 
Greenhouse Gas Emissions and Sinks,'' EPA 430-R-18-003 (Apr. 12, 
2018), available at https://www.epa.gov/sites/production/files/2018-01/documents/2018_complete_report.pdf.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 25054]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.683


[[Page 25055]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.684

BILLING CODE 4910-59-C

[[Page 25056]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.685

(3) Other Pollutant Damages--Criteria and Toxic Pollutants
    The CAFE model uses the entire on-road fleet, calculated VMT 
(discussed above), and emissions factors (which are an input to the 
CAFE model, specified by model year and age) to calculate tailpipe 
emissions associated with a given alternative. Just as it does for 
additional CO2 emissions associated with upstream emissions 
from fuel production, the model captures criteria pollutants that occur 
during other parts of the fuel life cycle. While this is typically a 
function of the number of gallons of gasoline consumed (and miles 
driven, for tailpipe criteria pollutant emissions), the CAFE model also 
estimates electricity consumption and the associated upstream emissions 
(resource extraction and generation, based on U.S. grid mix).
(a) Emissions Increases

[[Page 25057]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.686

BILLING CODE 4910-59-P

[[Page 25058]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.687


[[Page 25059]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.688


[[Page 25060]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.689


[[Page 25061]]


BILLING CODE 4910-69-C
(b) Air Quality Impacts of Other Pollutants
    Although this final rule focuses on standards for fuel economy and 
CO2, it will also have an impact on criteria and air toxic 
pollutant emissions, although as discussed above, it is expected that 
incremental impacts on criteria and air toxic pollutant emissions would 
be too small to observe under any of the regulatory alternatives under 
consideration. Nevertheless, the following sections detail the criteria 
pollutant and air toxic inventory impacts of this final rule; the 
methodology used to calculate those impacts; the health and 
environmental effects associated with the criteria and toxic air 
pollutants that are being impacted by this final rule; the potential 
impact of this final rule on concentrations of criteria and air toxic 
pollutants in the ambient air; and other unquantified health and 
environmental effects.
    Today's analysis reflects the combined result of several underlying 
impacts, all discussed above. CAFE and CO2 standards are 
estimated to impacts new vehicle prices, fuel economy levels, and 
CO2 emission rates. These changes are estimated to impact 
the size and composition of the new vehicle fleet and to impact the 
retention of older vehicles (i.e., vehicle survival and scrappage) that 
tend to have higher criteria and toxic pollutant emission rates. Along 
with the rebound effect, these lead to changes in the overall amount of 
highway travel and the distribution among different vehicles in the on-
road fleet. Vehicular emissions depend on the overall amount of highway 
travel and the distribution of that travel among different vehicles, 
and emissions from ``upstream'' processes (e.g., petroleum refining, 
electricity generation) depend on the total consumption of different 
types of fuels for light-duty vehicles.
(i) Impacts
    As discussed above, in addition to affecting fuel consumption and 
emissions of carbon dioxide or its equivalent, this rule would also 
influence other pollutants, i.e., ``criteria'' air pollutants and their 
precursors, and air toxics. The final rule would affect emissions of 
carbon monoxide (CO), fine particulate matter (PM2.5), 
sulfur dioxide (SOX), volatile organic compounds (VOC), 
nitrogen oxides (NOX), benzene, 1,3-butadiene, formaldehyde, 
acetaldehyde, and acrolein. Consistent with the evaluation conducted 
for the Environmental Impact Statement accompanying today's rule, the 
agency analyzed criteria air pollutant impacts in 2025, 2035, and 2050 
(as a representation of future program impacts). Estimates of these 
other emission impacts are shown by pollutant in Table VII-124 through 
Table VII-127 and are broken down by the two drivers of these changes: 
a) ``downstream'' emission changes, reflecting the estimated effects of 
VMT rebound (discussed in Section VIII of the FRIA), changes in vehicle 
fleet age, changes in vehicle emission standards, and changes in fuel 
consumption; and b) ``upstream'' emission increases because of 
increased refining and distribution of motor vehicle gasoline relative 
to the baseline. Program impacts on criteria and toxics emissions are 
discussed below.
    As discussed above, these changes in total annual criteria 
pollutant emissions attributable to passenger cars and light trucks 
reflect trends in both vehicular and upstream emissions, and these 
trends can either be mutually reinforcing or mutually offsetting, 
depending on the pollutant and year. Above, Figure VII-9 places these 
total changes in emissions in context, showing that, except for 
SO2, these changes in criteria pollutant emissions are very 
small. For SO2 emissions, changes are also very small 
through the late 2030s, after which reduced upstream emissions cause 
net emission reductions to exceed 10 percent of 2017 emissions by 2050.
BILLING CODE 4910-59-P

[[Page 25062]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.690


[[Page 25063]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.691


[[Page 25064]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.692


[[Page 25065]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.693

BILLING CODE 4910-59-C
    As shown in Table VII-128 through Table VII-131, it is estimated 
that the new final program would result in small changes for air toxic 
emissions

[[Page 25066]]

compared to total U.S. inventories across all sectors. These changes 
also reflect the changing balance between vehicular and upstream 
emissions.
BILLING CODE 4910-59-P

[[Page 25067]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.694


[[Page 25068]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.695


[[Page 25069]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.696


[[Page 25070]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.697


[[Page 25071]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.698

BILLING CODE 4910-59-C
    Changes in emissions of other pollutants due to these rules will 
impact air quality. Information on current air quality and the results 
of our air quality

[[Page 25072]]

modeling of the projected impacts of these rules are summarized in the 
following section.
(ii) Other Unquantified Health and Environmental Effects
    In the proposal, the agencies sought comment on whether there are 
any other health and environmental impacts associated with advancements 
in technologies that should be considered. For example, the use of 
technologies and other strategies to reduce fuel consumption and/or 
CO2 emissions could have effects on a vehicle's life-cycle 
impacts (e.g., materials usage, manufacturing, end of life disposal), 
beyond the issues regarding fuel production and distribution (upstream) 
CO2 emissions discussed in Section VI.D.2. The agencies 
sought comment on any studies or research in this area that should be 
considered in the future to assess a fuller range of health and 
environmental impacts from the light-duty vehicle fleet shifting to 
different technologies and/or materials. At this point, the agencies 
find there is insufficient information about the lifecycle impacts of 
the myriad of available technologies, materials, and cradle-to-grave 
pathways to conduct the type of detailed assessments that would be 
needed in a regulatory context, especially considering the 
characterization of specific vehicles in the analysis fleet and the 
characterization of specific technology options.
(c) Health Effects of Other Pollutants
    This section presents results of the analysis showing health 
effects associated with exposure to some of the criteria and air toxic 
pollutants impacted by the new final vehicle standards. As discussed 
above, the health impacts presented here are subject to a number of 
uncertainties, some of which arise from the less complex benefits-per-
ton approach relied on in this analysis, and some of which arise from 
the uncertainty surrounding many of the assumptions and other inputs 
relied on in the agencies' analysis. As the agencies conclude above, 
although it may seem that the agencies' estimates of increases in 
premature mortality resulting from the final standards are more likely 
to be too high than too low, it is extremely difficult to anticipate 
whether this is actually the case.
BILLING CODE 4910-59-P

[[Page 25073]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.699


[[Page 25074]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.700


[[Page 25075]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.701


[[Page 25076]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.702


[[Page 25077]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.703


[[Page 25078]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.704


[[Page 25079]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.705


[[Page 25080]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.706


[[Page 25081]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.707


[[Page 25082]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.708


[[Page 25083]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.709


[[Page 25084]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.710


[[Page 25085]]


BILLING CODE 4910-59-C

B. Impacts on Calendar Year Basis

    As with the NPRM, the agencies' analysis primarily examines 
regulatory impacts on a model year basis, accounting for the physical 
impacts and monetized costs and benefits attributable to vehicles 
produced prior to model year 2030 and occurring throughout these 
vehicles' useful lives. EDF submitted comments arguing that the 
agencies should examine impacts on a calendar year basis, as discussed 
above in VI.A.\2438\ CAFE analysis has historically examined effects of 
the standards on a model year basis, because CAFE (and CO2) 
standards are enforced on a model year basis, and manufacturers' 
responses to these standards (i.e., their costs), which are the direct 
effects of the standards, occur on a model year basis. On the other 
hand, overall impacts on national energy consumption and the 
environment result from the evolution and operation of the overall on-
road fleet, and this motivates consideration of results on a calendar 
year basis. As also discussed in VI.A., the agencies have expanded the 
presentation of results in today's rulemaking and FRIA by presenting 
some impacts for each of CYs 2017-2050 and, to enable doing so, have 
extended the analysis to cover model years through 2050.
---------------------------------------------------------------------------

    \2438\ EDF, NHTSA-2018-0067-12108, Appendix A at 9, et seq., and 
Appendix B at 11-14.
---------------------------------------------------------------------------

    For this analysis, the CAFE model reports impacts for each model 
year through 2050, and, to capture the entire useful lives of these 
vehicles, for each of calendar years 2017-2089.\2439\ One way to 
illustrate the model's outputs is to consider three cohorts of model 
years: MYs 1978-2017 (MYs to which the analysis applies no additional 
fuel-saving technology), MYs 2018-2029 (MYs included in both the ``MY 
basis'' and ``CY basis'' approaches), and MYs 2030-2050 (MYs included 
only the ``CY basis'' approach). On a calendar year basis, impacts of 
the final standards on annual CO2 emissions (impacts on fuel 
consumption would follow essentially the same trends) may be attributed 
to these cohorts as follows:
---------------------------------------------------------------------------

    \2439\ As for the NPRM, DOT has made the model and all inputs 
and outputs for today's analysis available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system. The model documentation available at the same location 
explains, among other things, the structure and contents of each 
type of input and output file. The 
``annual_societal_effects_report.csv ``and 
``annual_societal_costs_report.csv'' reports contain, respectively, 
estimates of physical impacts and monetized costs and benefits 
attributable to each model year in each calendar years. Other output 
file types contain corresponding aggregations either all calendar 
years, or across all model years.
[GRAPHIC] [TIFF OMITTED] TR30AP20.711

BILLING CODE 4910-59-P
    Here, the large lower area of the chart shows annual CO2 
emissions estimated to occur under the baseline/augural CAFE standards, 
through calendar year

[[Page 25086]]

2089, which is the last year any MY 2050 vehicles are estimated still 
to be on the road. The steady declines through 2050 reflect turnover to 
more efficient vehicles produced under either regulatory alternative, 
and the steep decline after 2050 reflects vehicles included in the 
analysis being removed from service. Of the increased annual emissions 
under the final standards, the black area shows the portion 
attributable to vehicles produced during MYs 2018-2029, and the topmost 
area shows the portion attributable to vehicles produced during MYs 
2030-2050. The final standards are estimated to reduce emissions from 
vehicles produced during MYs 1978-2017 by accelerating scrappage of 
these vehicles, but these changes are too small to be visible in this 
chart.
    The bulk of the reporting of results here and in the FRIA examines 
impacts over the useful lives of vehicles produced prior to MY 2030. In 
terms of the above chart, this means excluding the topmost area, 
producing the following:
[GRAPHIC] [TIFF OMITTED] TR30AP20.712

    On the other hand, calendar year accounting, as considered for this 
analysis, includes all model years included in the analysis (i.e., 
through MY 2050), and examines impacts in all calendar years for which 
a full on-road fleet is simulated. In terms of the first of the above 
charts, this means ``cutting off'' results at calendar year 2050:

[[Page 25087]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.713

    Here, the horizontal axis extends through 2089 to make clear that 
this calendar year accounting involves excluding emissions impacts over 
most of the useful lives of the latest model years included in the 
analysis. On a scale covering just those calendar years included in the 
calendar year analysis, the same chart appears as follows:

[[Page 25088]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.714

    Viewed on the same calendar year basis, technology costs appear as 
follows, with differences between costs under the baseline/augural 
standards and under the final standards shown as amounts by which the 
former exceed the latter (e.g., in 2025, the final standards are 
estimated to avoid about $19 billion in technology costs that would 
have been incurred under the baseline/augural standards):

[[Page 25089]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.715

    Present value analysis considered involves discounting all 
estimated future costs and benefits to 2019. At a 7 percent discount 
rate, the undiscounted technology costs shown above correspond to 
discounted costs shown in the following chart:

[[Page 25090]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.716

    Without discounting, therefore, the final standards avoid $457 
billion in technology costs through 2050, each additional year of 
analysis after 2036 adding about $14 billion to that total. At a 7 
percent discount rate, the final standards still avoid $183 billion in 
technology costs, while incremental amounts attributable to each 
additional year of analysis are (of course) lower than the undiscounted 
amounts--declining to about $5 billion during 2035-2036 and, by 2045, 
about $2 billion.
    For each of the regulatory alternatives considered here, the 
following tables summarize results of such aggregations for each 
reported category of monetized costs and benefits. The first three 
tables focus on the final CAFE standards, presenting total amounts 
through 2050 at 3 percent and 7 percent discount rates. The second 
three tables show results for corresponding CO2 standards.

[[Page 25091]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.717


[[Page 25092]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.718


[[Page 25093]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.719


[[Page 25094]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.720


[[Page 25095]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.721


[[Page 25096]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.722


[[Page 25097]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.723


[[Page 25098]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.724

    As illustrated above, the model year analysis answers the question 
``what impacts do we think might eventually be attributable to vehicles 
produced before 2030?,'' and the calendar year analysis answers the 
question ``what do

[[Page 25099]]

we think might happen between now and 2050?'' Again, CAFE and 
CO2 standards are enforced on a model year basis, and the 
agencies accordingly simulate manufacturers' responses to these 
standards--and estimate manufacturers' corresponding costs--on a model 
year basis. This motivates consideration of results on a model year 
basis. On the other hand, overall impacts on national energy 
consumption and the environment result from the evolution and operation 
of the overall on-road fleet, and this motivates consideration of 
results on a calendar year basis.
    These different perspectives produce results that, without careful 
consideration, appear to conflict. The model year perspective as 
applied through MY 2029 shows less stringent standards producing 
environmental benefits (compared to the augural standards) attributable 
to the aggregate of vehicles produced prior to MY 2030. While the 
calendar year perspective also shows similar trends prior to (calendar 
year) 2035, with the estimated transition to electric vehicles 
accelerating over time, the calendar year perspective shows less 
stringent standards mostly increasing emissions (SO2 being 
an exception) relative to the augural standards.
    Still, some important aspects of estimated social benefits and 
costs are common to both the model year and calendar year perspectives. 
For each of the regulatory action alternatives, the magnitude of total 
incremental benefits (relative to the baseline augural standards) is 
similar to the magnitude of total incremental costs. This stands in 
marked contrast to the agencies' 2012 rulemaking announcing the augural 
standards, and finding of estimated benefits that were 3-4 times larger 
than costs.\2440\ Under today's analysis, estimated benefits and costs 
are instead of similar magnitude, with estimated net benefits, by 
comparison, small enough to be even directionally uncertain, such that 
an alternative estimated to produce small positive net benefits under 
one perspective and applying a 7 percent discount rate might be 
estimated to produce small negative net benefits under the other 
perspective and/or applying a 3 percent discount rate. While the 
agencies obviously must consider benefits, costs, and net benefits, our 
decisions are based on wider considerations. Consistent with the 
agencies' 2012 final rule, today's final rule finds--from both the 
model year and calendar year perspectives--that forgone fuel savings 
(forgone because today's final rule involves relaxing rather than 
increasing the stringency of CAFE and CO2 standards) account 
for the bulk of estimated forgone social benefits. These are private 
benefits, which raises a significant question of whether there is a 
meaningful market failure that needs to be addressed by more stringent 
regulation.
---------------------------------------------------------------------------

    \2440\ 77 FR at 62629 (Oct. 15, 2012).
---------------------------------------------------------------------------

    Section VI contains an extensive discussion and analysis of the 
existence and nature of various market failures related to fuel economy 
standards. These potential market failures include the well-established 
externalities of environmentally harmful emissions, congestion, and 
safety; as well the debatable and hypothetical market failures related 
to the ``energy paradox.'' The energy paradox refers to an observation 
that some consumers appear voluntarily to forgo investments in energy 
conservation even when those initial investments appear to repay 
themselves--in the form of savings in energy costs--over the relatively 
near term. Section VI.D.1 discussion casts doubt on the theoretical 
underpinnings that the energy paradox represents a market failure, 
discusses recent research that suggests the extent consumers are 
undervaluing fuel economy has been overstated, and suggests the 
analysis supporting claims of an energy paradox overlooks the 
opportunity costs of other vehicle attributes that consumers and 
manufacturers trade off with fuel efficiency technology. As stated in 
Section VI, while the agencies have reservations about the extent to 
which a market failure capable of driving very large net private 
financial harm to consumers exists, the agencies do not take a position 
on the existence of an energy paradox in this rulemaking.
    The primary analysis shows that the CAFE final rule would generate 
$12.9 billion in total social net benefits using a 7 percent discount 
rate, but without the large net private loss of $26.4 billion, the net 
social benefits would equal the external net benefits, or $39.3 
billion. Therefore, given significant questions about whether 
government action to impose restrictions in private markets could 
improve net social benefits absent a market failure, if no market 
failure exists to motivate the $26.4 billion in private losses to 
consumers, the net benefits of these final standards would be $39.3 
billion. The CY analysis produces similar results, though the estimated 
private losses are exacerbated relative to the external gains. The CY 
analysis shows the CAFE final rule would generate -$6 billion in total 
net social benefits using a 7 percent discount rate, but without the 
large net private loss of $65 billion, the net social benefits would 
equal the external net benefits of $59 billion.
    One commenter suggested that the agencies should elect to use CY 
accounting in the primary analysis because the MY accounting approach 
resulted in an inconsistent accounting of costs and benefits owing to 
the scrappage effect. While the CY accounting approach does reduce non-
rebound safety benefits from $9 billion to $8 billion (combined fatal 
and non-fatal benefits), the total external net benefits of the rule 
actually increase by $20 billion using the CY approach. This result is 
driven primarily by a significant increase in congestion cost savings 
from less rebound driving, from $44 billion to $69 billion. Any changes 
in the net benefits in the opposite direction using CY accounting 
result from increased net private costs to consumers own financial 
wellbeing from allowing more consumer choice. These increased net 
private costs occur because the CY analysis captures model years far 
into the future, which are more uncertain and not subject to today's 
CAFE final rule. Therefore, the agencies see little evidence that the 
inconsistency suggested by the commenter is important, or that the 
primary conclusions of the analysis are meaningfully influenced by it.

Sensitivity Analysis

    As discussed at the beginning of this section, results presented 
today reflect the agencies' best judgments regarding many different 
factors. Based on analyses in past rulemakings, the agencies recognize 
that some analytical inputs are especially uncertain, some are likely 
to exert considerable influence over specific types of estimated 
impacts, and some are likely to do so for the bulk of the analysis. To 
explore the sensitivity of estimated impacts to changes in model 
inputs, analysis was conducted using alternative values for a range of 
different inputs. Results of this sensitivity analysis are summarized 
in the Final Regulatory Impact Analysis (FRIA) accompanying today's 
rulemaking, and detailed model inputs and outputs are available on 
NHTSA's website.\2441\ The following table lists the cases included in 
the sensitivity analysis.
---------------------------------------------------------------------------

    \2441\ The CAFE model and all inputs and outputs supporting 
today's rulemaking are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 25100]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.725


[[Page 25101]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.726


[[Page 25102]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.727

BILLING CODE 4910-59-C

VIII. How do the final standards fulfill the agencies' statutory 
obligations?

A. How Does the technical assessment support the final CO2 
standards as compared to the alternatives that EPA has considered?

1. Introduction
    Title II of the Clean Air Act provides for comprehensive regulation 
of mobile sources, authorizing EPA to regulate emissions of air 
pollutants from all mobile source categories. Under Section 202(a) and 
relevant case law, as discussed below, EPA considers such issues as 
technology emission reduction effectiveness, its cost (both per 
vehicle, per manufacturer, and per consumer), the lead time necessary 
to implement the technology, and based on this the feasibility of 
potential standards; the impacts of potential standards on emissions 
reductions of both GHGs and non-GHGs; the impacts of standards on oil 
conservation and energy security; the impacts of standards on fuel 
savings by consumers; the impacts of standards on the auto industry; 
other energy impacts; as well as other relevant factors such as impacts 
on safety.
    EPA is afforded considerable discretion under section 202(a) when 
assessing issues of technical feasibility and availability of lead time 
and in weighing these factors. In light of its consideration of the 
relevant factors, EPA has concluded, for the reasons discussed below, 
that the previous standards (which increase stringency at a rate of 
about 5% per year) are not appropriate, and the best action is to 
revise the standards to increase stringency by 1.5% per year. Beginning 
in 2009, EPA and NHTSA have worked together jointly to establish fuel 
economy and tailpipe CO2 emission standards for light duty 
vehicles. The first rulemaking, finalized in 2010, established 
standards for the 2012 through 2016 model years. Shortly thereafter, in 
2012, the agencies established standards for the 2017 through 2025 
model years--but given the limitation in EPCA that only allows for 
standards to be set five years at a time, the 2022-2025 model year 
standards were only final for EPA's tailpipe CO2 emissions 
regulation. This rapid period of rulemaking to establish standards over 
a decade in advance may have marked a departure for NHTSA, but it 
followed EPA's longstanding

[[Page 25103]]

approach when regulating vehicular criteria pollutant emissions to 
provide a significant period of time for the industry to develop 
technologies to achieve standards.
    While EPA had decades of experience regulating light duty vehicle 
emissions, it did not previously have experience regulating tailpipe 
CO2 emissions. And regulating CO2 emissions is 
quite different from regulating criteria pollutant emissions. With 
criteria pollutants, technological emission controls exist primarily in 
the form of engine controls and catalytic conversion. Today's emission 
controls for criteria pollutants have only a de minimis effect on 
performance or functionality of the vehicle.
    Controlling tailpipe CO2 emissions for an internal 
combustion engine requires controlling the amount of energy used to 
propel the vehicle. All else being equal, better performance (in 
acceleration or passing speed) requires more energy. Similarly, 
vehicles with more storage capacity tend to be larger, and moving an 
object with larger mass requires more energy than objects with smaller 
mass. Vehicles with greater towing performance likewise require more 
energy. Maintaining utility and performance requires sophisticated and 
expensive technological solutions, such as reducing mass through 
advanced materials, changing engine combustion cycles, increasing 
compression ratios, or turbo-charging the engine. Consumers often can 
feel the difference in vehicle performance as a result of these 
controls, and as will be discussed herein.
    As discussed when issuing the 2012 Final Rule, the economic and 
market assumptions underlying the standards the agencies finalized were 
crucial, and long-term projections are inherently uncertain. Upon 
review of those assumptions, such as the price of gas and the sales mix 
of pick-up trucks and sport-utility vehicles as compared to passenger 
cars, the agencies have now concluded that many of these assumptions 
have not proven to be accurate and therefore have been updated. Given 
the uncertainty about the 2012 assumptions at the time of that 
rulemaking, the agencies incorporated a mid-term evaluation process for 
EPA's 2022-2025 model year standards that would be ``collaborative, 
robust and transparent,'' and ``based on information available at the 
time of the mid-term evaluation and an updated assessment of all the 
factors considered in setting the standards and the impacts of those 
factors on the manufacturers' ability to comply.'' \2442\
---------------------------------------------------------------------------

    \2442\ 77 FR at 62633.
---------------------------------------------------------------------------

    While that process was expected to take place throughout 2017, and 
a final determination issued in the Spring of 2018, this process was 
expedited. On July 27, 2016, the agencies published a Federal Register 
notice making the public aware of the availability of a draft Technical 
Assessment Report, with comments due at the end of September 2016. On 
December 6, 2016, EPA published a notice in the Federal Register making 
the public aware of its proposed Final Determination and extensive 
Technical Support Document to keep the standards set in 2012 in place 
through the 2025 model year without change. The public was given until 
December 30, 2016 to comment on the proposed determination. Less than 
two weeks later, on January 12, 2017, EPA finalized its determination.
    Industry commenters stated that the 2017 Final Determination ``is 
the product of egregious procedural and substantive defects and EPA 
should withdraw it,'' that EPA had ``fail[ed] to provide an adequate 
period for meaningful notice and comment,'' that EPA had 
``acknowledg[ed] that the Proposed Determination adjusted a number of 
EPA assumptions in response to commenters who pointed out errors at 
earlier stages'' while stating that ``there was no need for more time 
because [it] did not include much new material,'' and that ``EPA [had] 
underestimated the burden [of the standards],'' ``EPA [made] cursory 
assertions that downplayed the impact of its mandate on auto sales and 
employment,'' and ``EPA refused to consider many of the [industry's] 
technical concerns even when supported by an outside consultant, 
asserted [industry] provided insufficient data, and then refused 
further meetings for clarification.'' \2443\
---------------------------------------------------------------------------

    \2443\ Alliance letter to Administrator Pruitt, Feb. 21, 2017, 
available at https://autoalliance.org/wp-content/uploads/2017/02/Letter-to-EPA-Admin.-Pruitt-Feb.-21-2016-Signed.pdf.
---------------------------------------------------------------------------

    In light of commenters' concerns about EPA's 2017 final 
determination, in March 2017, EPA announced its intent to reconsider 
the final determination in order to allow additional opportunity to 
hear from the public, and additional consultation and coordination with 
NHTSA in support of a national harmonized program. In August 2017, EPA 
published a notice in the Federal Register requesting comment on its 
reconsideration of the initial determination, and held a public hearing 
on the matter in September 2017. Then, in April 2018, EPA issued a 
revised final determination finding that the 2022-2025 model year GHG 
standards set in 2012 were not appropriate and a rulemaking should be 
initiated to revise the standards, as appropriate.
    In this proceeding, in order to determine what standards are 
appropriate, EPA and NHTSA sought comment on a wide range of potential 
standards--ranging from holding the 2020 standards flat through the 
2026 model year to retaining the standards finalized in 2012. Similar 
to the 2012 rulemaking, EPA considered a number of different 
alternatives--ranging from the standards finalized in 2012, to holding 
the 2020 MY standards flat through MY 2026. As in 2012, the manner in 
which different factors are weighed can yield very different result--
more stringent standards would improve CO2 emissions, reduce 
energy consumption, and save consumers fuel. Less stringent standards 
would reduce technology costs for manufacturers and save consumers in 
upfront purchase prices, enabling the fleet to turnover more quickly. 
While weighing these factors, EPA has considered compliance results 
that have been observed throughout the fleet. While the agencies have 
seen extraordinary reductions in tailpipe CO2 emissions 
since EPA has begun regulation in this area, manufacturers are 
increasingly falling short of meeting their performance targets, and 
are increasingly using acquired or earned credits to comply with 
requirements. For the 2016 model year, the overall fleet failed, for 
the first time in regulation history, to meet emission targets--
achieving 272 grams per mile, when the standard was 263 grams per 
mile.\2444\ The 2016 model year saw only five major manufacturers 
perform at or better than their CO2 footprint standards--
Honda, Hyundai, Mazda, Nissan, and Subaru. For the 2017 model year, 
only three major manufacturers--BMW, Honda, and Subaru--performed 
better than their CO2 standards, and the total fleet 
underperformed compared to the standards--achieving 263 grams per mile, 
when the fleetwide standard was 258 grams per mile.\2445\ The emissions 
averaging, credit banking and trading system was established to allow

[[Page 25104]]

manufacturers greater flexibility and lead time to address technical 
feasibility and cost without sacrificing effectiveness of the 
standards, but widespread reliance upon credits across the industry may 
raise concerns about compliance in future years, particularly since the 
more significant increases in stringency in the 2012 rulemaking have 
yet to be effective. Taken together, the agencies now believe this 
information supports the conclusion that the lead time EPA estimated 
would be sufficient to achieve compliance with the previous standards 
for MYs 2021-26, was not sufficient.
---------------------------------------------------------------------------

    \2444\ EPA Greenhouse Gas Emission Standards for Light-Duty 
Vehicles: Manufacturer Performance Report for the 2016 Model Year. 
EPA-420-R-18-002 (January 2018).
    \2445\ 2018 EPA Automotive Trends Report: Greenhouse Gas 
Emissions, Fuel Economy, and Technology since 1975, available at: 
https://www.epa.gov/automotive-trends/download-automotive-trends-report.
---------------------------------------------------------------------------

    In this action, EPA is reducing the rate of stringency increases 
from those adopted in the 2012 rulemaking in part to ensure that the 
standards remain reasonable and appropriate. As in 2012, EPA is 
deciding against selecting alternatives that are more stringent or less 
stringent than appropriate. The final rule analysis projects that the 
1.5 percent alternative would result in less significant shortfalls 
compared to more stringent alternatives, which will ease compliance 
burdens while nonetheless pushing the market beyond what it would 
demand in the absence of standards or what would be achieved with less 
stringent standards. The standards finalized today will result in 
continuing improvements compared to the 2020 model year, and are best 
viewed in the context of the larger rulemaking, as shown in the chart 
below:
BILLING CODE 4910-59-P
[GRAPHIC] [TIFF OMITTED] TR30AP20.728

BILLING CODE 4910-59-C
2. Basis for the CO2 Standards Under Section 202(a) of the 
Clean Air Act
    Title II of the Clean Air Act (CAA) provides for comprehensive 
regulation of mobile sources, authorizing EPA to regulate emissions of 
air pollutants from all mobile source categories. This rule implements 
a specific provision from Title II, section 202(a).\2446\ Section 
202(a)(1) states that ``[t]he Administrator shall by regulation 
prescribe (and from time to time revise) . . . standards applicable to 
the emission of any air pollutant from any class or classes of new 
motor vehicles or new motor vehicle engines, which in his judgment 
cause, or contribute to, air pollution which may reasonably be 
anticipated to endanger public health or welfare.'' If EPA makes the 
appropriate endangerment and cause or contribute findings, then section 
202(a) directs EPA to issue standards applicable to emissions of those 
pollutants.
---------------------------------------------------------------------------

    \2446\ 42 U.S.C. 7521(a).

---------------------------------------------------------------------------

[[Page 25105]]

    Any standards under CAA section 202(a)(1) ``shall be applicable to 
such vehicles and engines for their useful life.'' Emission standards 
set by the EPA under section 202(a)(1) are technology-based, as the 
levels chosen must be premised on a finding of technological 
feasibility. Thus, standards promulgated under section 202(a) are to 
take effect only after ``such period as the Administrator finds 
necessary to permit the development and application of the requisite 
technology, giving appropriate consideration to the cost of compliance 
within such period.'' \2447\ EPA must consider costs to those entities 
which are directly subject to the standards.\2448\ Thus, ``the 
[s]ection 202(a)(2) reference to compliance costs encompasses only the 
cost to the motor-vehicle industry to come into compliance with the new 
emission standards.'' \2449\ EPA is afforded considerable discretion 
under section 202(a) when assessing issues of technical feasibility and 
availability of lead time to implement new technology. Such 
determinations are ``subject to the restraints of reasonableness,'' 
which ``does not open the door to `crystal ball' inquiry.'' \2450\ In 
developing such technology-based standards, EPA has the discretion to 
consider different standards for appropriate groupings of vehicles 
(``class or classes of new motor vehicles''), or a single standard for 
a larger grouping of motor vehicles.\2451\
---------------------------------------------------------------------------

    \2447\ CAA section 202 (a)(2); see also NRDC v. EPA, 655 F.2d 
318, 322 (DC Cir. 1981).
    \2448\ Motor & Equipment Mfrs. Ass'n Inc. v. EPA, 627 F. 2d 
1095, 1118 (DC Cir. 1979).
    \2449\ Coalition for Responsible Regulation, 684 F.3d at 128; 
see also id. at 126-27 (rejecting arguments that EPA was required to 
consider or should have considered costs to other entities, such as 
stationary sources, which are not directly subject to the emission 
standards).
    \2450\ NRDC, 655 F.2d at 328 (quoting International Harvester 
Co. v. Ruckelshaus, 478 F.2d 615, 629 (DC Cir. 1973)).
    \2451\ NRDC, 655 F.2d at 338.
---------------------------------------------------------------------------

    Although standards under CAA section 202(a)(1) are technology-
based, they are not based exclusively on technological capability. EPA 
has the discretion, and in some instances has been specifically 
directed by Congress, to consider and weigh various factors along with 
technological feasibility, such as the cost of compliance, \2452\ lead 
time necessary for compliance, \2453\ safety,\2454\ other impacts on 
consumers,\2455\ and energy impacts associated with use of the 
technology.\2456\
---------------------------------------------------------------------------

    \2452\ See section 202(a)(2).
    \2453\ Id.
    \2454\ See NRDC, 655 F.2d at 336 n. 31.
    \2455\ Since its earliest Title II regulations, EPA has 
considered the safety of pollution control technologies. See 45 FR 
14496, 14503 (March 5, 1980). (``EPA would not require a particulate 
control technology that was known to involve serious safety 
problems. If during the development of the trap-oxidizer safety 
problems are discovered, EPA would reconsider the control 
requirements implemented by this rulemaking.'').
    \2456\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-624 
(DC Cir. 1998) (ordinarily permissible for EPA to consider factors 
not specifically enumerated in the CAA).
---------------------------------------------------------------------------

    Unlike standards set under provisions such as section 202(a)(3) and 
section 213(a)(3), EPA is not required to set technology-forcing 
standards when such standards would not be appropriate. EPA has 
interpreted a similar statutory provision, CAA section 231,\2457\ as 
follows:

    \2457\ Section 231(a)(2)(A) of the CAA provides: ``The 
Administrator shall, from time to time, issue proposed emission 
standards applicable to the emission of any air pollutant from any 
class or classes of aircraft engines which in his judgment causes, 
or contributes to, air pollution which may reasonably be anticipated 
to endanger public health or welfare.'' Section 231(a)(3) provides 
in part: ``Within 90 days after the issuance of such proposed 
regulations, he shall issue such regulations with such modifications 
as he deems appropriate. Such regulations may be revised from time 
to time.'' Sectiion 231(b) provides: ``Any regulation prescribed 
under this section (and any revision thereof) shall take effect 
after such period as the Administrator finds necessary (after 
consultation with the Secretary of Transportation) to permit the 
development and application of the requisite technology, giving 
appropriate consideration to the cost of compliance within such 
period.''
---------------------------------------------------------------------------

    While the statutory language of section 231 is not identical to 
other provisions in title II of the CAA that direct EPA to establish 
technology-based standards for various types of engines, EPA 
interprets its authority under section 231 to be somewhat similar to 
those provisions that require us to identify a reasonable balance of 
specified emissions reduction, cost, safety, noise, and other 
factors. See, e.g., Husqvarna AB v. EPA, 254 F.3d 195 (D.C. Cir. 
2001) (upholding EPA's promulgation of technology-based standards 
for small non-road engines under section 213(a)(3) of the CAA). 
However, EPA is not compelled under section 231 to obtain the 
``greatest degree of emission reduction achievable'' as per sections 
213 and 202 of the CAA, and so EPA does not interpret the Act as 
requiring the agency to give subordinate status to factors such as 
cost, safety, and noise in determining what standards are reasonable 
for aircraft engines. Rather, EPA has greater flexibility under 
section 231 in determining what standard is most reasonable for 
aircraft engines, and is not required to achieve a ``technology 
forcing'' result.\2458\
---------------------------------------------------------------------------

    \2458\ 70 FR 69664, 69676 (Nov. 17, 2005).

    This interpretation was upheld as reasonable in NACAA v. EPA.\2459\ 
CAA section 202(a), as with section 231, does not specify the degree of 
weight to apply to each factor, and EPA accordingly interprets its 
authority under section 202(a) similarly to its interpretation of 
section 231 as set forth above: EPA has discretion in choosing an 
appropriate balance among the statutory factors.\2460\
---------------------------------------------------------------------------

    \2459\ 489 F.3d 1221, 1230 (DC Cir. 2007).
    \2460\ See Sierra Club v. EPA, 325 F.3d 374, 378 (D.C. Cir. 
2003) (even where a provision is technology-forcing, the provision 
``does not resolve how the Administrator should weigh all [the 
statutory] factors in the process of finding the 'greatest emission 
reduction achievable'''); see also Husqvarna AB v. EPA, 254 F. 3d 
195, 200 (D.C. Cir. 2001) (great discretion to balance statutory 
factors in considering level of technology-based standard, and 
statutory requirement ``[to give] appropriate consideration to the 
cost of applying . . . technology'' does not mandate a specific 
method of cost analysis); Hercules Inc. v. EPA, 598 F. 2d 91, 106-07 
(D.C. Cir. 1978) (``In reviewing a numerical standard, we must ask 
whether the agency's numbers are within a `zone of reasonableness,' 
not whether its numbers are precisely right''); Permian Basin Area 
Rate Cases, 390 U.S. 747, 797 (1968) (same); Federal Power 
Commission v. Conway Corp., 426 U.S. 271, 278 (1976) (same); Exxon 
Mobil Gas Marketing Co. v. FERC, 297 F. 3d 1071, 1084 (D.C. Cir. 
2002) (same).
---------------------------------------------------------------------------

    As noted above, EPA has found that the elevated concentrations of 
greenhouse gases in the atmosphere may reasonably be anticipated to 
endanger public health and welfare.\2461\ EPA defined the ``air 
pollution'' referred to in CAA section 202(a) to be the combined mix of 
six long-lived and directly emitted GHGs: carbon dioxide 
(CO2), methane (CH4), nitrous oxide 
(N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), 
and sulfur hexafluoride (SF6). The EPA further found under 
CAA section 202(a) that emissions of the single air pollutant defined 
as the aggregate group of these same six greenhouse gases from new 
motor vehicles and new motor vehicle engines contribute to air 
pollution. As a result of these findings, section 202(a) requires EPA 
to issue standards applicable to emissions of that air pollutant. New 
motor vehicles and engines emit CO2, CH4, 
N2O, and HFC. EPA has established standards and other 
provisions that control motor vehicle emissions of CO2, 
HFCs, N2O, and CH4. EPA has not set any standards 
for PFCs or SF6 as they are not emitted by motor vehicles.
---------------------------------------------------------------------------

    \2461\ 74 FR 66496 (Dec. 15, 2009).
---------------------------------------------------------------------------

3. EPA's Conclusion That the Final CO2 Standards Are 
Appropriate and Reasonable
    In this section, EPA discusses the factors, data and analysis the 
Administrator has considered in the selection of the EPA's revised 
CO2 emission standards for MYs 2021 and later and the 
comments received on EPA's consideration of these factors (see further 
discussion below on EPA's summary and analysis of comments).
    As discussed in Section VIII.A.1 above, the primary purpose of 
Title II of the Clean Air Act is the protection of public health and 
welfare, and GHG

[[Page 25106]]

emissions from light-duty vehicles have been found by EPA to endanger 
public health and welfare.\2462\ The goal of the light-duty vehicle GHG 
standards is to reduce these emissions which cause or contribute to air 
pollution which may reasonably be anticipated to endanger public health 
or welfare, while taking into account other factors as discussed above.
---------------------------------------------------------------------------

    \2462\ Id.
---------------------------------------------------------------------------

    CAA section 202(a)(2) states when setting emission standards for 
new motor vehicles, the standards ``shall take effect after such period 
as the Administrator finds necessary to permit the development and 
application of the requisite technology, giving appropriate 
consideration to the cost of compliance within such period.'' 42 U.S.C. 
7521(a)(2). That is, when establishing emission standards, the 
Administrator must consider both the lead time necessary for the 
development of technology that can be used to achieve the emission 
standards and the resulting costs of compliance on those entities that 
are directly subject to the standards. In previous rulemakings, 
including the rulemaking that established the current standards, EPA 
considered lead time-related elements, including comparative per-
vehicle cost increases by manufacturer for both cars and trucks, 
comparative penetration rates of advanced technologies by manufacturers 
for both cars and trucks, and lead time concerns about increasing 
technology penetration rates for these advanced technologies beyond 
current levels. EPA also considered comparative industry-wide costs and 
differences between alternatives, framed in terms of total costs and 
percentage differences between alternatives. These elements are 
discussed in detail throughout the analysis. As mentioned previously, 
however, the performance of the fleet in recent years indicates that 
the lead time deemed as adequate in the 2012 rulemaking was not 
sufficient.
    EPA is not limited to consideration of the factors specified in CAA 
section 202(a)(2) when establishing standards for light-duty vehicles. 
In addition to feasibility and cost of compliance, EPA may (and 
historically has) considered such factors as safety, energy use and 
security, degree of reduction of both GHG and non-GHG pollutants, 
technology cost-effectiveness, and costs and other impacts on 
consumers.
    EPA also considers relevant case law. Critical to this series of 
joint rulemakings with NHTSA, the Court in Massachusetts v. EPA,\2463\ 
recognized EPA's argument that ``it cannot regulate carbon dioxide 
emissions from motor vehicles'' without ``tighten[ing] mileage 
standards . . . .''--a task assigned to DOT. The Court found that 
``[t]he two obligations may overlap, but there is no reason to think 
the two agencies cannot both administer their obligations and yet avoid 
inconsistency.'' \2464\ Accordingly, the agencies have worked closely 
together in setting standards, and many of the factors that NHTSA 
considers to set maximum feasible standards overlap with factors that 
EPA considers under the Clean Air Act. Just as EPA considers energy use 
and security, NHTSA considers these factors when evaluating the need of 
the nation to conserve energy, as required by EPCA. Just as EPA 
considers technological feasibility, the cost of compliance, 
technological cost-effectiveness and cost and other impacts upon 
consumers, NHTSA considers these factors when weighing the 
technological feasibility and economic practicability of potential 
standards. EPA and NHTSA both consider implications of the rulemaking 
on CO2 emissions as well as criteria pollutant emissions. 
And, NHTSA's role as a safety regulator inherently leads to the 
consideration of safety implications when establishing standards. The 
balancing of competing factors by both EPA and NHTSA are consistent 
with each agency's statutory authority and recognize the overlapping 
obligations the Supreme Court pointed to in directing collaboration.
---------------------------------------------------------------------------

    \2463\ 549 U.S. 497, 531 (2007).
    \2464\ Id. at 532.
---------------------------------------------------------------------------

    As discussed in prior rulemakings setting GHG standards,\2465\ EPA 
may establish technology-forcing standards under section 202(a), but it 
must provide a rationale for concluding that the industry can develop 
the needed technology in the available time. However, EPA is not 
required to set technology-forcing standards under section 202(a). 
Rather, because section 202(a), unlike the text of section 202(a)(3) 
and section 213(a)(3),\2466\ does not specify that standards shall 
obtain ``the greatest degree of emission reduction achievable,'' EPA 
retains considerable discretion under section 202(a) in deciding how to 
weigh the various factors, consistent with the language and purpose of 
the Clean Air Act, to determine what standards are appropriate.
---------------------------------------------------------------------------

    \2465\ See, e.g., 77 FR 62624, 62673 (Oct. 15, 2012), EPA and 
NHTSA final rule for 2017 and later model year light-duty GHG 
emissions and CAFE standards.
    \2466\ Section 202(a)(3) provides that regulations applicable to 
emissions of certain specified pollutants from heavy-duty vehicles 
or engines ``shall contain standards which reflect the greatest 
degree of emission reduction achievable through the application of 
technology which the Administrator determines will be available . . 
. giving appropriate consideration to cost, energy, and safety 
factors associated with the application of such technology.'' 42 
U.S.C. 7521(a)(3). Section 213(a)(3) contains a similar provision 
for new nonroad engines and new nonroad vehicles (other than 
locomotives or engines used in locomotives). 42 U.S.C. 7547(a)(3).
---------------------------------------------------------------------------

    The proposed rule presented an analysis of alternatives, in support 
of the Administrator's consideration of a range of alternative 
CO2 standards as potential revisions of the existing 
standards for model years 2021 and later, from the previous standards 
(representing an increase in stringency of approximately 5 percent per 
year from MY 2021 through MY 2025) to several less stringent 
alternatives. These alternatives ranged from a zero percent increase in 
stringency to a stringency increase for passenger cars of 2 percent per 
year and for light trucks of 3 percent per year, in addition to the 
baseline alternative consisting of the previous standards.\2467\ The 
analysis supported the range of alternative standards based on factors 
relevant to the EPA's exercise of its section 202(a) authority, such as 
emissions reductions of GHGs and other air pollutants, the necessary 
technology and associated lead-time, the costs of compliance for 
automakers, the impact on consumers with respect to cost and vehicle 
choice, and effects on safety. The proposed rule identified the 
alternative composed of a zero percent increase in stringency as the 
preferred alternative.
---------------------------------------------------------------------------

    \2467\ 83 FR 42990, Table I-4 (August 24, 2018).
---------------------------------------------------------------------------

    EPA received numerous public comments on the range of stringency 
alternatives in the proposed rule and the Administrator's consideration 
of various factors in determining appropriate GHG standards under 
section 202(a) of the CAA. Below EPA responds to comments on these 
issues. EPA notes that many comments concerned the technical foundation 
and analysis upon which EPA was basing its regulatory decisions, such 
as the modeling of emission control technologies and costs, the safety 
analysis, and consumer issues. Comments specific to these analyses are 
discussed elsewhere in this preamble. The section below addresses 
comments specifically addressing EPA's considerations in finalizing 
appropriate CO2 emissions standards under the CAA.
    EPA's conclusion, after consideration of the factors described 
below, public comments, and other information in the administrative 
record for this action is that holding CO2 emissions 
standards for MY 2020 flat through MY 2026 is not appropriate or 
reasonable. EPA

[[Page 25107]]

concludes steady stringency increases year over year are warranted, but 
that the MY 2021-2026 standards first established in 2012 are not 
appropriate taking into account lead time and the various factors 
described below. Accordingly, the Administrator has concluded that 1.5 
percent annual increases in stringency from the MY 2020 standards 
through MY 2026 (Alternative 3 of this final rule analysis) \2468\ are 
reasonable and appropriate.
---------------------------------------------------------------------------

    \2468\ The numbered Alternatives presented in the SAFE proposed 
rule (see Table I-4 at 83 FR 42990, August 24, 2018) were in some 
cases defined differently than those presented in this final rule 
(see Section V). Unless otherwise stated, the Alternatives described 
in this section refer to those presented in this final rule.
---------------------------------------------------------------------------

a) Consideration of the Development and Application of Technology To 
Reduce CO2 Emissions
    When EPA establishes emission standards under CAA section 202, it 
considers both what technologies are currently available and what 
technologies under development may become available. For today's final 
rule, EPA considered the analysis of the potential penetration into the 
future vehicle fleet of a wide range of technologies that both reduce 
CO2 and improve fuel economy (see FRIA Chapter X). The 
majority of these technologies have already been developed, have been 
commercialized, and are in-use on vehicles today. These technologies 
include, but are not limited to, engine and transmission technologies, 
vehicle mass reduction technologies, technologies to reduce aerodynamic 
drag, and a range of electrification technologies. The electrification 
technologies include 12-volt stop-start systems, 48-volt mild hybrids, 
strong hybrid systems, plug-in hybrid electric vehicles, and dedicated 
electric vehicles.
    This consideration is especially important given current 
projections about relatively lower fuel prices than what was projected 
in 2012. In that rulemaking, EPA expressed concern that some 
alternatives may require too much advanced technologies (including 
electrification) in light of uncertain consumer acceptance of added 
costs, as well as the technologies themselves.\2469\ There, EPA 
concluded that more stringent increases in technology penetration rates 
raise serious concerns about the ability and likelihood that 
manufacturers can smoothly implement additional technologies to meet 
requirements.\2470\
---------------------------------------------------------------------------

    \2469\ 77 FR 62879.
    \2470\ See 77 FR at 62875, discussion about certain alternatives 
may require too much electrification and ``may well be overly 
aggressive in the face of uncertain consumer acceptance of both the 
added costs and the technologies themselves. EPA continues to 
believe these technology penetration rates are inappropriate given 
the concerns just voiced.'' At 62877, ``This increase in tech 
penetration rates raises serious concerns about the ability and 
likelihood manufacturers can smoothly implement. . . .''
---------------------------------------------------------------------------

    As shown in Section VII of this preamble and in FRIA Section VII, 
the projected penetration of technologies varies across the 
Alternatives considered for this final rule. In general, the baseline 
alternative consisting of the previous EPA standards as finalized in 
2012 was projected to result in the highest penetration of advanced 
technologies into the vehicle fleet, in particular mild hybrids at 7.1 
percent penetration and strong hybrids at 9 percent penetration by MY 
2030. By contrast, the revised final standards adopted today (1.5 
percent per year stringency improvement from MY 2021 through MY 2026) 
are projected to result in a significantly lower level of mild and 
strong hybrids used to meet the standards, at 1.6 percent mild hybrids 
and 2.2 percent strong hybrids by MY 2030. Further, the final rule 
analysis indicates that the previous CO2 standards would 
have led to a projected 5.7 percent penetration of dedicated electric 
vehicles (EV), with 0.4 percent penetration of plug-in hybrid electric 
vehicles (PHEV); the revised final standards reduce this projected 
level to 3.7 percent EV penetration (with 0.2 percent PHEV 
penetration), which again is more in line with what the EPA believes is 
a more appropriate projected level of market penetration.
    The technology penetration rates in the analysis for the final rule 
are changed since EPA's prior analysis. These changes in the estimated 
penetrations in this rulemaking are due to changes in the model that 
are meant to reflect consumer response to the standards, as well as 
changes to estimates for technology costs and effectiveness. In the 
2017 Final Determination on Model Year 2022-2025 standards, where EPA 
found there was available and effective technology to meet the MY 2022-
2025 standards, the technology was available at reasonable cost to the 
vehicle manufacturers and consumers, there was adequate lead time, and 
the standards were feasible and practicable. EPA also found that the 
previous MY 2022-2025 standards could be met largely through advanced 
gasoline vehicle technologies, with low levels of electrified 
vehicles.\2471\ The levels of electrified vehicle technologies 
projected in this final rule to meet the baseline Alternative (the 
previous GHG standards) differ slightly from those projected in the 
2017 Final Determination. In this final rule, EPA projects a combined 
strong and mild hybrid penetration of 16 percent (compared to 20 
percent in the 2017 Final Determination), with the share of mild 
hybrids somewhat lower (7 percent compared to 18 percent in the 2017 
Final Determination) and the share of strong hybrids higher (9 percent 
compared to 2 percent in in the 2017 Final Determination). EPA projects 
a total level of plug-in vehicles of 6 percent, similar to the 5 
percent total projected in the 2017 Final Determination, but with a 
slightly different mix of plug-in hybrid electric vehicles (0.4 percent 
compared to 2 percent in the 2017 Final Determination) and dedicated 
electric vehicles (5.7 percent compared to 3 percent in the 2017 Final 
Determination).
---------------------------------------------------------------------------

    \2471\ ``Final Determination on the Appropriateness of the Model 
Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards 
under the Midterm Evaluation,'' EPA-420-R-17-001, January 2017. See 
Table ES-1, page 4-5, and Section II (i), (ii), and (iii), pages 28-
24. Hereafter ``2017 Final Determination.''
---------------------------------------------------------------------------

    Another aspect of the analysis that EPA considered related to 
technology development and application is manufacturers' projected 
level of over-compliance under the alternatives considered for the 
final rule. Under the least stringent Alternatives (Alternative 1, zero 
percent stringency improvement, and Alternative 2, 0.5 percent per year 
stringency improvement), manufacturers overall are projected to over-
comply with those levels of stringency. For example, under Alternative 
1, manufacturers are projected to achieve a CO2 level of 206 
g/mi in MY 2029, 16 g/mi below (more stringent than) the required 
target level of 222 g/mi. Similarly, for Alternative 2, manufacturers 
are projected to achieve a CO2 level of 205 g/mi in MY 2029, 
10 g/mi below the required target level of 215 g/mi. Thus, the industry 
is projected to considerably over-comply with the Alternative 1 and 2 
standards. Under the final standards, the projected level of over-
compliance is much narrower, only 4 g/mi (198 g/mi by MY 2029 compared 
to a 202 g/mi target), and for other alternatives that are more 
stringent than the final standards, that gap is similar or even more 
narrow as shown in Table VII-7. This is an indication that the 
standards in Alternatives 1 and 2 may not represent

[[Page 25108]]

an appropriate level of stringency when compared to the pace at which 
manufacturers would be applying technologies. While some level of over-
compliance is expected so that manufacturers retain a reasonable 
compliance margin, Alternatives 1 and 2 would, based on the final rule 
analysis, result in manufacturers retaining a compliance margin more 
than 2-3 times that of the other alternatives. The Administrator has 
rejected those lower stringency Alternatives in part for this reason 
and believes that the final standards (Alternative 3, 1.5 percent per 
year stringency improvement) represent an appropriate margin of 
compliance that can be attained given the projected pace of 
manufacturers' application of technologies.
    EPA received several comments regarding its consideration of the 
development and application of GHG reducing technologies. The 
California Air Resources Board (CARB) commented that, despite what they 
characterize as evidence of widely available technology, EPA has 
proposed to promulgate emission standards that are less stringent than 
existing standards and that would lead to increased emissions of GHGs. 
The New York State Department of Environmental Conservation commented 
that the proposal did not ``appropriately value, or consider, 
technology advancement and innovation by OEMs and automotive parts 
suppliers'' and noted the role of technology innovation in reducing 
technology costs. EPA notes that the agencies specifically considered 
technology cost-savings attributable to experience with technology--in 
other words, the analysis provides that technology costs reduce over 
time.
    The Center for Biological Diversity (CBD) et al. commented that 
since technologies exist today that can achieve the current standards, 
reducing the standards to the level proposed in the NPRM is contrary to 
the objectives of the Clean Air Act. These parties further commented 
that EPA failed to make a proposed finding that additional lead-time is 
necessary, as they argue is required by Section 202(a)(2). The Green 
Energy Institute at Lewis and Clark Law School and others similarly 
commented that EPA lacks a reasonable justification for extending the 
phase-in period for the current standards because compliant 
technologies currently exist and are already commercially available.
    The Attorney General of California and others commented that EPA 
acknowledges that most or all technology necessary to meet the current 
standards is available, and does not provide evidence to support how 
additional lead time is ``necessary to permit the development and 
application of the requisite technology.''
    In response to the public comments, and as EPA indicated in the 
proposal and in the 2012 Final Rule establishing the previous 
standards, the technologies projected to be used to meet the GHG 
standards, including the alternatives in the proposal as well as the 
final standards, are currently available and in production. If the 
appropriateness of the standards were based solely on an assessment of 
technology availability, and lead time considerations were limited to 
the development of such technology, EPA might consider more stringent 
CO2 standards to be potentially appropriate. But this is not 
the sole or predominant factor to be weighed. In 2012, EPA had to 
balance this issue as well. As in 2012, manufacturers today are capable 
of building vehicles that can meet the standards that any of the 
regulatory alternatives evaluated in the final rule would require. 
However, greater uncertainty about consumer acceptance of those 
technologies (as compared to what EPA believed was likely in 2012) 
means that providing more lead time is appropriate.\2472\
---------------------------------------------------------------------------

    \2472\ See 77 FR at 62871 (``As stated above, EPA's analysis 
indicates that there is a technology pathway for all manufacturers 
to build vehicles that would meet their final standards as well as 
the alternative standards. The differences between the final 
standards and these analyzed alternatives lie in the per-vehicle 
costs and the associated technology penetration rates.'').
---------------------------------------------------------------------------

    As in 2012, EPA disagrees with commenters that a finding that 
necessary technology is available is, by itself, determinative of the 
appropriate emission standard under CAA section 202(a). As described in 
the proposed rule and in this section of the final rule, the 
Administrator weighs technology availability and lead time along with 
several other factors, including costs, emissions impacts, safety, and 
consumer impacts in determining the appropriate standards under section 
202(a) of the CAA.
    Under this analysis, given the factors discussed later in this 
Section, the previous standards would yield technology penetration 
rates for advanced technologies beyond what is appropriate and 
reasonable. By contrast, the final standards are projected to result in 
more modest penetration rates for advanced technologies that 
nonetheless will achieve an increased level of technology penetration 
compared to the standards applicable for MY 2020. For example, the 
final rule analysis projects that dynamic cylinder deactivation 
penetration for MY 2030 would be 39.2 percent under the previous 
standards for, but 34.4 percent under today's final standards. 
Similarly, turbocharged engine penetration would be a projected 48 
percent by MY 2030 under the previous standards, compared to 36.4 
percent under the final standards. In addition, mild hybrids are 
projected to change from 7.1 percent to 1.6 percent, strong hybrids 
from 9 percent to 2.2 percent, and dedicated electric vehicles from 5.7 
percent to 3.7 percent (all for MY 2030) under the final standards 
instead of the previous standards. The Administrator believes that the 
level of technology development and application for the final standards 
is an appropriate balance, in light of the relevant factors considered 
as a whole, as discussed below.
(b) Consideration of the Cost of Compliance
    EPA is required to consider costs of compliance when setting 
standards under section 202(a). The standards finalized today would 
reduce required technology costs for the industry by an estimated $108 
billion for the vehicles produced from MY 2017 through MY 2029 (at 3 
percent discount rate, see Section VII) compared to the EPA standards 
established in 2012. While less-stringent increases would result in 
additional technology cost savings ($129 billion and $126 billion for 
Alternatives 1 and 2, respectively), technology cost savings are only 
one element that EPA considers.
    In addition to capital cost savings, the final standards would 
reduce the per-vehicle costs by $1,250 per vehicle in MY 2030, compared 
to the standards set in 2012, as shown in Table VII-77. While less-
stringent increases would result in greater per-vehicle technology 
cost-savings, cost-savings alone do not dictate the appropriate 
standards. For example, Alternatives 1 and 2 would save manufacturers 
$1,218 and $1,181 in per-vehicle costs in MY 2030 compared to the 
previously issued standards. Alternatives more stringent than the final 
standards would be more burdensome to manufacturers, with Alternatives 
4 through 8 ranging from a cost savings to manufacturers of $927 to 
$351 per-vehicle compared to the previous standards.
    The costs to comply projected in this final rule are higher than 
those previously projected by EPA in the 2017 Final Determination: In 
2017 EPA projected that the per-vehicle cost to meet the MY 2025 
standards would be $875 on average, with a range of $800 to $1,115 
considering a range of

[[Page 25109]]

sensitivities (in 2015 dollars).\2473\ The costs to the auto industry 
for complying with the previous MY 2022-2025 standards projected in the 
2017 Final Determination were $24 billion to $33 billion (in 2015$ at 7 
percent and 3 percent discount rates, respectively).\2474\ Again, EPA 
notes that the values in this final rule analysis and the values in the 
2017 Final Determination have different points of reference making them 
not directly comparable, as discussed above.
---------------------------------------------------------------------------

    \2473\ See 2017 Final Determination Table ES-1, page 4-5, and 
II(v), page 24-26.
    \2474\ Id. at Table ES-4, page 7.
---------------------------------------------------------------------------

    Several public comments addressed EPA's consideration of costs of 
compliance in setting the revised standards. The Alliance of Automobile 
Manufacturers (Alliance) commented that the proposal's cost estimates 
for the current MY 2021 and later standards differed from what EPA 
projected in 2012 when setting those standards. The Alliance argued 
that that those changes in the expected costs of the previously issued 
standards provide significant reasoned support for EPA's view that the 
existing standards should be reduced.
    The Association of Global Automakers (Global Automakers) commented 
on the importance of lead time for technology investment. While it 
agreed that the existing standards are too stringent, it stated that 
vehicle manufacturers and suppliers have invested $76 billion in 
manufacturing facilities, and that much of that was for improvement in 
CO2 emission reductions and fuel economy improvements. At 
least some of that investment, according to Global Automakers, was made 
to meet the standards set in 2012. Global Automakers expressed concern 
with an abrupt halt to gradual fuel economy improvements, as such an 
approach could result in stranded capital investments for automakers 
and suppliers.
    CBD and others disagreed with EPA's conclusion that the cost of 
broader adoption of technologies is unreasonable in light of other 
factors considered by EPA. CBD and others claimed that the Clean Air 
Act narrowly allows for consideration of cost only as a question of 
whether costs of compliance make it infeasible for manufacturers to 
meet standards within the relevant period. They argue that this 
consideration relates to lead time, and not to a broader consideration 
of costs. They assert that broader compliance cost considerations apply 
only to the motor vehicle industry. They also claim that compliance 
costs to meet the standards set in 2012 for the 2017-2025 model years 
are not challenging to the industry.
    These commenters also state that the costs to industry to meet the 
standards are not high enough to require reducing standards, to permit 
development and application of the required technology. They claim that 
the only burden that Congress intended to impose as a constraint on 
emission reduction requirements are costs that are ``so severe as to 
preclude the deployment of required technology during the relevant 
period.''
    The New York State Department of Environmental Conservation 
commented on the role of technology innovation in considering 
technology feasibility, while acknowledging that the feasibility 
analysis allows for consideration of numerous factors argues that since 
technology exists today to meet the standards for MY 2026, no lead time 
is necessary. It further states that EPA did not appropriately balance 
or consider in the proposal future technological advancements and OEM 
innovation that will further constrain the costs of new technology.
    In response to the Alliance's comment that the projected compliance 
costs have changed significantly from EPA's 2012 rule, EPA agrees. 
Indeed, this is a significant factor in EPA's conclusion that the 
previous standards were too stringent. EPA notes that the projected 
difference between the cost to comply with the previous standards and 
the costs to comply with the standards established today is lower in 
this final rule analysis as compared to the projected difference 
between the proposal's preferred alternative and the previous 
standards. EPA concludes that the final standards nevertheless result 
in significant reductions in required technology costs for auto 
manufacturers compared to the previous standards.
    EPA also considered the Global Automakers' concern that freezing 
the standards from MY 2021-2026 as proposed could result in stranded 
capital for the auto industry and automotive suppliers who have 
invested significantly in meeting the previous standards. The standards 
EPA is finalizing today, unlike the proposed preferred alternative, 
will require the gradual increase in CO2 improvements across 
the fleet, at a rate of 1.5 percent per year stringency improvement, 
thus supporting investments in GHG-reducing technologies, at a pace 
that EPA believes is more reasonable than that of the previous 
standards.
    EPA disagrees with CBD et al.'s comments that the agency's 
consideration of costs is inappropriate or not supported by the record. 
EPA disagrees that Congress intended section 202(a)(2)'s requirement to 
give ``appropriate consideration to the cost of compliance within such 
period'' to mean that the agency ``only consider compliance costs if 
they are so severe as to preclude deployment of the requisite 
technology during the period.'' EPA does not interpret the Clean Air 
Act as limiting EPA's consideration of costs to manufacturers only to 
the question of whether such costs are so high that a manufacturer 
could not afford to deploy the technology in question for a given model 
year--that would be tantamount to suggesting that EPA must always set a 
standard to achieve ``the greatest degree of emission reduction 
achievable through the application of technology,'' which as discussed 
above is not EPA's approach to setting standards such as these under 
section 202(a). And this is particularly important when setting 
CO2 standards, which, as described above, have a significant 
impact on vehicle utility and performance that differs from other 
standards established under Section 202. As discussed above, Congress 
specified such technology-forcing standards elsewhere in section 202 
and could have done so here (or otherwise specified that standards 
shall take effect ``as soon as practicable'' while taking into 
consideration costs and other factors)--but did not do so. Section 
202(a) prevents EPA from implementing standards sooner than feasible, 
taking into account lead time considerations and the cost of 
compliance, but does not require standards be implemented as soon as 
feasible or at the limit of feasibility, taking into account the cost 
of compliance. EPA notes that it received numerous comments on the 
analysis underlying the proposed rule, and the analysis for this final 
rule in fact was changed from the proposal in consideration of these 
comments, as discussed in Section VI.B. Nevertheless, the projected 
costs to comply with the previous MY 2021-2026 standards remain 
significant as discussed above, and EPA has considered these costs 
along with other factors under the CAA in determining the final 
standards, as discussed in Section VIII.A.3.h) below.
(c) Consideration of Costs to Consumers
    In this section EPA considers the cost impacts on consumers. First, 
the initial up-front costs to consumers are discussed, then the costs 
associated with fuel expenditures, and finally the total ownership 
costs to consumers over the life of the vehicles.

[[Page 25110]]

    In addition to the $1,250 per-vehicle technology costs to the 
automotive industry described above, which EPA expects could, and 
likely would, be passed on to consumers, the analysis estimates other 
per-vehicle costs that could be borne by consumers, specifically costs 
attributed to changes in financing, insurance, taxes, and other fees, 
as shown in Section VII. Considering these additional costs, EPA's 
final standards (Alternative 3) would result in reduced costs to 
consumers of $1,385 in MY 2029 (at a 3 percent discount rate) compared 
to EPA's previously issued standards. While alternatives lower in 
stringency than the final standards would save consumers more (i.e., 
Alternatives 1 and 2 would save consumers $1,665 and $1,637, 
respectively, in MY 2029 at 3 percent discount rate), while 
alternatives more stringent than the final standards would save 
consumers less (i.e., Alternatives 4 through 7 would save consumers a 
range of from $1,329 to $620, for MY 2029 at 3 percent discount rate), 
this is only one of the factors EPA considers in setting standards. On 
balance, EPA believes that further increases in stringency, compared to 
the proposal, are appropriate and reasonable.
    Compared to the previously issued CO2 standards, the 
standards finalized today will result in increased fuel consumption and 
associated expenditures for consumers. The analysis detailed in the 
Final RIA and summarized in Section VII of this preamble projects the 
increased fuel consumption for owners of the vehicle over the projected 
life of the vehicle, up to 39 years, as compared to the previously 
issued standards as the baseline. For example, as shown in Table VII-84 
(at a 3 percent discount rate), consumers will spend $1,461 more in 
fuel costs over the vehicle lifetime, which the analysis assumes can be 
up to 39 years,\2475\ under today's final standards (Alternative 3) 
compared to the previously issued standards.
---------------------------------------------------------------------------

    \2475\ For further information of on the modeled distribution of 
registrations by age see, e.g., Table VI-238--Registrations, Total 
VMT, and Proportions of Total VMT by Vehicle Age (in Section 
VII.D.2.b).2.(d)) which shows the distribution of registrations by 
vehicle age.
---------------------------------------------------------------------------

    EPA notes that, when comparing lifetime fuel savings for all owners 
of a vehicle to the upfront additional ownership costs--generally borne 
by the initial purchaser, a net reduction in benefits of $175 is seen 
under the final standards. That said, as noted by several commenters, 
consumers keep vehicles for a much shorter period of time prior to 
trading the vehicle in for another or selling the vehicle.\2476\ CFA, 
for instance mentioned that consumers retain vehicles for more than 
five years, and a group of State Comptrollers and Treasurers referred 
to an IHS Markit report that the average length of time a consumer 
keeps a new car is approximately 6.6 years. Accordingly, such a 
simplistic comparative approach would anticipate that a consumer 
account for fuel savings over a much longer period of time than would 
be rational. Further, it is important to note that consumers are 
informed of estimated average annual fuel costs for the vehicle, as 
well as a comparison of the difference between five years'-worth of 
fuel costs or savings compared to an average new vehicle on the 
Monroney label that must be posted on every new vehicle offered for 
sale.
---------------------------------------------------------------------------

    \2476\ It should be noted, however, that, all else being equal, 
improved fuel economy can improve resale value of a vehicle. That 
said, it is not at all clear that consumers generally anticipate 
potential future incremental trade-in value attributable to improved 
fuel economy when making a decision as to which new vehicle to 
purchase.
---------------------------------------------------------------------------

    In the 2017 Final Determination, EPA projected that the previous MY 
2022-2025 standards compared to the MY 2021 standards would provide 
fuel savings of $52 billion to $92 billion and total net benefits of 
$59 billion to $98 billion (in 2015 dollars and at 7 percent and 3 
percent discount rates, respectively, and based on AEO2016 reference 
case fuel prices). The up-front vehicle costs to consumers were 
projected to be approximately $926 per vehicle, including the vehicle 
technology costs, taxes and insurance.\2477\ EPA projected that 
consumers would realize net savings of $1,650 over the lifetime of a 
new MY 2025 vehicle (net of increased lifetime costs and lifetime fuel 
savings).\2478\ Under the final standards, vehicle sales are expected 
to increase by 2.2 million vehicles over MY 2017-2029 compared to 
projected sales under the previous standards. EPA views this projection 
of vehicle sales increases resulting from the final standards as 
important in facilitating the turnover of the fleet to newer, safer 
vehicles, all of which will be subject to increasingly stringent 
criteria pollutant emission requirements as federal Tier 3 emission 
standards continue to phase in from MY 2017 through MY 2025.
    Below the major comments are summarized regarding EPA's 
consideration of the impact of the revised standards on consumers. 
Securing America's Future Energy (SAFE) commented that vehicle prices 
are influenced by many factors beyond the GHG standards, and that costs 
to improve fuel economy make up only a portion of the vehicle price. 
SAFE notes that fuel savings from efficient vehicles offsets increase 
ownership costs. SAFE further claims, without support, that standards 
``do not have a major role in creating higher vehicle prices, or in 
suppressing sales.'' Accordingly, SAFE argues that pausing fuel economy 
increases, as proposed in the NPRM, is not justified. SAFE suggests 
that fuel savings impacts should be discussed along with technology 
cost increases.
    CBD and others commented that EPA's consideration of consumer 
costs, including finance and insurance costs, cannot outweigh its 
public health mandate. Such commenters noted that some of the options 
analyzed in the notice showed that fuel savings of the lifetime of the 
vehicle outweighed upfront vehicle price increases, and that not 
choosing such an alternative is not justified. CBD then goes on to 
argue that the analysis inflates technology costs and undercounts fuel 
savings.
    The California Attorney General and others claim that EPA's 
consideration of the potential increased costs for consumers related to 
maintenance, financing, insurance, taxes, and other fees is 
unjustified, unlawful, and contrary to its prior position that 
compliance cost considerations include only costs to the motor-vehicle 
industry.
    EPA notes that fuel efficiency and GHG standards affect labor and 
materials costs, technology add-ons, and sales mix, and expects the 
estimated cost decrease from these final standards to have a positive 
effect on the auto market and vehicle buyers. As described in the 
notice and throughout this preamble, EPA disagrees that standards have 
no major impact on increasing prices or suppressing sales. Fuel-saving 
technology adds costs, and as prices increase, fewer consumers can 
afford to buy new cars--either because they cannot afford a new car, or 
because they decide to purchase an older vehicle, or because they 
decide to keep their existing vehicle. EPA also notes that both the 
notice and this preamble discusses fuel savings from the various 
alternatives analyzed. Some commenters suggest EPA calculate and 
consider fuel savings, spread over the lifetime of the vehicle up to 39 
years and experienced by multiple owners--compared to the upfront 
vehicle costs, which are generally paid for by the original purchaser 
either in cash or through additional finance costs over a much shorter 
period of time. This approach, which would yield a projected $175 in 
additional costs (additional lifetime outlays for fuel minus avoided 
upfront vehicle costs)

[[Page 25111]]

over the multi-owner, lifetime of a vehicle beyond the initial 
ownership savings, distorts the comparison. Instead, EPA concludes that 
the upfront vehicle technology costs (and associated financing costs) 
are a more important factor. In other words, a consumer is more likely 
to buy a new vehicle at a lower up-front price even if that vehicle 
will incur a more-than offsetting level of fuel costs over its lifetime 
that will be borne by the first and all subsequent owners of the 
vehicle.\2479\ By reducing upfront costs, more consumers will be able 
to afford new vehicles, which will result in a quicker fleet turnover 
to safer, more efficient vehicles that emit lower amounts of criteria 
pollutants than the existing fleet. In fact, the agencies project that 
the revised standards will result in 2.2 million additional new 
vehicles sold--all of which would meet the latest safety standards and 
be subject to the phase-in of the Tier 3 criteria pollutant emission 
standards.
---------------------------------------------------------------------------

    \2479\ For further discussion regarding consumers valuation of 
fuel economy, see preamble section VI.D.1.b).(2) (sales), preamble 
section VI.D.1.b).(8), and Final Regulatory Impact Analysis section 
III.C.
---------------------------------------------------------------------------

    With respect to the comments that consideration of costs to 
consumers is contrary to CAA section 202(a)(2), EPA disagrees. As 
discussed above, section 202(a)(2) requires EPA to consider the cost of 
compliance, which EPA has done, and it allows EPA to consider other 
costs, including costs to consumers, which EPA also have done, in this 
rule and past rules setting standards under section 202(a). The statute 
sets some minimum requirements for EPA's consideration, but permits a 
wider range of concerns to be considered, including public health and 
welfare but also safety, costs to consumers, and other factors 
discussed herein. As discussed above, and below, EPA has considered the 
effects of a range of potential standards across this entire set of 
factors. The agency is permitted to take all of these factors into 
account, and that is what it has done in selecting the final standards.
d) Consideration of GHG Emissions and Other Air Pollutant Emissions
    As discussed above, the purpose of GHG standards established under 
CAA section 202 is to reduce GHG emissions, which EPA has found to 
endanger public health and welfare, in an appropriate manner that takes 
into account other factors as directed by Congress and in the 
reasonable exercise of EPA's discretion under the statute. Today's 
final standards are projected to increase CO2 emissions 
compared to the previously issued standards, by a total of 867 million 
metric tons (MMT) over the lifetime of MY 1977 through MY 2029 vehicles 
(see Section VII of this preamble)--i.e., by 2.9% of the amount 
projected to be attributable to passeners cars and light trucks under 
the baseline/augural standards. Of this CO2 emissions 
increase, 731 MMT would come from tailpipe emissions, and an additional 
136 MMT from upstream sources, both being nearly 3% greater than 
projected to occur under the baseline/augural standards. The analysis 
projects that Alternatives more stringent than the final standards 
would result in smaller increases in CO2 emissions. Also 
compared to the baseline/augural standards, and also over the lifetime 
of MY 1977-2029 vehicles, Alternatives 4 through 7 are projected to 
increase CO2 emissions by 826 MMT (2.8%) to 361 MMT (1.2%). 
Alternatives less stringent than the final standards would increase 
CO2 emissions by a greater amount, 1,074 MMT (3.5%) and 
1,044 MMT (3.6%), for Alternatives 1 and 2 respectively.\2480\
---------------------------------------------------------------------------

    \2480\ This preamble and the FRIA document estimate annual GHG 
emissions from light-duty vehicles under the baseline CO2 
standards, the final standards, and the standards defined by each of 
the other regulatory alternatives considered. For the final rule 
issued in 2012, EPA estimated changes in atmospheric CO2, 
global temperature, and sea level rise using GCAM and MAGICC with 
outputs from its OMEGA model. Because the agencies are now using the 
same model and inputs, outputs from NHTSA's EIS (that used more 
recent versions of GCAM and MAGICC) were analyzed. Today's analysis 
estimates that annual GHG emissions from light-duty vehicles under 
the CO2 standards and corresponding CAFE standards, which 
are very similar. Especially considering the uncertainties involved 
in estimating future climate impacts, the very similar estimates of 
future GHG emissions under CO2 standards and 
corresponding CAFE standards means that climate impacts presented in 
NHTSA's EIS represent well the climate impacts of the CO2 
standards.
---------------------------------------------------------------------------

    In addition to GHG emissions, EPA has considered the change in 
criteria air pollutant emissions impacts due to the revised 
CO2 standards. EPA has considered both tailpipe emissions 
and upstream emissions associated with increased fuel consumption. 
Unlike with CO2 emissions, which EPA found to be a long-
lived greenhouse gas well-mixed throughout the global atmosphere, 
criteria pollutant emissions contribute primarily to local and regional 
air pollution. Generally, tailpipe emissions for volatile organic 
compounds (VOC), nitrogen oxides (NOX), and particulate 
matter (PM) decrease under the final standards compared to the previous 
standards, leading to improvements in human health in areas where air 
quality improves. Upstream emissions attributable to refining and 
transportation of the additional fuel needed under less stringent 
standards increase under the final standards, leading to adverse 
impacts on public health in locations where air quality worsens. The 
additional upstream emissions generally exceed the reduced tailpipe 
emissions, leading to net increases in these pollutants and net 
increases in adverse health effects. Under the model year analysis 
(changes in pollutants summed over the lifetimes of MY 1977-2029 
vehicles for calendar year 2017 and later), and relative to total 
emissions projected to be attributable to passenger cars and light 
trucks under the baseline/augural standards, these increases range from 
0.1% (for NOX) to 0.7% (for SO2 and PM). On the 
other hand, projected net emissions of carbon monoxide (CO) are 0.4% 
lower under the final standards than under the baseline/augural 
standards, and emissions of air toxics (e.g., benzene) are 0.1-0.4% 
lower under the final standards, varying among different toxic 
compounds.
    In addition to evaluating emissions impacts under the model year 
analysis described above, EPA has considered the emissions impacts 
under a calendar year analysis, which provides information over a 
longer time horizon about the interactions between all vehicle model 
years on the road in any given calendar year--that is, considering the 
effects of the revised MY 2021 and later standards on fleet turnover 
and utilization from calendar year 2017 out to 2050. Both the model 
year analysis and the calendar year analysis provide relevant 
information about the impacts of EPA's standards. When viewed from the 
calendar year analysis perspective that extends through 2050, the 
emissions impacts of the revised MY 2021 and later standards compared 
to the baseline/augural standards vary over time, with cumulative 
differences generally being greater in magnitude than under the model 
year analysis: EPA's analysis shows cumulative VOC emissions through 
2050 under the final standards increasing by a total of nearly 575 
thousand tons (1.9%) relative to the cumulative amount projected to 
accrue through 2050 under the baseline/augural standards. On the same 
basis, estimated NOX and PM emissions increase by about 173 
thousand tons (0.8%) and 16.5 thousand tons (1.7%), respectively. On 
the other hand, also on the same basis, estimated CO and SO2 
emissions decrease by about 278 thousand tons (0.1%) and 38 thousand 
tons (0.8%), respectively.
    As shown in the NHTSA Final Environmental Impact Statement (FEIS),

[[Page 25112]]

NHTSA's analysis indicates small air quality improvements in some areas 
and small decrements in others which could help or hinder individual 
areas' efforts to attain the NAAQS in the future.
    EPA has also considered the health effects of air pollution 
associated with today's final standards. As discussed above, it is the 
cumulative contribution of the lower projected vehicle tailpipe 
emissions with the higher projected upstream emissions (primarily from 
the production and distribution of gasoline) which impact air quality. 
As noted above and presented in detail elsewhere in this preamble and 
the Final RIA, vehicle emissions are generally reduced due to the SAFE 
final rule.
    Due largely to the projected increase in upstream emissions 
resulting from the increased production and transportation of gasoline 
resulting from the standards finalized today compared to the previous 
EPA standards, the Final Rule analysis projects increases in premature 
deaths, asthma exacerbation, respiratory symptoms, non-fatal heart 
attacks, and a wide range of other health impacts. While these health 
impacts are presented in detail elsewhere in this preamble and in the 
Final RIA, two factors suggest that the forgone premature mortality 
benefits are overstated. First, in the last year, EPA has completed 
analysis that demonstrated the likelihood that the air quality modeling 
approach used here (i.e., benefits per ton) overestimates foregone PM 
premature mortality benefits. Second, the 2012 rulemaking significantly 
overestimated gasoline price projections in its baseline, predicting 
lower fuel consumption, thus overestimating the premature mortality 
benefits in that rule. While gasoline price projections in this 
rulemaking have been updated to reflect recent data, the potential for 
this kind of unanticipated fluctuation in gasoline prices remains, thus 
estimates of fuel consumption and the correlated foregone premature 
mortality benefits may not capture actual market outcomes.
    The valuation of premature mortality effects rely on the results of 
``benefits per ton'' approach (BPT). This approach is a reduced form 
approach, which is less complex than full-scale air quality modeling, 
requiring less agency resources and time. Based on EPA's work to 
examine reduced form approach, the BPT may yield estimates of 
PM2.5-benefits for the mobile sector that are as much as 10 
percent greater than those estimated when using full air quality 
modeling.
    The EPA is currently working on a systematic comparison of results 
from its BPT technique and other reduced-form techniques with results 
from full-form photochemical modelling. While this analysis employed 
photochemical modeling simulations, we acknowledge that the Agency has 
elsewhere applied reduced-form techniques. The summary report from the 
``Reduced Form Tool Evaluation Project'', which has not yet been peer 
reviewed, is available on EPA's website at https://www.epa.gov/benmap/reduced-form-evaluation-project-report. Under the scenarios examined in 
that report, EPA's BPT approach in the 2012 rule (which was based off a 
2005 inventory) may yield estimates of PM2.5-benefits for 
the mobile sector that are as much as 10 percent greater than those 
estimated when using full air quality modeling. The estimate increases 
to 30 percent greater for the electricity sector. The EPA continues to 
work to develop refined reduced-form approaches for estimating 
PM2.5 benefits.
    Also, in this regulation, a key projection that influences the 
estimation about car purchase and driving behavior is the gasoline 
price projection. From 2008 through 2018, the average monthly gasoline 
price ranged from less $1/gallon to $4/gallon.\2481\ The gasoline price 
level and the volatility of price changes are major drivers of car 
purchasing behavior thereby gasoline consumption and the resulting 
criteria pollutant emissions. If gasoline prices are lower than 
projected in an analysis, consumers are more likely to purchase less 
fuel efficient cars, resulting in more emissions and vice versa.
---------------------------------------------------------------------------

    \2481\ https://www.eia.gov/energyexplained/gasoline/price-fluctuations.php.
---------------------------------------------------------------------------

    With a lower fuel price projection and an expectation that new 
vehicle buyers respond to fuel prices, the 2012 rule would have shown 
much smaller fuel savings attributable to the more stringent standards. 
Projected fuel prices are considerably lower today than in 2012. The 
agencies now understand new vehicle buyers to be at least somewhat 
responsive to fuel prices, and the agencies have therefore updated 
corresponding model inputs to produce an analysis the agencies consider 
to be more realistic.
    The first of these assumptions, fuel prices, was simply an artifact 
of the timing of the rule. Following recent periodic spikes in the 
national average gasoline price and continued volatility after the 
great recession, the fuel price forecast then produced by EIA (as part 
of AEO 2011) showed a steady march toward historically high, sustained 
gasoline prices in the United States. However, the actual series of 
fuel prices has skewed much lower. As it has turned out, the observed 
fuel price in the years between the 2012 final rule and this rule has 
frequently been lower than the ``Low Oil Price'' sensitivity case in 
the 2011 AEO, even when adjusted for inflation. The discrepancy in fuel 
prices is important to the discussion of differences between the 
current rule and the 2012 final rule, because that discrepancy leads in 
turn to differences in analytical outputs and thus to differences in 
what the agencies consider in assessing what levels of standards are 
reasonable, appropriate, and/or maximum feasible. Long-term predictions 
are challenging and the fuel price projections in the 2012 rule were 
within the range of conventional wisdom at the time. However, it does 
suggest that fuel economy and tailpipe CO2 regulations set 
almost two decades into the future are vulnerable to surprises, in some 
ways, and reinforces the value of being able to adjust course when 
critical assumptions are proven inaccurate. This value was codified in 
regulation when EPA bound itself to the mid-term evaluation process as 
part of the 2012 final rule.\2482\
---------------------------------------------------------------------------

    \2482\ See 40 CFR 86-1818-12(h).
---------------------------------------------------------------------------

    Because of these uncertainties surrounding air quality modeling of 
premature mortality effects, the projections of foregone PM premature 
mortality benefits are uncertain and may be over-stated. Fluctuations 
in gasoline prices contribute to this uncertainty, making it difficult 
to accurately project gasoline consumption and its related premature 
mortality benefits.
    The analysis projects that the air pollution emission increases 
associated with the revised standards will lead to an increase of 440 
to 1,000 premature deaths--deaths that occur before the normally 
expected life span--0.5% more than the number of such deaths projected 
to occur under the baseline/augural standards and over the lifetime of 
the MY 1977-MY 2029 vehicles. In addition, a wide range of health 
impacts are projected to increase by 0.4-0.6% under the final standards 
compared to occurrences projected to occur the standards established in 
2012, as summarized in Table VII-132 et seq.
    When quantified using the calendar year (CY) analysis perspective 
(CYs 2018-2050), under the revised final standards (compared to the 
previous standards), premature mortality is expected to increase from 
460 to 1,010 deaths (i.e., by 0.4%), upper and lower respiratory 
symptoms are expected to increase by 22,000 cases (0.4%), asthma 
exacerbations are projected to increase by 16,000 cases (0.4%), acute 
bronchitis

[[Page 25113]]

cases are projected by increase by 720 (0.4%), non-fatal heart attacks 
are projected to increase by 450 (0.4%), hospital admissions for 
cardiovascular and respiratory issues are projected to increase by 225 
(0.4%) cases, and emergency room visits for respiratory issues are 
projected to increase by 260 (0.4%). In addition, these additional 
health impacts are expected to result in an additional 61,000 work loss 
days (0.3% of the number projected under the baseline/augural 
standards) and 355,000 minor restricted activity days (0.4% more than 
under that baseline/augural standards) for the public. Compared to the 
baseline/augural standards, the agencies estimate that the final 
standards rule will increase by 0.3-0.4% each of the various health 
impacts accumulated through 2050 (e.g., premature deaths, upper and 
lower respiratory symptoms, asthma exacerbations, acute bronchitis 
cases, hospital admissions for cardiovascular and respiratory issues, 
emergency room visits for respiratory issues).
    In the 2017 Final Determination, EPA projected GHG emissions 
reductions of 540 million metric tons over the lifetimes of MY 2022-
2025 vehicles.\2483\ EPA also projected criteria pollutant emission 
reductions for CY2040 of 97,000 tons of VOC, 24,000 tons of 
NOX, 3,600 tons of PM2.5, and 15,000 tons of 
SO2.\2484\ EPA projected that these emissions reductions 
would result in positive health benefits through CY2050.\2485\ In this 
final rule, the revised final standards compared to the previous 
standards are projected to result in an increase in emissions and 
health incidences, as discussed above, resulting in $5 billion or $3 
billion (in 2018 $, and reflecting, respectively, either a 7 percent or 
3 percent discount rate) in foregone public health benefits (see Table 
VII-103 and Table VII-104).
---------------------------------------------------------------------------

    \2483\ 2017 Final Determination at Table ES-3, page 6, and 
Section II (iv), page 24.
    \2484\ 2016 Proposed Determination at Appendix C, Table C.54, 
page A-163.
    \2485\ Id. at Table C.87, page A-183.
---------------------------------------------------------------------------

    In public comments on these topics, the Attorney General of 
California and others commented that, in adopting the previous 
standards, EPA focused on obtaining significant CO2 emission 
reductions, but now proposed to increase emissions relative to the 
previous standards without sufficient justification. They claim that 
EPA offered no justification of acknowledgement of a change in 
position, stating that none of the alternatives further the goal of 
CO2 emission reductions. They argue that EPA justifies its 
proposal on the limited impact of the rule on global climate change, 
and that failing to seek incremental improvements is contrary to the 
EPA's duties under the Clean Air Act.
    The United States Conference of Catholic Bishops commented that 
considering public safety of any set of standards requires giving 
significant weight to the effect of air pollution, and that the 
proposal failed to promote public health and safety.
    The Chesapeake Bay Foundation (CBF) claims that the proposal would 
have significant health consequences that disproportionately impact 
minority and low-income communities in the Chesapeake Bay. They discuss 
general impacts of climate change CBF argues that criteria pollutant 
health impacts of the proposal, should be more heavily weighed against 
safety impacts of the rule.
    The State of Washington commented that the agencies did not analyze 
public health effects from increased criteria pollutant emissions 
arising from increased petroleum consumption or environmental justice 
concerns. They claim that the NPRM's discussion of the negligible 
impact of the rulemaking on global climate change is ``deeply 
concerning.''
    As noted above, EPA agrees that the purpose of Title II emission 
standards is to protect the public health and welfare from air 
pollution, and in establishing emission standards, the agency is 
cognizant of the importance of this goal. At the same time, EPA 
balances multiple factors in determining what standards are reasonable 
and appropriate. And, contrary to some commenters' views, unlike other 
provisions in Title II, section 202(a) does not require the 
Administrator to set standards which result in the greatest degree of 
emissions control achievable. Thus, in setting these standards, the 
Administrator has taken into consideration other factors discussed 
above and below, including not only technological feasibility, lead-
time, and the cost of compliance, but also potential impacts of vehicle 
emission standards on safety and other impacts on consumers.
    Several commenters claimed that the agencies did not analyze health 
impacts of the various alternatives, but this is not accurate. First, 
the notice and PRIA included this information in monetized terms to 
facilitate the balancing of various factors. Further, NHTSA conducted a 
comprehensive Draft Environmental Impact Statement, which discussed 
these effects in detail. For this final rule, these health impacts have 
been separately itemized, as summarized above. Other commenters claimed 
that the agencies did not sufficiently consider environmental justice 
elements in the proposal. This, too, is inaccurate, as discussed 
elsewhere in this preamble.
    In response to comments of the California Attorney General and 
others, that the Clean Air Act cannot allow for increases in a 
regulated emission, EPA notes that the 2012 Final Rule specifically 
called for a Mid Term Evaluation process that envisioned the potential 
for an adjustment of the standards in case the stringency increases 
established in 2012 were no longer reasonable and appropriate. As 
discussed above, the increases in stringency of the standards for MY 
2021-2025 are, on balance, not reasonable and appropriate based on a 
consideration of the factors described in this preamble. EPA now 
recognizes based on updated information and analysis that industry 
should be provided additional lead time to meet the later model years 
of standards set in the 2012 rule, and, as discussed in this preamble, 
industry is having unanticipated difficulties complying with earlier 
years of the standards, with fleetwide performance failing to meet 
CO2 emission targets in MY 2016 and MY 2017. That is not to 
say that CO2 and criteria pollutant emissions are not 
significant factors in this rulemaking. Indeed, they are weighed 
heavily along with other important factors considered by EPA, which has 
led to increasing stringency on a 1.5 percent annual basis for the 
2021-2026 model years. Importantly, the agencies project that the 
revised standards will result in an additional 2 million new vehicles 
sold before 2030 compared to under the baseline/augural standards. This 
means that an additional 2 million vehicles will be produced during the 
phase-in of the Tier 3 emission standards, which implement more 
stringent tailpipe standards for criteria pollutants, displacing 
greater numbers of higher-emitting older vehicles and providing 
significant health benefits. As discussed, when finalizing the Tier 3 
standards in 2014, ``[t]he final Tier 3 vehicle and fuel standards 
together will reduce dramatically emissions of NOX, VOC, 
PM2.5, and air toxics.'' \2486\
---------------------------------------------------------------------------

    \2486\ 79 FR 23425.
---------------------------------------------------------------------------

    Although GHG emissions reductions would be lessened under the 
standards finalized today compared to the previously issued EPA 
standards, in light of this assessment indicating higher vehicle costs 
and associated impacts on consumers, EPA believes that, on balance, the 
final standards

[[Page 25114]]

(Alternative 3) are justified and appropriate.
(e) Consideration of Consumer Choice
    EPA believes that consumer demand is an important consideration in 
setting CO2 emission standards, because one of EPA's goals 
in setting the standards has been and continues to be to allow 
manufacturers to provide, and consumers to purchase, vehicles with 
varying attributes and functionality rather than to shift demand to 
certain vehicle types or sizes. Societal and economic trends play a 
role in this area as well--if fuel prices are relatively high, demand 
for fuel-efficient vehicles increase and, as a result, compliance with 
standards is easier to achieve. If fuel prices are relatively low--as 
they are now and are projected to be in the mid-term--consumer demand 
for fuel-efficiency is less strong, making it harder for manufacturers 
to comply with the standard. While manufacturer difficulty in complying 
due to lack of consumer demand may not be the deciding factor in 
determining the appropriate levels of stringency for standards, it is 
relevant to understanding lead time difficulties, which EPA is required 
to consider under Section 202(a)(2).
    As discussed previously, the EPA CO2 standards are based 
on vehicle footprint, and in general smaller footprint vehicles have 
individual CO2 targets that are lower (more stringent) than 
larger footprint vehicles. The passenger car fleet has footprint curves 
that are distinct from the light-truck fleet. One of EPA's goals in 
designing the footprint-based standards, in considering the shape, 
slope, and stringency of the footprint standard curves, and in adopting 
various compliance flexibilities (e.g., emissions averaging, banking, 
and trading, air-conditioning credits, off-cycle credits) was to 
maintain consumer choice. The EPA standards are designed to require 
reductions of CO2 emissions over time from the vehicle fleet 
as a whole, but also to provide sufficient flexibility to the 
automotive manufacturers so that firms can produce vehicles that serve 
the needs of their customers. The past several model years in the 
marketplace show that, while this approach reduces the impact of 
increased fuel economy on consumer choice, it does not adequately 
account for changes in consumer preference. As a result, as discussed 
throughout this preamble, manufactures are struggling to meet 
CO2 emission standards based upon their fleet performance. 
In fact, the 2017 model year saw that only three major manufacturers 
had fleets that met the standards. One reason behind these challenges 
is that, while the footprint-based attribute standards account for 
vehicle length and width, they do not account for vehicle height or 
weight. And, since many crossovers sold today are classified as 
passenger cars and not light trucks, the additional weight of such 
vehicles to provide for requisite ride height puts pressure on 
CO2 emission compliance for automaker passenger car fleets. 
Similarly, large SUVs are subject to the same footprint-based standards 
as lighter trucks, putting pressure on CO2 emission standard 
compliance. For the 2017 model year, 12 percent of the fleet consisted 
of car-based SUVs, and 32 percent of the fleet consisted of truck-based 
SUVs.\2487\ Taller and heavier vehicles, including crossovers and SUVs, 
are more popular today than was expected at the time the standards were 
set. While automobile manufacturers have continued to offer a broad 
range of vehicles (e.g., full-size pick-up trucks with high towing 
capabilities, minivans, cross-over vehicles, SUVs, and passenger cars; 
vehicles with off-road capabilities; luxury/premium vehicles, 
supercars, performance vehicles, entry level vehicles, etc.) despite 
continuing required increases in fuel economy stringency, this has 
largely been possible because of well-stocked over-compliance credit 
banks from when standards were less stringent and the ability to 
acquire credits from other manufacturers. As mentioned earlier, the 
agencies have concerns whether this is sustainable. Automotive 
companies have been able to reduce their fleet-wide CO2 
emissions while continuing to produce and sell the many diverse 
products that serve the needs of consumers in the market. The agencies 
recognize that automotive customers are diverse, that automotive 
companies do not all compete for the same segments of the market, and 
that increasing stringency in the standards can be expected to have 
different effects not only on certain vehicle segments but also on 
certain manufacturers that have developed market strategies around 
those vehicle segments. Taking into consideration this diversity of the 
automotive customer base, and of the strategies which have developed to 
meet specific segments, EPA concludes that the previous standards are 
not reasonable or appropriate.
---------------------------------------------------------------------------

    \2487\ 2018 EPA Automotive Trends Report: Greenhouse Gas 
Emissions, Fuel Economy, and Technology since 1975, available at: 
https://www.epa.gov/automotive-trends/download-automotive-trends-report.
---------------------------------------------------------------------------

    In the initial determination, EPA assessed several factors related 
to consumer choice, including the costs to consumers of new vehicles 
and fuel savings to consumers, as described above under Section 
VII.A.2.c). In 2017, EPA found that the previous standards would 
increase the upfront costs of vehicles but overall would have positive 
net benefits because lifetime fuel savings outweighed the lifetime 
vehicle costs for consumers. As discussed above, the costs of 
technology to comply with the standards are generally borne by the 
initial purchaser, with understanding of fuel cost implication given 
statutorily required disclosures. In contrast, the fuel savings are 
realized by many subsequent owners over the vehicles' lifetime, which 
this analysis assumes can be up to 39 years. New vehicle purchasers are 
not likely to place as much weight on fuel savings that will be 
realized by subsequent owners. Accordingly, EPA is placing greater 
weight on the up-front vehicle cost savings to consumers in light of 
the goal of accelerating the turnover of the motor vehicle fleet to 
safer cars that emit fewer criteria pollutants.
    EPA received many comments regarding the agency's consideration of 
consumer choice in determining appropriate standards under section 
202(a) of the CAA. The Alliance commented that EPA's concerns regarding 
consumer choice are well founded, stating ``in the years since 2012 
(and in part due to the unexpected decrease in fuel prices), consumers 
have demonstrated less interest in high-efficiency/low-emission 
vehicles than EPA and NHTSA projected in issuing the 2012 Final Rule. 
As such, compliance with the existing standards would require a 
substantially greater variance than EPA expected from the vehicle fleet 
that consumers would otherwise choose.''
    Global Automakers agreed that consumer acceptance is an important 
factor, but does not justify holding standards flat through the 2026 
model year. Global Automakers further commented that ``[f]uel economy 
remains a factor in vehicle purchase decisions, though perhaps not a 
dominant one.''
    CBD and others commented that the Clean Air Act does not allow EPA 
to reduce stringency based upon consumer choice factors. They point to 
the diversity of the vehicle fleet and argue that EPA's consideration 
of projected tech levels and associated costs as ``speculative'' and 
not grounded in fact.
    U.S. Congressman Mark DeSaulnier claimed that the justification for 
the proposal appeared to be consumer

[[Page 25115]]

willingness to buy new vehicles. He claimed that absent any standards 
whatsoever, automakers could produce more vehicles that consumers would 
want to purchase. He stated that the standards require all vehicles to 
become more efficient and that EPA has an overly simplistic 
understanding of American consumers, who, according to him, are ``wary 
of the price tag'' when shopping, but, nonetheless, ``overwhelmingly 
want more efficient vehicles, and they want to reduce the health burden 
of air pollution.''
    The Institute for Policy Integrity (IPI) claims, without support, 
that as fuel efficiency technology is introduced and becomes 
widespread, consumer attitudes will change and will start focusing on 
such technology. IPI also claims that manufacturers can change consumer 
preference through advertising. IPI implies that manufactures play a 
larger role in shaping consumer options of their needs that consumers 
do themselves. IPI also comments that academic literature relating to 
demand- and supply-side obstacles to fuel economy indicates that the 
proposal's justification runs counter to available evidence.
    The University of California Berkeley Environmental Law Clinic 
(Berkeley) argued against EPA's consideration of consumer choice in 
setting standards, claiming that low-income households bear exposure to 
operating costs, fuel price fluctuations, and environmental impacts. 
Berkeley also claimed that EPA's purported list of features consumers 
may favor over fuel economy is not supported by evidence, and, in any 
event, should be categorized into lists of ``needs'' versus ``wants.''
    Consumer choice is a complex consideration when setting standards. 
As Congressman DeSaulnier correctly notes, EPA cannot disregard its 
consideration of public health and welfare based upon the agency-
projected whims of consumers. At the same time, the willingness of 
consumers to pay for fuel economy improvements, which as described 
above affects vehicle performance and utility in a manner 
distinguishable from criteria pollutant emissions, has a direct effect 
upon the ability of manufacturers to sell their product. And as 
consumers demand vehicles with increased ride height (which, all else 
being equal, increases CO2 emissions), establishing 
standards that account for this--but still require manufacturers to 
focus on improving emission performance, is reasonable and appropriate.
    In response to Global Automakers' comment that consumers do not 
heavily focus on fuel economy in making purchase decisions, EPA agrees, 
but notes that this is a consumer's choice, as federal law requires 
that consumers are made aware of fuel economy impacts, pursuant to 49 
U.S.C. 32908. EPA also agrees that the willingness to pay for fuel 
economy improvements is ``not zero.''
    EPA agrees with the Global Automakers comment that while consumer 
choice is an important consideration in determining the appropriate 
level of the revised standards, the final rule analysis does not 
support holding the standards constant. Although EPA proposed standards 
at the level of 0 percent increase in stringency from MY 2021 and 
later, after considering the comments received and based on the updated 
analysis for this final rule, EPA is finalizing standards with a 1.5 
percent per year improvement in stringency from MY 2021 to MY 2026. As 
indicated in the comments on this topic, there is a range of views and 
relevant information concerning the extent of consumers' interest in 
fuel economy and on the role fuel savings plays in consumer purchase 
decisions.\2488\ EPA's understanding is that some consumers value fuel 
economy more than others, and EPA finds it unnecessary to identify the 
precise role of fuel economy in consumer purchase decisions because the 
Administrator believes that the standards should encourage a range of 
vehicles meeting a range of consumer preferences. Further, as described 
above, consumers are made aware of the relative fuel price impacts of 
new vehicles, given the required information label on new vehicles, 
thus indicating that, in all likelihood, consumers do take fuel 
expenses into account when making new vehicle purchase decisions.
---------------------------------------------------------------------------

    \2488\ Studies of the role of fuel economy in consumer purchase 
decisions have found a wide range of values (Greene, D., A. Hossain, 
J. Hofmann, G. Helfand, and R. Beach. ``Consumer Willingness to Pay 
for Vehicle Attributes: What Do We Know?'' Transportation Research 
Part A 118 (2018), p. 258-79). The National Academy of Sciences in 
2015 judged that ``there is a good deal of evidence that the market 
appears to undervalue fuel economy relative to its expected present 
value, but recent work suggests that there could be many reasons 
underlying this, and that it may not be true for all consumers.'' 
National Research Council of the National Academies (2015). Cost, 
Effectiveness, and Deployment of Fuel Economy Technologies for 
Light-Duty Vehicles. Washington, DC: National Academies Press, p. 9-
16.
---------------------------------------------------------------------------

    EPA disagrees with Congressman DeSaulnier's assertion that EPA 
seeks to set standards that do not affect what manufacturers produce--
instead, the agencies examine what consumers are purchasing in the 
market to determine what standards are appropriate. The agency's 
assumptions in 2012--that consumers would gravitate toward the purchase 
of compact sedans and coupes in response to exceedingly high fuel 
prices--have proved incorrect. Fuel prices have fallen and remained 
relatively low, and are projected to remain relatively low throughout 
the period covered by this rulemaking. EPA seeks to achieve 
improvements in CO2 emissions, but it is not realistic to 
expect the high demand for crossover vehicles to abate, or for those 
vehicles to meet more-stringent standards set for compact sedans. That 
said, EPA agrees with Congressman DeSaulnier that American consumers 
are wary of the price of vehicles--popular reporting that consumers may 
reference explain affordability concerns in crisis terms--even 
indicating that the average price of a vehicle is now beyond that which 
is affordable to the median household income of every city outside of 
Washington, DC \2489\ This results in significant adverse economic 
impacts--higher finance charges, taxes, registration fees, and 
insurance costs, all of which result in challenges qualifying for 
financing and longer finance terms, which increase the likelihood of 
negative equity scenarios. EPA also agrees with Congressman DeSaulnier 
that consumers want increased fuel efficiency and to reduce the impacts 
of harmful air pollution. These are all true. But direct health impacts 
of vehicles emissions stem more from criteria pollutant emissions than 
from CO2 emissions. And CO2 emission technology 
has a significant relationship to the price of vehicles for which 
consumers are so wary. EPA, with this rulemaking, is attempting to 
strike the correct balance between a number of factors, including 
improving efficiency and affordability, which should yield additional 
sales and an improved rate of fleet turnover to vehicles that have 
better criteria pollutant emissions--particularly since the vehicles 
sold subject to this rulemaking will be sold during the phase-in of 
Tier 3 criteria pollutant emission standards.
---------------------------------------------------------------------------

    \2489\ See., e.g., Car and Driver, ``For Middle-Class Shoppers, 
New Cars Are Moving out of Reach'' November 30, 2019. Available at: 
https://www.caranddriver.com/news/a30061910/middle-class-car-shoppers-priced-out/; New York Times, ``New Cars Are Too Expensive 
for the Typical Family, Study Finds'' July 2, 2016. Available at: 
https://www.nytimes.com/2016/07/02/your-money/new-cars-are-too-expensive-for-the-typical-family-study-finds.html.
---------------------------------------------------------------------------

    In response to Berkeley, low-income consumers are even more 
sensitive to upfront vehicle purchase prices than they are to the 
smaller delta between weekly or monthly fuel costs

[[Page 25116]]

experienced over time between the previous standards and the standards 
finalized today--they may well take note of the fact that one cannot 
pay today's bills with tomorrow's savings. They may also want to take 
note that the standards finalized today are projected to improve fleet 
turnover into newer vehicles that emit reduced criteria pollutants.
    EPA disagrees with the assertion by CBD and others that the agency 
has not provided a rationale for its consideration of consumer choice 
in determining the appropriate standards. EPA notes that despite a 
variety of vehicles on the market today and over the past several 
years, the fleet has failed to comply with standards based upon 
performance beginning with the 2016 model year, and has fallen further 
behind in the 2017 model year, when only three major automakers 
complied with CO2 emission standards based upon performance 
alone.
    In response to IPI's comment that the deployment of more fuel-
efficient technologies, combined with manufacturer advertising, will 
change consumer preference, this runs counter to historical trends. 
Manufacturers have continuously deployed additional fuel efficiency 
technology in each model year--which is why EPA continues to see 
fleetwide improvements in CO2 emissions on new vehicles. And 
manufacturers have consistently advertised the fuel economy performance 
of their vehicles. Federal law requires the physical posting fuel 
economy performance, as well as estimated and comparative fuel cost 
information, on every new vehicle offered for sale. Notwithstanding 
this activity, consumer demand, and willingness to pay for technology 
that reduces CO2 emissions and improves fuel economy, has not matched 
required standards--which is one of the reasons that EPA is revising 
the standards today. As discussed in the proposal, EPA recognizes that 
the diversity in the automotive customer base, combined with the facts 
and analysis developed by the agency in this rulemaking, raises 
concerns that the previous standards, if they are not adjusted, may not 
continue to fulfill the agency's goal of providing sufficient 
manufacturer flexibility to meet consumer needs and consumer choice 
preferences in their vehicle purchasing decisions. In the 2012 Final 
Rule and the Initial Determination, EPA expected that consumers would 
readily accept fuel-saving technologies in their new vehicles, despite 
the agency's uncertainty about the role of fuel savings in consumers' 
purchase decisions. Given low fuel prices and the pronounced market 
shift to crossovers and SUVs, notwithstanding required disclosers of 
fuel costs and relative fuel economy performance, EPA now concludes 
that it is appropriate to account for the shift in consumer preference 
in concluding that the standards set in 2012 did not provide sufficient 
lead time for manufacturers to achieve the standards set at that time. 
EPA remains concerned that the projected level of hybridization and 
other advanced technologies and the associated vehicle costs necessary 
to achieve the previous standards are too high from a consumer-choice 
perspective, and not sufficiently account for consumer acceptance of 
such technology. While consumers have benefited from improvements over 
several decades in traditional vehicle technologies, such as 
advancements in transmissions and internal combustion engines, 
electrification technologies are a departure from what consumers have 
traditionally purchased. Strong hybrid and other advanced 
electrification technologies have been available for many years (20 
years for strong hybrids and eight years for plug-in and all electric 
vehicles), and sales levels have been relatively low, in the 2-3 
percent range.\2490\ As discussed above, the analysis projects that the 
2012 EPA standards would be projected to require a significant increase 
in hybridization (up to 8 percent for mild hybrids and 10 percent for 
strong hybrids in MY 2030). This large increase in technology demand 
over the next decade could lead to automotive companies needing to 
change the choice of vehicle types they are able to offer to consumers, 
compared to what the companies would otherwise have offered in the 
absence of the previously issued standards. As discussed above, 
manufacturers are, by and large, not meeting existing standards based 
upon actual fleet performance in CO2 emissions and are 
instead relying upon the use of earned or acquired credits. As the 
previous standards were set to increase significantly through MY 2020 
and thereafter, reducing the rate of increase is appropriate and 
reasonable. Doing so will provide manufacturers with sufficient lead 
time to meet the standards being set today.
---------------------------------------------------------------------------

    \2490\ For instance, the 2019 calendar year saw only a 1.4% 
penetration of battery electric vehicles in the light duty fleet, 
following 1.2% for 2018, 0.6% for 2017, 0.5% for 2016, and 0.4% for 
2015. Wards Auto Monthly Sales reports, available at https://wardsintelligence.informa.com/.
---------------------------------------------------------------------------

    EPA recognizes that one possibility for automotive companies who 
wish to retain their current vehicle offerings, but face compliance 
challenges is to purchase GHG emissions credits. In EPA's annual 
Automotive Trends Report, EPA has reported that credit trading has 
occurred frequently in the past several years to achieve compliance 
with the GHG standards.\2491\ Credit trading can lower a manufacturer's 
costs of compliance, both for those selling and those purchasing 
credits, and this program compliance flexibility is another tool 
available to auto firms to allow them to continue offering the types of 
vehicles that customers want. Between MY 2010 and MY 2017, these trades 
have included 11 firms, with five firms selling CO2 credits 
to seven firms.\2492\ The number of firms participating in the GHG 
credits market represents about one-half of the automotive companies 
selling vehicles in the U.S. market, but since several of these firms 
are small players, they represent less than half of the vehicle 
production volume. In total, approximately 48 million Megagrams of 
CO2 credits have been traded between firms, which represents 
19 percent of the MY 2017 industry-wide bank of credits. That said, 
more manufacturers have relied upon previously earned credits to 
achieve compliance. Between MY 2010 and MY 2017, 80% of firms applied 
previously earned credits. However, long-term planning is an important 
consideration for automakers, and an automaker who may need to purchase 
credits as part of a future compliance strategy is not guaranteed to 
find credits. The automotive industry is highly competitive, and firms 
may be reluctant to base their future product strategy on an uncertain 
future credit availability, but face struggles in achieving 
CO2 emission reductions in a manner that meets consumer 
expectations for cost, utility, and performance. Also, pools of 
available credits continue to decline over time as the standards become 
more stringent and previously banked credits are either used or expire; 
indeed, this has happened in recent years.\2493\ EPA's views on the 
availability of the credit market to aid in manufacturers' compliance 
have changed since the Initial Determination. Based upon the 
information available to the EPA in early January 2017, the auto 
industry had outperformed its standards in the four previous compliance 
years (MYs 2012-2015) and EPA had viewed that as

[[Page 25117]]

a positive trend.\2494\ Since then, however, overall manufacturer 
performance failed to meet the standard fleetwide, and many 
manufacturers relied on credits to meet their individual compliance 
targets. Furthermore, recent experience suggests that availability of 
the credit bank is becoming a more uncertain means to achieve 
compliance.\2495\ Thus, while credit trading may be a useful 
flexibility to reduce the overall costs of the program and to smooth 
the pathway to compliance realizing necessary transitions from vehicle 
redesign cycles, EPA believes it is important to set standards that 
preserve consumer choice without relying on credit purchasing 
availability as a compliance mechanism. As discussed in Section VII, 
the agencies project that the EPA final standards (Alternative 3, 1.5 
percent year over year stringency improvement), will require more 
realistic penetration of advanced CO2 emission technologies 
such as electrification--better ensuring that manufacturers will be 
able to provide vehicles that meet consumer demand.
---------------------------------------------------------------------------

    \2491\ 2018 EPA Automotive Trends Report at Figures 5.15 and 
5.17.
    \2492\ EPA Greenhouse Gas Emission Standards for Light-Duty 
Vehicles: Manufacturer Performance Report for the 2016 Model Year. 
EPA-420-R-18-002. January 2019.
    \2493\ 2018 EPA Automotive Trends Report at Figure 5.17 and 
Table 5.17.
    \2494\ See Initial Determination at page 7-8.
    \2495\ Id. at Figure ES-8.
---------------------------------------------------------------------------

(f) Consideration of Safety
    As discussed above, EPA has long considered the safety implications 
of its emission standards.\2496\ More recently, EPA has considered the 
potential impacts of emission standards on safety in past rulemakings 
on GHG standards, including the 2010 rule which established the 2012-
2016 light-duty vehicle GHG standards, and the 2012 rule which 
previously established 2017-2025 light-duty vehicle GHG standards. 
Indeed, section 202(a)(4)(A) specifically prohibits the use of an 
emission control device, system or element of design that will cause or 
contribute to an unreasonable risk to safety.\2497\ The relationship 
between CO2 emissions and safety is more nuanced. Safety 
impacts relate to changes in the use of vehicles in the fleet, relative 
mass changes, and the turnover of fleet to newer and safer vehicles.
---------------------------------------------------------------------------

    \2496\ See, e.g., 45 FR 14496, 14503 (1980) (``EPA would not 
require a particulate control technology that was known to involve 
serious safety problems.'').
    \2497\ 42 U.S.C. 7521(a)(4)(A).
---------------------------------------------------------------------------

    The analysis for the final rule projects that there will be a 
change in vehicle miles traveled (VMT) under the final standards, 
specifically 607 billion less miles traveled compared to the previous 
standards case. Based on these projections about reduced VMT in the 
light-duty fleet, the analysis estimates that fatalities will be 
reduced by 2584 (out of a total impact of 3269) over the lifetime of MY 
1977-2029 vehicles compared to the previous CO2 
standards.\2498\ In other words, the reduction in fatalities under the 
final standards compared to the previous standards is primarily driven 
by the modeling's projected changes in VMT and associated changes in 
mobility (i.e., people driving less). The details of the safety 
assessment are discussed in Section VI of this preamble and in Section 
VI of the FRIA. Under alternatives with stringency levels lower than 
the final standards, the analysis projects greater reductions in VMT, 
and thus projects somewhat greater reductions in fatalities based on 
these VMT changes. Under alternatives with stringency levels higher 
than the final standards, the analysis projects lower reductions in 
VMT, and thus projects fewer fatalities reduced, See Table VI-271.
---------------------------------------------------------------------------

    \2498\ The number of fatalities projected is a product of two 
contributing factors: the number of miles driven (VMT) and the risk 
of driving (i.e., fatalities per mile). Overall in this final rule 
analysis, the change in fatalities projected is primarily caused by 
the changes in VMT.
---------------------------------------------------------------------------

    EPA notes that the magnitude of the changes in fatalities stemming 
from changes in mobility projected in this final rule is less than what 
was presented in the proposed rule. In response to comments, the 
agencies took a conservative approach to modeling the effects of 
standard stringency upon safety. The agencies held VMT constant across 
alternatives. The reasons for the differences in fatality estimates in 
the final rule compared to the proposed rule, including changes to the 
modeling inputs and projections based on the agencies' assessment of 
public comments.
    The approach for reporting fatality impacts for this final rule is 
different than the previous analyses for the Initial Determination and 
the 2012 rulemaking. First, the analysis quantifies the number of 
fatalities caused by changes in VMT between each Alternative and the 
previous standards, whereas previous analyses did not. Second, the 
safety analysis itself is different from previous analyses that assumed 
that automakers would not reduce the weight of approximately the 
lightest half of passenger cars--discounting the safety impacts of mass 
reduction. Third, while the agencies qualitatively discussed the effect 
of price increases attributable to increased stringency on vehicle 
sales, fleet turnover, and the improved safety of newer vehicles, the 
agencies never attempted to quantify these impacts.
    With respect to public comments, the Alliance commented that ``EPA 
has discretion to consider all the relevant factors in setting 
appropriate emissions standards under Sec.  202(a)(1), including 
vehicle safety. Moreover, given NHTSA's greater expertise in evaluating 
motor vehicle safety, it is appropriate for EPA to respect the views of 
its companion agency on those issues.'' The Alliance commented that 
``[t]he new safety analysis likewise provides support for EPA's 
conclusion that the MY 2021-2025 GHG standards are not appropriate and 
should be reduced in stringency. Indeed, given that the `primary 
purpose' of Sec.  202(a)(1) is `the protection of public health and 
welfare,' EPA would be abdicating its statutory duty if it ignored 
these concerns.''
    Global Automakers commented that safety impacts due to the rebound 
effect should not be attributed to the standards and should not serve 
as a basis for keeping the standards flat. They further argued that the 
dynamic scrappage model is flawed and should be removed from the 
modeling for purposes of the final rule. They also argued, that 
Congress expressed interest in improving efficiency, emissions, and 
safety (without no recognition of cost as a factor), and that 
therefore, improvement in all such areas should provide that 
improvements in efficiency would not lead to negative safety impacts.
    CBD and others commented that safety concerns should not be 
considered because the record does not indicate that vehicles must be 
unsafe to meet the previous standards. They further commented that EPA 
cannot justify reduced stringency upon ``rebound'' fatalities, and they 
argue that those fatalities cannot be considered by EPA, since they 
``stem from voluntary choices by individuals to drive more--not the 
`operation or function'of the technologies at issue'' (quoting CAA 
Section 202(a)(4)(A)).
    Environmental Defense Fund (EDF) similarly commented that the 
estimates of fatalities are unsound, as is considering total fatalities 
resulting from increased stringency, rather than fatality rates. They 
added that the projected fatalities stem from consumer and manufacture 
behaviors that are removed from the stringency requirements. They 
further argue that considering fatalities that are attributable to the 
standards--particularly rebound fatalities--are inappropriate. EDF, 
UCS, and Consumers Union argue that fatalities attributable to 
increased driving are not relevant to agency decisions.
    In response to the Alliance comments, EPA has considered safety, as 
described in this section, and agrees that the

[[Page 25118]]

potential impacts of emission standards on safety is an important 
consideration in determining appropriate standards under CAA section 
202(a). In response to comments from Global Automakers that the safety 
analysis in the proposed rule did not support freezing the standards, 
EPA agrees that safety considerations alone do not justify such an 
approach, and notes that the safety analysis performed for this final 
rule has changed from the analysis for the proposed rule based on 
consideration of public comments. EPA is finalizing standards that are 
more stringent (1.5 percent per year stringency improvement for MY 
2021-2026) than the proposed rule's preferred alternative (0 percent 
stringency improvement).
    Several commenters argued that the proposal's claims of reduced 
fatalities were based upon projected changes in driving, arguing that 
that EPA should not decide the level of the standards based on these 
assumed changes in travel. As discussed above, EPA acknowledged that 
the reduction in fatalities under the final standards compared to the 
previous standards are in large part driven by projected changes in 
driving behavior (i.e., people driving less). While EPA is not seeking 
to restrict mobility or driving, ignoring impacts associated with this 
rule would be inappropriate. Moreover, the provisions of Section 
202(a)(4) do not preclude EPA from considering such impacts. While EPA 
has considered the safety assessment for this final rule, as discussed 
in the following section below, safety was one of several factors 
considered in deciding on the level of today's final standards.
g) Consideration of Energy Security Impacts
    Among other factors EPA considered in selecting the previous 
standards in the 2012 Final Rule was the effect of the standards on 
U.S. petroleum imports and energy security.\2499\ As discussed in the 
PRIA, Final RIA and in Section Energy Security, the energy security 
position of the United States has changed dramatically since 2012. The 
U.S. has become a net exporter of petroleum and additional payments by 
United States consumers resulting from upward pressure on oil price due 
to additional demand are a transfer that occurs within the United 
States economy.\2500\ Additional petroleum use necessarily increases 
demand and thus subjects the nation to additional risk of price shocks, 
but this risk is significantly reduced as the United States has 
dramatically increased domestic petroleum production and has additional 
capacity to do so. Accordingly, energy security concerns are reduced 
compared to the assessment in the 2012 rulemaking and do not alter 
EPA's selection of final revised standards in this rule.
---------------------------------------------------------------------------

    \2499\ See 77 FR 62938, et seq.
    \2500\ The U.S. Energy Information Administration EIA estimates 
that the United States exported more total crude oil and petroleum 
products in September and October 2019, and expects the United 
States to continue to be a net exporter. See Short Term Energy 
Outlook November 2019, available at https://www.eia.gov/outlooks/steo/archives/nov19.pdf.
---------------------------------------------------------------------------

(h) Balancing of Factors and EPA's Revised Standards for MY 2021 and 
Later
    As discussed in this section, the Administrator is required to 
consider a number of factors when establishing emission standards under 
section 202(a)(2) of the Clean Air Act: The standards ``shall take 
effect after such period as the Administrator finds necessary to permit 
the development and application of the requisite technology, giving 
appropriate consideration to the cost of compliance within such 
period.'' \2501\ For this Final Rule, the Administrator has considered 
a wide range of potential emission standards (Baseline/No Action 
Alternative and Alternatives 1 through 7), ranging from the previous 
EPA standards (Baseline/No Action Alternative), through a number of 
less stringent alternatives, including the proposed preferred 
alternative (Alternative 1, 0 percent per year stringency improvement) 
and what has been chosen as the final standards (Alternative 3, 1.5 
percent per year stringency improvement). The Administrator has 
determined that the revised final standards, which would increase the 
stringency of the MY 2020 standards by 1.5 percent per year for both 
passenger cars and light-trucks from MY 2021 through 2026, are 
appropriate under section 202(a) of the CAA. In addition to 
technological feasibility, lead-time, and the costs of compliance, the 
Administrator has also considered the impact of the standards on GHG 
and non-GHG emissions reductions, the costs to consumers, and vehicle 
safety.
---------------------------------------------------------------------------

    \2501\ 42 U.S.C. 7521(a)(2).
---------------------------------------------------------------------------

    In addition to comments on each of the factors the Administrator 
considered discussed above, comments also were received on how the 
Administrator should balance these factors in determining the 
appropriate final standards.
    The Alliance commented that the CAA provides EPA with significant 
latitude to exercise its expert judgment in determining the level at 
which emissions standards should be set. The Alliance commented further 
that unlike other CAA provisions, Sec.  202(a)(1) does not require EPA 
to set standards that will result in the greatest degree of emission 
reduction achievable. Instead, the statute leaves EPA flexibility to 
decide what factors are relevant, and how to weigh those factors, in 
its decision-making process. The Alliance also commented ``EPA also has 
'significant latitude' regarding the 'coordination of its regulations 
with those of other agencies,' '' ``EPA has discretion to defer to the 
judgment of other agencies regarding issues within their areas of 
expertise,'' and the CAA ``gives the agency authority to engage in 
reasoned decision-making, balancing all of the relevant factors in 
light of the available facts. EPA has done that here and has provided a 
reasoned explanation of its determination that the environmental 
benefits of the existing MY 2021-2025 GHG standards are outweighed by 
their negative effects on costs and safety.''
    The American Iron and Steel Institute commented that it favors the 
general direction taken in the SAFE proposal, including the preferred 
option for CO2 standards, and that it believes a final SAFE 
rule that ``balances the priorities of costs to consumers, safety 
design considerations, employment impacts and total GHG emissions will 
result in the best outcome.''
    CBD and others claimed that the justifications EPA offered in the 
notice are untethered from the statute, and that EPA used a flawed 
analysis. Further, they claim that EPA did not exercise its own 
judgment and delegated its responsibilities impermissibly to NHTSA, 
failing to consider ``relevant EPA information.''
    EPA's analysis is described in detail in this preamble. EPA decided 
to use the CAFE model for a number of reasons, described in more detail 
in Section IV, including that using two models results in an 
inefficient use of resources, the CAFE model can analyze both EPA's and 
NHTSA's statutory programs, the CAFE model is capable of modeling 
incremental improvements of discrete technologies, and EPA believes 
that the CAFE model provides reasonable results. Merely because EPA has 
a set of its own analytical tools that model similar effects does not 
mean that it must use those tools to perform the analysis, and doing so 
would create unnecessary complication and lead to potential 
inconsistencies. Since the agencies are establishing standards jointly 
and seeking to avoid

[[Page 25119]]

inconsistencies in a manner consistent with Supreme Court direction, 
using the same model for the analysis is reasonable. Nonetheless, EPA 
has exercised its own judgment in this final rule.
    The California Attorney General and others claim that EPA failed 
adequately to acknowledge, explain, or justify its departure from the 
prior determination. They claim that EPA failed to propose or make a 
finding required by Section 202(a)(2) relating to adequate lead time, 
inconsistent with EPA's prior explanation that it is provided with 
limited flexibility in making such a determination.
    The California Attorney General and others also claim that EPA's 
analysis improperly weighs the factors it considers, and that it 
insufficiently weighed certain factors required under the Clean Air 
Act, including air pollution. In response, EPA notes that the Clean Air 
Act does not specify how the Administrator should weigh the factors 
considered, as discussed elsewhere in this section.
    The California Attorney General and others further noted that the 
purpose of the Clean Air Act is to is to ``protect and enhance the 
quality of the Nation's air resources so as to promote the public 
health and welfare and the productive capacity of its population.''
    The Institute for Policy Integrity claimed that the agencies 
balanced the factors in a way that conflicts with their controlling 
statutes and weighed the statutory factors without regard for the 
accuracy of the accompanying cost-benefit analysis.
    The National Coalition for Advanced Transportation claimed that the 
proposal appeared to be based on heightened concerns with cost, 
consumer acceptance, and safety, and insufficiently on technology 
availability and emissions reductions. As discussed in this section, 
EPA is neither relying solely on cost or safety nor ignoring any 
factors, but rather is balancing a number of factors.
    Green Energy Institute at Lewis and Clark Law School et al. 
commented that the Clean Air Act does not authorize the weakening or 
freezing of existing standards due to industry costs or consumer 
preferences. While EPA has broad discretion to revise standards based 
upon a balancing of factors, the final rule will provide for increasing 
stringency of 1.5 percent per year from MY 2021 through MY 2026.
    Motor & Equipment Manufacturers Association (MEMA) commented that 
the technology costs from their preferred alternative (Alternative 8 in 
the notice) were not significant and did not justify holding MY 2020 
standards flat in light of other elements, such as preserving 
investments in fuel saving technology. EPA disagrees, and considers the 
reductions in costs resulting from the revised final standards, $1,250 
per vehicle by MY 2029, to be one important aspect of the justification 
of these standards.
    EPA believes the previously issued standards for MY 2021 and later, 
considered as a whole, are too stringent. Factors in favor of reduced 
stringency include manufacturer compliance costs, and the related per-
vehicle cost savings. As described above, the agencies project that the 
final CO2 standards will reduce manufacturers' MY 2018-2029 
compliance costs by $108 billion (when applying a 3% discount rate),and 
will reduce average MY 2030 vehicle prices $977 (also applying a 3% 
discount rate). Including other costs, such as financing and insurance, 
consumers the standards finalized today will result in reduced costs of 
$1,286 per-vehicle for a MY 2030 vehicle. EPA expects that the final 
standards will not impede consumers from being able to purchase a new 
vehicle of their choice or require significant changes in product lines 
for any manufacturer. In fact, under the final standards, vehicle sales 
are expected to increase by 2.2 million vehicles over MY 2017-2029 
compared to projected sales under the augural standards, a significant 
increase in vehicles sold over this timeframe see Table VI-155. EPA 
views this projection of vehicle sales increases resulting from the 
final standards as important in facilitating the turnover of the fleet 
to newer, safer vehicles, all of which will be subject to increasingly 
stringent criteria pollutant emission requirements as federal Tier 3 
emission standards continue to phase in from MY 2017 through MY 2025.
    Another factor weighing toward reduced stringency is safety. As 
discussed previously, reduced stringency results in less pressure on 
manufacturers to reduce mass in vehicles, which, for smaller passenger 
cars has negative safety implications when involved in accidents with 
heavier vehicles. Further, as vehicle prices decrease compared to the 
previous standards, more consumers will be able to afford newer 
vehicles, which are significantly safer. Lastly, as vehicles will not 
be required to be as fuel efficient as under the previous standards, 
``rebound'' driving will be reduced. The agencies project a reduction 
in 605 billion miles traveled by light-duty vehicles produced through 
MY 2029, and project that this reduced VMT will lead to 2,584 fewer 
highway fatalities under the final standards compared to the previous 
CO2 standards (i.e., people are projected to drive less 
under the final standards with an associated reduction in driving-
related fatalities). While, notwithstanding EPA's involvement with 
State and local Transportation Control Measures (TCMs), the 
Administrator does not seek to change the way people drive--EPA's 
intention is not to restrict mobility, or to discourage driving, based 
on the level of the standards--EPA nonetheless believes it is 
appropriate to consider this projection.\2502\ The agencies also 
project that accelerated fleet turnover attributable to the change in 
standards will lead to the avoidance of a further 447 fatalities, and 
that the reduced need for reductions of vehicle mass will lead to the 
avoidance of a further 238 fatalities. In other words, the agencies 
project that the change in CO2 standards will lead to 3,269 fewer 
fatalities over the useful lives of vehicles produced through MY 2029.
---------------------------------------------------------------------------

    \2502\ Information regarding TCMs is available at https://www.epa.gov/statelocalenergy/transportation-control-measures.
---------------------------------------------------------------------------

    Factors that weigh in favor of increased stringency options are 
increased upstream criteria pollutant emissions attributable to 
additional refining and other fuel-related activities, as well as 
increased CO2 emissions and consumer fuel expenditures.
    As described above, the agencies project that the revised final 
standards will have a negative impact on air quality health outcomes, 
including a projected increase of 444 to 1,000 premature deaths from 
increased air pollution over the lifetime of the MY 1977-2029 vehicles 
on the road after calendar year 2017 cumulative through CY 2068, under 
EPA's CO2 program.\2503\ EPA recognizes that the final 
standards are projected to increase CO2 emissions compared 
to the previous EPA standards. However, EPA notes that, unlike other 
provisions in Title II referenced above, section 202(a) does not 
require EPA to set standards for light-duty vehicles which result in 
the ``greatest degree of emission reduction achievable.'' EPA has not 
chosen the standard that has the highest estimated net social benefits. 
However, as discussed elsewhere in this preamble, from a cost benefit 
perspective, the differences among the various alternatives are 
relatively narrow. EPA believes consideration of costs and benefits is 
certainly relevant to its

[[Page 25120]]

exercise of discretion in selecting appropriate standards, but also 
recognizes that some costs and benefits are difficult to quantify, and 
additional factors can prove material under the Clean Air Act as well 
in those policy decisions. For example, EPA notes that the agency 
decided against pursuing more stringent alternatives analyzed in both 
the rulemaking establishing 2012-2016 standards and the rulemaking 
establishing 2017-2025 standards.
---------------------------------------------------------------------------

    \2503\ The agencies believe that these premature mortality 
estimates may be over-estimated. Please see more detailed 
discussions in Sections VI.D.3.d) and VIII.A.3.d) in this preamble, 
and similar discussions in the final regulatory impact analysis.
---------------------------------------------------------------------------

    EPA has also given weight to the policy goal of establishing 
CO2 standards which are coordinated with NHTSA's CAFE 
standards. While not a statutory requirement, EPA has considered the 
importance of having coordinated and harmonized EPA CO2 and 
CAFE programs, while recognizing the different statutory authorities 
for those programs, since the establishment of the EPA CO2 
program. The agencies discussed the importance of having one national 
program in the SAFE Vehicles Part 1 joint action.\2504\ In today's 
joint final rule, DOT is establishing CAFE standards for MY 2021-2026 
which increase in stringency at a level of 1.5 percent per year. The 
revised EPA standards will also increase in stringency at a rate of 1.5 
percent per year. Coordinating revisions to the GHG and CAFE standards 
in order to maintain one national program is a factor the Administrator 
has consideration in determining the revised GHG standards.
---------------------------------------------------------------------------

    \2504\ 84 FR 51,310 (Sept. 27, 2019).
---------------------------------------------------------------------------

    In light of available statutory discretion and the range of factors 
that the statute authorizes and permits the Administrator to consider, 
and his consideration of the factors discussed above, the EPA concludes 
that reducing the stringency of the MY 2021-2026 standards is an 
appropriate approach under section 202(a). Therefore, based on the data 
and analysis detailed in this final rule, the Administrator concludes 
that the previous MY 2021 and later CO2 standards are too 
stringent, and is establishing revised standards for MY 2021 through MY 
2026 at a level of 1.5 percent per year improvement in stringency.
    In response to comments concerned about EPA's proposal to freeze 
the MY 2021-2026 standards at MY 2020 levels, EPA notes that it is 
finalizing the 1.5 percent per year improvement in stringency level and 
not the 0 percent improvement level proposed, after considering the 
somewhat higher costs to industry and up-front vehicle costs to the 
consumer and slightly lower GHG emissions and health-related impacts 
compared to the proposed preferred alternative. The Administrator has 
taken these tradeoffs into account in his balancing of factors under 
section 202(a) of the CAA.
    While the set of factors considered by EPA under section 202(a) of 
the CAA in today's final rule and under the midterm evaluation 
regulations \2505\ in the Initial Determination are similar and 
overlapping, the Administrator recognizes that he is balancing these 
factors differently in this final rule than in the Initial 
Determination. In the Initial Determination, EPA's decision that the 
previous MY 2022-2025 standards were appropriate was based on 
conclusions that the standards were feasible within the lead time 
provided at reasonable costs, the standards would result in significant 
reductions in GHG emissions and oil consumption and associated fuel 
savings for consumers, and the standards would yield significant 
benefits to public health and welfare and positive net benefits 
overall, without adverse impacts on industry, safety, or 
consumers.\2506\
---------------------------------------------------------------------------

    \2505\ 40 CFR 86.1818-12(h).
    \2506\ Initial Determination, Section III, page 29-30.
---------------------------------------------------------------------------

    Since the Initial Determination, EPA has completed its compliance 
review of the first two model years covered by the 2012 final rule. 
Notwithstanding widespread availability of vehicles that meet or exceed 
their CO2 emission targets, consumers are not expressing 
sufficient interest in fuel economy in their purchasing decisions to 
enable manufacturers to meet the standards based upon fleet 
performance. Although manufacturers earned significant credits in the 
early years of the agency's CO2 regulation history, these 
credits are being applied broadly across the industry and well in 
advance of the more aggressive model year stringency increases. While 
some manufacturers, including alternative fuel automakers are earning 
significant tradable credits, they do not have to trade them. And 
building a program around the potential for acquiring credits from 
competing manufacturers is not the intention of this action. While EPA 
is analyzing the differences between these standards and the previous 
standards for this rulemaking, EPA cannot ignore that this rulemaking 
was foreseen in the 2012 rulemaking. The prospect of revising the 
standards was expressly envisioned in that rulemaking based upon the 
uncertainty in the assumptions and future projections at that time. 
When viewed from the perspective of the larger set of MY 2017 through 
MY 2026 standards rulemakings, the standards finalized today fit the 
pattern of gradual, tough, but feasible stringency increases that take 
into account real world performance, shifts in fuel prices, and changes 
in consumer behavior toward crossovers and SUVs and away from more 
efficient sedans. This approach ensures that manufacturers are provided 
with sufficient lead time to achieve standards, considering the cost of 
compliance.
    In this final rule, the EPA is placing greater weight on the costs 
to industry and the up-front vehicle costs to consumers. EPA believes 
that the costs to both industry and automotive consumers would have 
been too high under the previous standards, and that the standards 
should be revised to be less stringent to lower these costs. EPA 
believes that by lowering the auto industry's costs to comply with the 
program, with a commensurate reduction in per-vehicle costs to 
consumers, the final rule is enhancing the ability of the fleet to turn 
over to newer, cleaner and safer vehicles.
    EPA believes that the characteristics and impacts of these and 
other alternative standards generally reflect a continuum in terms of 
technical feasibility, cost, lead time, consumer impacts, emissions 
reductions, and oil savings, and other factors evaluated under section 
202(a). In determining the appropriate standard to adopt in this 
context, EPA judges that the final standards are appropriate and 
preferable to more stringent alternatives based largely on 
consideration of cost--both to manufacturers and to consumers--and the 
potential for overly aggressive penetration rates for advanced 
technologies relative to the penetration rates seen in the final 
standards, especially in the face of an unknown degree of consumer 
acceptance of both the increased costs and of the technologies 
themselves--particularly given current projections of fuel prices 
during that timeframe. At the same time, the final rule helps to 
address these issues by maintaining incentives to promote broader 
deployment of advanced technologies, and so provides a means of 
encouraging their further penetration while leaving manufacturers 
alternative technology choices. EPA thus judges that more stringent 
alternatives, which would necessitate even more technology and more 
cost, would not be appropriate. Instead, EPA is adopting a more gradual 
increase in stringency to ensure that the benefits of reduced GHG 
emissions are achieved without the potential for disruption to 
automakers or consumers.

[[Page 25121]]

B. NHTSA's Statutory Obligations and Why the Selected Standards Are 
Maximum Feasible as Determined by the Secretary

    In this section, NHTSA discusses the factors, data and analysis 
that the agency has considered in the selection of the CAFE standards 
for MYs 2021 and later and the comments received on NHTSA's 
consideration of these factors (see further discussion below on NHTSA's 
summary and analysis of comments).
    As discussed in more detail below, the primary purpose of EPCA, as 
amended by EISA, and codified at 49 U.S.C. chapter 329, is energy 
conservation, and fuel economy standards help to conserve energy by 
requiring automakers to make new vehicles travel a certain distance on 
a gallon of fuel.\2507\ The goal of the CAFE standards is to conserve 
energy, while taking into account the statutory factors set forth at 49 
U.S.C. 32902(f), as discussed below.
---------------------------------------------------------------------------

    \2507\ While individual vehicles need not meet any particular 
mpg level, as discussed extensively elsewhere in this preamble, it 
is broadly true that fuel economy standards require vehicle 
manufacturers' fleets to meet certain fuel economy levels as set 
forth by NHTSA in regulation.
---------------------------------------------------------------------------

    49 U.S.C. 32902(f) states when setting maximum feasible CAFE 
standards for new vehicles, the Secretary of Transportation \2508\ 
``shall consider technological feasibility, economic practicability, 
the effect of other motor vehicle standards of the Government on fuel 
economy, and the need of the United States to conserve energy.'' In 
previous rulemakings, including the 2012 final rule that established 
CAFE standards for MY 2021 and set forth augural standards for MYs 
2022-2025, NHTSA considered technological feasibility, including the 
availability of various fuel-economy-improving technologies to be 
applied to new vehicles in the timeframe of the standards depending on 
the ultimate stringency levels, and also considered economic 
practicability, including the differences between a range of regulatory 
alternatives in terms of effects on per-vehicle costs, industry-wide 
costs, the ability of both the industry and individual manufacturers to 
comply with standards at various levels, as well as effects on vehicle 
sales, industry employment, and consumer demand. NHTSA also considered 
how compliance with other motor vehicle standards of the Government 
might affect manufacturers' ability to meet CAFE standards represented 
by a range of regulatory alternatives, and how the need of the U.S. to 
conserve energy could be more or less met under a range of regulatory 
alternatives, in terms of considerations like costs to consumers, the 
national balance of payments, environmental implications like climate 
and smog effects, and foreign policy effects like the likelihood that 
U.S. military and other expenditures could change as a result of more 
or less oil consumed by the U.S. vehicle fleet. These elements are 
discussed in detail throughout this analysis. As will be discussed in 
greater detail below, while NHTSA is considering all of the same 
factors in setting today's CAFE standards that it considered in 
previous rulemakings, and in many instances in a similar way as it 
considered those factors in previous rulemakings, the facts on the 
ground have changed and NHTSA is therefore choosing to set CAFE 
standards at a different level from what the 2012 final rule set forth.
---------------------------------------------------------------------------

    \2508\ By delegation, NHTSA.
---------------------------------------------------------------------------

    NHTSA is not limited to consideration of the factors specified in 
49 U.S.C. 32902(f) when establishing CAFE standards for passenger cars 
and light trucks. In addition to the factors enumerated above, NHTSA 
may (and historically has) considered such factors as safety and the 
environment.
    NHTSA also considers relevant case law. Critical to this series of 
joint rulemakings with EPA, the Court in Massachusetts v. EPA,\2509\ 
recognized EPA's argument that ``it cannot regulate carbon dioxide 
emissions from motor vehicles'' without ``tighten[ing] mileage 
standards . . . .''--a task assigned to DOT. The Court found that 
``[t]he two obligations may overlap, but there is no reason to think 
the two agencies cannot both administer their obligations and yet avoid 
inconsistency.'' \2510\ Accordingly, the agencies have worked closely 
together in setting standards, and many of the factors that NHTSA 
considers to set maximum feasible standards overlap with factors that 
EPA considers under the Clean Air Act. Just as EPA considers energy use 
and security, NHTSA considers these factors when evaluating the need of 
the nation to conserve energy, as required by EPCA. Just as EPA 
considers technological feasibility, the cost of compliance, 
technological cost-effectiveness and cost and other impacts upon 
consumers, NHTSA considers these factors when weighing the 
technological feasibility and economic practicability of potential 
standards. EPA and NHTSA both consider implications of the rulemaking 
on CO2 emissions as well as criteria pollutant emissions. 
And, NHTSA's role as a safety regulator inherently leads to the 
consideration of safety implications when establishing standards. The 
balancing of competing factors by both EPA and NHTSA are consistent 
with each agency's statutory authority and recognize the overlapping 
obligations the Supreme Court pointed to in directing collaboration. 
NHTSA also considers the Ninth Circuit's decision in Center for 
Biological Diversity v. NHTSA \2511\ which remanded NHTSA's 2006 final 
rule establishing standards for MYs 2008-2011 light trucks and 
underscored that ``the overarching purpose of EPCA is energy 
conservation.''
---------------------------------------------------------------------------

    \2509\ 549 U.S. 497, 531 (2007).
    \2510\ Id. at 532.
    \2511\ 538 F.3d 1172 (9th Cir. 2008).
---------------------------------------------------------------------------

    The proposed rule presented an analysis of a wide range 
alternatives as potential revisions of the existing standards for model 
year 2021 and new standards for model years 2022-2026. These 
alternatives ranged from a zero percent increase in stringency to a 
stringency increase for passenger cars of 2 percent per year and for 
light trucks of 3 percent per year, in addition to the baseline 
alternative consisting of the augural standards.\2512\ The analysis 
supported the range of alternative standards based on factors relevant 
to NHTSA's exercise of its 49 U.S.C. 32902(f) authority, such as fuel 
saved and emissions reduced, the technologies available to meet the 
standards, the costs of compliance for automakers and their abilities 
to comply by applying technologies, the impact on consumers with 
respect to cost and vehicle choice, and effects on safety. The proposed 
rule identified the alternative composed of a zero percent increase in 
stringency as the preferred alternative.
---------------------------------------------------------------------------

    \2512\ 83 FR 42990, Table I-4 (Aug. 24, 2018).
---------------------------------------------------------------------------

    NHTSA received numerous public comments on the range of stringency 
alternatives in the proposed rule and NHTSA's consideration of various 
factors in determining maximum feasible CAFE standards under 49 U.S.C. 
chapter 329. Below NHTSA responds to comments on these issues. NHTSA 
notes that many comments concerned the technical foundation and 
analysis upon which NHTSA was basing its regulatory decisions, such as 
the modeling of fuel economy-improving technologies and costs, the 
safety analysis, and consumer issues. Comments specific to these 
analyses are discussed elsewhere in this preamble. The section below 
addresses comments specifically addressing NHTSA's considerations in 
finalizing maximum

[[Page 25122]]

feasible CAFE standards under 49 U.S.C. chapter 329.
    NHTSA's conclusion, after consideration of the factors described 
below, public comments, and other information in the administrative 
record for this action is that 1.5 percent annual increases in 
stringency from the MY 2020 standards through MY 2026 (Alternative 3 of 
this final rule analysis) \2513\ are maximum feasible. Holding CAFE 
standards for MY 2020 flat through MY 2026, as proposed, would unduly 
weigh economic practicability concerns more heavily than the need of 
the United States to conserve energy, while finalizing the MY 2021 and 
augural standards first established and set forth in 2012 would place 
undue weight on the need of the U.S. to conserve energy while being 
beyond economically practicable, as described in more detail below.
---------------------------------------------------------------------------

    \2513\ The numbered Alternatives presented in the SAFE proposed 
rule (see Table I-4 at 83 FR 42990, August 24, 2018) were in some 
cases defined differently than those presented in this final rule 
(see Section V). Unless otherwise stated, the Alternatives described 
in this section refer to those presented in this final rule.
---------------------------------------------------------------------------

    The following sections discuss in more detail the statutory 
requirements and considerations involved in NHTSA's determination of 
maximum feasible CAFE standards, comments received on those issues, and 
NHTSA's explanation of its balancing of factors for this final rule.
1. EPCA, as Amended by EISA
    EPCA, as amended by EISA, contains a number of provisions regarding 
how to set CAFE standards. DOT (by delegation, NHTSA) \2514\ must 
establish separate CAFE standards for passenger cars and light trucks 
\2515\ for each model year,\2516\ and each standard must be the maximum 
feasible that the Secretary (again, by delegation, NHTSA) believes the 
manufacturers can achieve in that model year.\2517\ In determining the 
maximum feasible level achievable by the manufacturers, EPCA requires 
that NHTSA consider four statutory factors of technological 
feasibility, economic practicability, the effect of other motor vehicle 
standards of the Government on fuel economy, and the need of the United 
States to conserve energy.\2518\ In addition, NHTSA has the authority 
to consider (and traditionally does) other relevant factors, such as 
the effect of the CAFE standards on motor vehicle safety and consumer 
preferences.\2519\ The ultimate determination of what standards can be 
considered maximum feasible involves a weighing and balancing of 
factors, and the balance may shift depending on the information before 
NHTSA about the expected circumstances in the model years covered by 
the rulemaking. The agency's decision must also support the overarching 
purpose of EPCA, energy conservation, while balancing these 
factors.\2520\
---------------------------------------------------------------------------

    \2514\ EPCA and EISA direct the Secretary of Transportation to 
develop, implement, and enforce fuel economy standards (see 49 
U.S.C. 32901 et. seq.), which authority the Secretary has delegated 
to NHTSA at 49 CFR 1.95(a).
    \2515\ 49 U.S.C. 32902(b)(1) (2007).
    \2516\ 49 U.S.C. 32902(a) (2007).
    \2517\ Id.
    \2518\ 49 U.S.C. 32902(f) (2007).
    \2519\ Both of these additional considerations also can be 
considered part of economic practicability, but NHTSA also has the 
authority to consider them independently of that statutory factor.
    \2520\ Center for Biological Diversity v. NHTSA, 538 F. 3d 1172, 
1197 (9th Cir. 2008) (``Whatever method it uses, NHTSA cannot set 
fuel economy standards that are contrary to Congress's purpose in 
enacting the EPCA--energy conservation.'').
---------------------------------------------------------------------------

    Besides the requirement that the standards be maximum feasible for 
the fleet in question and the model year in question, EPCA/EISA also 
contain several other requirements, as explained below.
(a) Lead Time
    EPCA requires that NHTSA prescribe new CAFE standards at least 18 
months before the beginning of each model year.\2521\ Thus, if the 
first year for which NHTSA is proposing to set new standards in this 
NPRM is MY 2022, NHTSA interprets this provision as requiring the 
agency to issue a final rule covering MY 2022 standards no later than 
April 1, 2020.
---------------------------------------------------------------------------

    \2521\ 49 U.S.C. 32902(a) (2007).
---------------------------------------------------------------------------

    For amendments to existing standards, EPCA requires that if the 
amendments make an average fuel economy standard more stringent, at 
least 18 months of lead time must be provided.\2522\ EPCA contains no 
lead time requirement to amend standards if the amendments make an 
average fuel economy standard less stringent. NHTSA therefore 
interprets EPCA as allowing amendments to reduce a standard's 
stringency up until the beginning of the model year in question. In the 
NPRM, NHTSA proposed to amend the standards for model year 2021. NHTSA 
explained that since the agency was proposing to reduce these 
standards, the action was not subject to a lead time requirement.
---------------------------------------------------------------------------

    \2522\ 49 U.S.C. 32902(g)(2) (2007).
---------------------------------------------------------------------------

    The States and Cities commenters argued that NHTSA had counted 18 
months incorrectly, and that ``18 months prior to September 1, 2021 is 
in fact March 1, 2020.'' \2523\ NHTSA agrees that 18 months prior to 
September 1 would be March 1 of the year prior; the statement in the 
NPRM that ``NHTSA has consistently interpreted the ``beginning of the 
model year'' as September 1 of the CY prior'' was a typographical 
error. As prior Federal Register notices indicate, NHTSA has in fact 
long interpreted the beginning of the model year for CAFE compliance 
purposes as October 1 of the CY prior.\2524\ Thus, counting backwards, 
18 months prior to October 1 is properly identified as April 1, meaning 
that new standards for MY 2022 must be established by April 1, 2020.
---------------------------------------------------------------------------

    \2523\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 78, fn. 211.
    \2524\ See, e.g., 75 FR 25546 (May 7, 2010).
---------------------------------------------------------------------------

    With regard to the amendments to the MY 2021 standards, a coalition 
of environmental groups commented that NHTSA's legal construction of 
EPCA's lead time requirement as not applying to MY 2021 was ``not . . . 
permissible,'' arguing that section 32902(g)(1) only permits amendments 
to existing CAFE standards that ``meet[ ] the requirement of subsection 
(a) or (d) as appropriate,'' and that section 32902(a) requires fuel 
economy standards to be prescribed 18 months before the beginning of 
the model year.\2525\ The environmental group coalition therefore 
argued that the two identified provisions must be read together to 
compel all amendments to standards to be prescribed at least 18 months 
before a model year, and concluded that because it was impossible to 
finish a final rule 18 months before the start of MY 2021, that MY 2021 
standards could not be amended.\2526\ The States and Cities group 
provided similar comments, arguing that NHTSA's interpretation of 
(g)(2) rendered the reference in (g)(1) to (a) ``a nullity,'' and that 
the ``as appropriate'' language in (g)(1) referred to the determination 
of whether providing 18 months of lead time was appropriate, rather 
than to whether (a) or (d) was the relevant provision governing the 
standards in question.\2527\ NCAT commented that ``Congress in Sec.  
32902 has indicated that at least 18 months of lead time are 
appropriate when setting standards,'' and stated that ``Manufacturers' 
need for adequate lead time when designing products and developing 
compliance strategies is the same regardless of whether the agency

[[Page 25123]]

is making standards more stringent, less stringent, or simply changing 
the structure or compliance options provided under the standards.'' 
\2528\ NADA, in contrast, argued that NHTSA does ``have the authority 
and discretion to reopen the MY 2021 standards,'' and that the 
``mandate for at least 18 months of lead time before new standards may 
take effect does not apply to instances, such as for MY 2021, where 
standards are being relaxed.'' \2529\ CEI also agreed with NHTSA's 
interpretation of lead time set forth in the NPRM.\2530\
---------------------------------------------------------------------------

    \2525\ Center for Biological Diversity, Conservation Law 
Foundation, Earthjustice, Environmental Defense Fund, Environmental 
Law and Policy Center, Natural Resources Defense Council, Public 
Citizen, Sierra Club, Union of Concerned Scientists (hereafter, 
``environmental group coalition''), Appendix A, NHTSA-2018-0067-
12000, at 66.
    \2526\ Id.
    \2527\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 78-79.
    \2528\ NCAT, NHTSA-2018-0067-11969, at 46.
    \2529\ NADA, NHTSA-2018-0067-12064, at 9.
    \2530\ CEI, NHTSA-2018-0067-12015, at 3-4.
---------------------------------------------------------------------------

    NHTSA agrees that section 32902(g)(1) states that amendments must 
meet the requirements of subsection (a) or (d) as appropriate, and that 
32902(a) states that standards must be prescribed 18 months in advance 
of the model year. However, NHTSA cannot agree that the 18-month lead 
time requirement applies to amendments to existing standards that 
reduce stringency. Section 32902(g)(2) clearly states that ``[w]hen the 
Secretary of Transportation prescribes an amendment under this section 
that makes an average fuel economy standard more stringent (emphasis 
added), the Secretary shall prescribe the amendment . . . at least 18 
months before the beginning of the model year to which the amendment 
applies.'' Commenters' construction of the statute would render 
superfluous the words ``more stringent'' in 32902(g)(2), and there is a 
presumption against superfluity.\2531\ Congress purposely included the 
words ``more stringent'' in order to exclude the contrary situation--
``less stringent''--from the 18-month lead time requirement. A plain 
reading of (g)(1) simply provides that the Secretary (by delegation, 
NHTSA) should refer to the correct provision depending on whether the 
standard being amended is generally applicable (pointing to section 
(a)) or a standard applicable to low-volume manufacturer pursuant to an 
exemption (pointing to section (d)). Reading (g)(1) and (g)(2) together 
is the appropriate way to give effect to both provisions. This reading 
provides that NHTSA may amend the MY 2021 standard by following the 
requirements for generally-applicable standards; this reading also 
provides that 18 months' lead time is only required for amendments that 
increase stringency. NHTSA also does not agree that (g)(1) can be read 
to imply that the agency must provide 18 months of lead time ``if 
appropriate,'' as the States and Cities suggest, nor that there is any 
statutory basis to extend the lead time requirement to changes to the 
``structure or compliance options provided under the standards'' as 
NCAT suggests. If new off-cycle technologies could not be recognized 
toward compliance without providing 18 months' lead time, manufacturer 
efforts to rely on that compliance flexibility to redress past 
shortfalls would be frustrated.
---------------------------------------------------------------------------

    \2531\ See, e.g., Duncan v. Walker, 533 U.S. 167 (2001) (citing 
U.S. v. Menasche, 348 U.S. 528, 538-539 (1955)).
---------------------------------------------------------------------------

    Moreover, automakers need more time to respond when NHTSA amends 
standards to be more stringent--doing so would likely require 
automakers to change their product and/or sales plans to ensure that 
they will meet more-stringent standards than those standards for which 
they may have already prepared. But such product or sales plans would 
not necessarily need to be changed if standards were amended to be less 
stringent--in fact an automaker would be rewarded by keeping existing 
plans to comply in place with additional bankable and tradable 
overcompliance credits. However, the environmental group coalition 
argued that ``[c]hanging the MY 2021 standard at this late date would 
penalize technologically advanced automakers and parts suppliers, who 
have already made significant investments in updating their 
technology.'' \2532\ The States and Cities group made similar 
comments,\2533\ as did NCAT.\2534\ The environmental group coalition 
further suggested that amending the MY 2021 standard would reduce the 
need for (and thus the value) of overcompliance credits, ``which would 
be disruptive to the manufacturers that have done the most to further 
EPCA's conservation goals.'' \2535\ NCAT made similar comments, arguing 
that ``The practical and financial impact of the change accordingly is 
not materially different from increasing the stringency of a standard 
this late in the product cycle.'' \2536\
---------------------------------------------------------------------------

    \2532\ Environmental group coalition, NHTSA-2018-0067-12000, 
Appendix A, at 66.
    \2533\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 78, fn. 213.
    \2534\ NCAT, NHTSA-2018-0067-11969, at 46-47.
    \2535\ Environmental group coalition, NHTSA-2018-0067-12000, 
Appendix A. at 66-67.
    \2536\ NCAT, NHTSA-2018-0067-11969, at 47.
---------------------------------------------------------------------------

    NHTSA believes that to the extent that some manufacturers have 
already invested in future fuel economy improvements, those 
manufacturers will continue to be well-positioned both to respond to 
increasing standards in the future, and to take advantage of any market 
demand for higher fuel economy/reduced tailpipe CO2 
emissions from consumers who put a premium on those aspects. NHTSA is 
also aware that several companies have self-imposed emissions-reduction 
goals which may drive their decisions on technology application 
regardless of regulatory obligations. NHTSA does not believe that 
companies which have already invested in higher levels of technology 
consider those investments to be bad ones. The agencies note that 
manufacturer commenters, despite the concerns expressed by others, did 
not comment about a lack of lead time associated with changing the MY 
2021 standards; rather, many manufacturer commenters expressly cited 
the need to revise MY 2021 standards, arguing that the previously-
established values are beyond maximum feasible. Regarding the value of 
overcompliance credits under more or less stringent standards, NHTSA 
agrees that the need for credits may be less under less stringent 
standards, but this is true regardless of the lead time question. 
Further, NHTSA does not believe that this suggests only standards that 
compel reliance on overcompliance credits (especially those earned by 
competitors) can be maximum feasible; this topic will be addressed in 
further detail below, and regardless, NHTSA is prohibited from 
considering credit availability in determining maximum feasible CAFE 
standards.
(b) Separate Standards for Cars and Trucks, and Minimum Standards for 
Domestic Passenger Cars
    As discussed above, EPCA requires NHTSA to set separate CAFE 
standards for passenger cars and light trucks for each model 
year.\2537\ NHTSA interprets this requirement as preventing the agency 
from setting a single combined CAFE standard for cars and trucks 
together, based on the plain language of the statute. Congress 
originally required separate CAFE standards for cars and trucks to 
reflect the different fuel economy capabilities of those different 
types of vehicles,\2538\ and over the history of the CAFE program, has 
never revised this requirement. Even as many cars and trucks have come 
to resemble each other more closely over time--many crossover and 
sport-utility models, for example, come in versions today that may be 
subject to either the car standards or the truck standards depending on 
their characteristics--it is still accurate to say that vehicles with

[[Page 25124]]

truck-like characteristics such as 4 wheel drive, cargo-carrying 
capability, etc., consume more fuel per mile than vehicles without 
these characteristics. Thus, NHTSA believes that the different fuel 
economy capabilities of cars and trucks would generally make separate 
standards appropriate for these different types of vehicles, regardless 
of the plain language of the statute which requires such treatment.
---------------------------------------------------------------------------

    \2537\ 49 U.S.C. 32902(b)(1) (2007).
    \2538\ Indeed, EPCA initially only required NHTSA to establish 
CAFE standards for passenger cars; establishment of light truck 
standards was permissible.
---------------------------------------------------------------------------

    EPCA, as amended by EISA, also requires another separate standard 
to be set for domestically-manufactured \2539\ passenger cars. Unlike 
standards for passenger cars and light trucks described above, the 
compliance burden of the minimum domestic passenger car standard is the 
same for all manufacturers: The statute clearly states that any 
manufacturer's domestically-manufactured passenger car fleet must meet 
the greater of either 27.5 mpg on average, or
---------------------------------------------------------------------------

    \2539\ In the CAFE program, ``domestically-manufactured'' is 
defined by Congress in 49 U.S.C. 32904(b). The definition roughly 
provides that a passenger car is ``domestically manufactured'' as 
long as at least 75% of the cost to the manufacturer is attributable 
to value added in the United States, Canada, or Mexico, unless the 
assembly of the vehicle is completed in Canada or Mexico and the 
vehicle is imported into the United States more than 30 days after 
the end of the model year.

92 percent of the average fuel economy projected by the Secretary 
for the combined domestic and non-domestic passenger automobile 
fleets manufactured for sale in the United States by all 
manufacturers in the model year, which projection shall be published 
in the Federal Register when the standard for that model year is 
promulgated in accordance with [49 U.S.C. 32902(b)].\2540\
---------------------------------------------------------------------------

    \2540\ 49 U.S.C. 32902(b)(4) (2007).

    Since that requirement was promulgated, the ``92 percent'' has 
always been greater than 27.5 mpg. NHTSA published the 92-percent 
minimum domestic passenger car standards for model years 2017-2025 at 
49 CFR 531.5(d) as part of the 2012 final rule. For MYs 2022-2025, 
531.5(e) states that these were to be applied if, when actually 
proposing MY 2022 and subsequent standards, the previously identified 
standards for those years are deemed maximum feasible, but if NHTSA 
determines that the previously identified standards are not maximum 
feasible, the 92-percent minimum domestic passenger car standards would 
also change. This is consistent with the statutory language that the 
92-percent standards must be determined at the time an overall 
passenger car standard is promulgated and published in the Federal 
Register. Thus, any time NHTSA establishes or changes a passenger car 
standard for a model year, the minimum domestic passenger car standard 
for that model year will also be evaluated or reevaluated and 
established accordingly. NHTSA explained this in the rulemaking to 
establish standards for MYs 2017 and beyond and received no 
comments.\2541\
---------------------------------------------------------------------------

    \2541\ 77 FR 62624, 63028 (Oct. 15, 2012).
---------------------------------------------------------------------------

    The 2016 Alliance/Global petition for rulemaking asked NHTSA to 
revise the 92-percent minimum domestic passenger car standards 
retroactively for MYs 2012-2016 ``to reflect 92 percent of the required 
average passenger car standard taking into account the fleet mix as it 
actually occurred, rather than what was forecast.'' The petitioners 
stated that doing so would be ``fully consistent with the statute.'' 
\2542\
---------------------------------------------------------------------------

    \2542\ Automobile Alliance and Global Automakers Petition for 
Direct Final Rule with Regard to Various Aspects of the Corporate 
Average Fuel Economy Program and the Greenhouse Gas Program (June 
20, 2016) at 5, 17-18, available at https://www.epa.gov/sites/production/files/201609/documents/petition_to_epa_from_auto_alliance_and_global_automakers.pdf 
(hereinafter Alliance/Global Petition).
---------------------------------------------------------------------------

    NHTSA explained in the NPRM that NHTSA understood that determining 
the 92 percent value ahead of the model year to which it applies, based 
on the information then available to the agency, would result in a 
different mpg number than if NHTSA determined the 92 percent value 
based on the information available at the end of the model year in 
question. NHTSA further explained that it understood that determining 
the 92 percent value ahead of time could make the minimum domestic 
passenger car standard more stringent than it could be if it were 
determined at the end of the model year, if manufacturers end up 
producing more larger-footprint passenger cars than what NHTSA had 
originally anticipated.
    Accordingly, NHTSA sought comment on the request by Alliance/
Global. Additionally, recognizing the uncertainty inherent in 
projecting specific values far into the future, NHTSA also sought 
comment on whether it is possible to define the 92 percent valueas a 
range, if NHTSA defined the values associated with a CAFE standard 
(i.e., the footprint curve) as a range rather than as a single number. 
NHTSA referred to the sensitivity analysis included in the proposal and 
in the accompanying PRIA as a basis for such an mpg range ``defining'' 
the passenger car standard in any given model year. If NHTSA took that 
approach, 92 percent of that ``standard'' would also, necessarily, be a 
range. NHTSA broadly sought comment on that approach or other similar 
approaches.
    The Alliance and FCA commented that they ``supported the NHTSA 
proposal'' to calculate 92 percent as a range rather than as a single 
value, with the ultimate minimum domestic passenger car standard to be 
determined at the end of the MY to which it applies.\2543\ Both 
organizations cited compliance difficulties when the 92 percent 
calculated at the time of the rulemaking turns out to be more stringent 
than 92 percent of the final MY compliance obligations for passenger 
cars, and argued that minimum domestic passenger car standards should 
be recalculated as part of this rulemaking for all model years, rather 
than only MYs 2021-2026, in order to ameliorate that compliance 
difficulty retroactively. The Alliance argued that the 18 month lead 
time requirement should not be interpreted to apply to the minimum 
domestic passenger car standards, because if the 92 percent value is a 
range like the overall passenger car curve, then that value cannot be 
determined until after the model year is completed.\2544\ Because 
manufacturers' individual compliance obligations are not subject to the 
18 month lead time requirement, the Alliance requested that the 92 
percent should similarly not be.\2545\ Separately, Kreucher commented 
that NHTSA should expand the credit transfer provision to allow 
transferred credits to be used to meet the minimum domestic passenger 
car standard.\2546\
---------------------------------------------------------------------------

    \2543\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 41; 
FCA, NHTSA-2018-0067-11943, at 64.
    \2544\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 42-
43.
    \2545\ Id.
    \2546\ Kreucher, NHTSA-2018-0067-0444, at 11.
---------------------------------------------------------------------------

    In contrast, the States and Cities and ACEEE opposed changes to the 
minimum domestic passenger car standard, with the States and Cities 
commenting that NHTSA ``is proposing to retroactively revise the 92 
percent based on actual fleet mix'' \2547\ and ACEEE simply noting that 
the Alliance/Global had requested that NHTSA do this.\2548\ ACEEE 
stated that NHTSA did not have discretion to alter the statutory 
requirement, and argued that calculating 92 percent at the end of the 
model year was ``entirely counter to the intent of the law--the so-
called backstop is designed explicitly to protect against the market 
shifts for which the [industry is] asking the standard to be 
adjusted.'' \2549\

[[Page 25125]]

The States and Cities similarly argued that ``the 92 percent 
requirement is expressly intended to be a projection, not a 
retrospective recalculation,'' and ``the statute does not contemplate a 
`range,' but rather a `minimum' with a set value--92 percent. If 
Congress had intended the value to be a range, it would have included 
that language in the statute, and would not have determined the value 
with such specificity.'' \2550\
---------------------------------------------------------------------------

    \2547\ States and Cities, NHTSA-2018-0067-11735, at 79.
    \2548\ ACEEE, NHTSA-2018-0067-12122, Attachment (joint NGO 
comment to manufacturer petition for flexibilities), at 15.
    \2549\ Id. ACEEE cited a NHTSA statement in the 2010 final rule 
establishing standards for MYs 2012-2016 in support of this 
argument, noting that NHTSA had said ``this minimum standard was 
intended to act as a `backstop,' ensurng that domestically-
manufactured passenger cars reached a given mpg level even if the 
market shifted in ways likely to reduce overall fleet mpg.'' Id. 
(emphasis added).
    \2550\ States and Cities, NHTSA-2018-0067-11735, at 79.
---------------------------------------------------------------------------

    NHTSA considered comments about setting the MDPCS as a range. NHTSA 
recognizes that the approach discussed in the NPRM may not be within 
our statutory authority and therefore is setting the standards as 
specific values.
    NHTSA agrees that setting the MDPCS after the model year is 
completed and the total passenger car fleet standard is known would 
provide standards that adapt with changes in consumer demand. However, 
such an approach would not establish the final numerical value until 
significantly after the model year completed, only after final 
compliance data has been submitted by all manufacturers and EPA and 
NHTSA have completed compliance work for the total passenger car fleet. 
In addition, the standard would be based on the production of all 
manufacturers of passenger cars, providing no means for an individual 
manufacturer to have certainty over its final standard. Individual 
manufacturers likewise would have no control over the value by 
controlling their production mix. For these reasons, NHTSA is denying 
the Alliance/Global petition that the 92 percent value for the MDPCS be 
determined based on the information available at the end of the model 
year in question.
    That said, NHTSA agrees that the actual total passenger car fleet 
standards have differed significantly the 2012 projection, and examined 
the projections from past rulemakings in greater detail. NHTSA reviewed 
the total passenger car fleet (all domestic and import passenger cars) 
standard that was projected at the time of rulemakings for MYs 2011 to 
2018 and compared those projections to the actual total fleet passenger 
car standard for each of those model years from compliance data, based 
on the actual footprints and production volume of the models produced 
in those model years. Table VIII-1 shows the projected standards and 
the actual standards on a fuel economy basis, and Table VIII-2 shows 
the fuel economy values converted to fuel consumption values which was 
used as the basis for and analyzing the differences between the 
projected standards and actual standards.\2551\ Table VIII-2 also shows 
the percentage difference between the total passenger car fleet 
standard at the time of the rulemaking and the actual fleet standard 
based on compliance data.
---------------------------------------------------------------------------

    \2551\ Consistent with EPCA/EISA and corresponding regulations, 
CAFE compliance calculations have been conducted on a mile per 
gallon basis. However, engineering computations have almost 
exclusively been conducted on a fuel consumption basis (i.e., in 
gallons per mile), because the underlying engineering relationships 
are more meaningfully defined on a fuel consumption basis.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 25126]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.729

[GRAPHIC] [TIFF OMITTED] TR30AP20.730


[[Page 25127]]


    The data show that the standards projected in 2012 were 
consistently more stringent than the actual standards, by an average of 
1.9 percent. This difference indicates that in rulemakings conducted in 
2009 through 2012, the agencies' projections of passenger car vehicle 
footprints and production volumes consistently underestimated the 
consumer demand for larger passenger cars over the MYs 2011 to 2018 
period.
BILLING CODE 4910-59-C
    To establish minimum standards for domestic passenger cars in these 
past rulemakings, NHTSA computed the average of manufacturers' 
requirements given the attribute-based standards being issued, and 
given the projected distribution of passenger car footprints as 
indicated in the analysis fleet (aka market forecast) used to analyze 
impacts of the standards. The joint NHTSA-EPA rulemaking establishing 
standards for MYs 2012-2016 presented analysis that, in turn, used a 
``2008-based'' market forecast that combined detailed information 
regarding the MY 2008 fleet with a commercial market forecast (by brand 
and segment) and a range of agency assumptions. Importantly, the 
commercial market forecast showed Chrysler's production falling 
dramatically, and never recovering; as well as Chrysler passenger cars 
being distributed more than most OEMs (other than Jaguar and Mercedes) 
toward larger footprints, and this forecast impacted the NHTSA's 
projection of overall average requirements for passenger cars under the 
footprint-based standards. For example, the 2008-based forecast showed 
production of Chrysler brands (Chrysler, Dodge, Jeep, and Ram) for the 
U.S. market totaling 0.8 million units by MY 2017, and today's analysis 
fleet uses a MY 2017 fleet showing 1.9 million Chrysler-branded units. 
Also, among the agencies' assumptions, was that some manufacturers 
(Chrysler, Ford, Subaru, Mazda, and Mitsubishi) would rapidly increase 
production of small footprint vehicles not observed in the MY 2008 
fleet.
    The joint rulemaking establishing standards for MYs 2017-2025 also 
used this 2008-based fleet for the NPRM, showing more than 1.3 million 
units smaller than 41 square feet in MY 2017, far more than the 0.3m 
units shown in the model inputs for today's analysis. For the 2012 
final rule, the agencies conducted side-by-side analysis, one using the 
2008-based fleet, and one using a 2010-based fleet. The 2010-based 
fleet used a newer commercial forecast that was considerably more 
sanguine regarding, for example, FCA's prospects. Minimum standards for 
domestic passenger cars were based on an average of results for the 
2008-based and 2010-based total passenger car fleets.
    The analysis fleet underlying today's reference case analysis is 
discussed above in Section VI.A.2 and available in full detail with the 
model inputs and outputs accompanying today's notice.\2552\ For the 
current rulemaking, NHTSA also considered that, unlike the passenger 
car standards and light truck standards which are vehicle attribute-
based and automatically adjust with changes in consumer demand, that 
MDPCSs are not attribute-based, and therefore do not adjust with 
changes in consumer demand. They are fixed standards that are 
established at the time of the rulemaking. The MYs 2011-2018 MDPCS were 
more stringent and placed more burden on manufacturers of domestic 
passenger cars than was projected and expected at the time of the 
rulemakings. NHTSA agrees with the Alliance's concerns over the impact 
of changes in consumer demand on manufacturers' ability to comply with 
the MDPCS and in particular, manufacturers that produce larger 
passenger cars domestically.
---------------------------------------------------------------------------

    \2552\ https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
---------------------------------------------------------------------------

    Additionally, as discussed in more detail in Section VIII.B.4 
below, consumer demand may shift even more in the direction of larger 
passenger cars if fuel prices continue to remain low. The fuel prices 
used in the analysis for this final rule rely on EIA's future forecasts 
of fuel prices, which were made prior to the recent collapse of oil 
prices. If the former OPEC+ members continue to pursue market share, 
fuel prices will likely continue to drop. If, instead of pursuing 
market share, they try to control prices restricting supply, U.S. shale 
production could begin to ramp back up and exert downward pressure on 
price. If fuel prices end up even lower than our analysis assumes, 
benefits from saving additional fuel will be worth even less to 
consumers. Our analysis captures none of these effects. Sustained low 
oil prices can be expected to have real effects on consumer demand for 
additional fuel economy, and consumers may foreseeably be even less 
interested in smaller passenger cars than they are at present.
    To help avoid similar outcomes in the rulemaking timeframe to what 
has happened with the MDPCS over the last several model years, NHTSA 
determined it is reasonable and appropriate to consider the recent 
projection errors as part of estimating the projected total passenger 
car fleet fuel economy for MYs 2021-2026. As stated above the average 
difference over MYs 2011-2018 was 1.9 percent. As explained above, 
those differences are largely attributable to aspects of the forecasts 
that turned out to be far different from reality. NHTSA is projecting 
the total passenger car fleet fuel economy using the central analysis 
value in each model year and applying an offset based on the historical 
1.9 percent difference identified for MYs 2011-2018. Table VIII-3 hows 
the calculation values used to determine the total passenger car fleet 
fuel economy value for each model year.
    NHTSA will continue its practice of determining the MDPCS as 
specific values at the same time that it sets passenger car standards, 
at 92 percent of the projected passenger cars standard in each model 
year. Table VIII-3 also shows the computations for the MDPCS for each 
model year. The new MDPCS are prescribed in the regulatory text below.

[[Page 25128]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.731

    Table VIII-4 lists the minimum domestic passenger car standards 
reflecting the updated analysis discussed above, and comparing these to 
standards that would correspond to each of the other regulatory 
alternatives considered. NHTSA has updated these to reflect its overall 
analysis and resultant projection for the CAFE standards finalized 
today, highlighted below as ``Preferred (Alternative 3),'' and has 
calculated what those standards would be under the no action 
alternative (as issued in 2012, as updated for the NPRM, and as further 
updated by today's analysis) and under the other alternatives described 
and discussed further in Section V, above. As explained in a separate 
memorandum to the document, while the CAFE Model analysis underlying 
the FEIS, FRIA, and final rule does not reflect this change, separate 
analysis that does reflect the change demonstrates that doing so does 
not change estimated impacts of any of the regulatory alternatives 
under consideration.
[GRAPHIC] [TIFF OMITTED] TR30AP20.732


[[Page 25129]]


Attribute-Based and Defined by Mathematical Function
    EISA requires NHTSA to set CAFE standards that are ``based on 1 or 
more attributes related to fuel economy and express[ed] . . . in the 
form of a mathematical function.'' \2553\ Historically, NHTSA has based 
standards on vehicle footprint and proposes to continue to do so for 
all the reasons described in previous rulemakings. As in previous 
rulemakings, NHTSA proposed to define the standards in the form of a 
constrained linear function that generally sets higher (more stringent) 
targets for smaller-footprint vehicles and lower (less stringent) 
targets for larger-footprint vehicles. These footprint curves are 
discussed in much greater detail in Section V above. NHTSA sought 
comment both on the choice of footprint as the relevant attribute and 
on the rationale for the constrained linear functions chosen to 
represent the standards; those comments and NHTSA's responses are 
discussed above in Section V.
---------------------------------------------------------------------------

    \2553\ 49 U.S.C. 32902(b)(3)(A).
---------------------------------------------------------------------------

d) Number of Model Years for Which Standards May Be Set at a Time
    EISA also states that NHTSA shall ``issue regulations under this 
title prescribing average fuel economy standards for at least 1, but 
not more than 5, model years.'' \2554\ In the 2012 final rule, NHTSA 
interpreted this provision as preventing the agency from setting final 
standards for all of MYs 2017-2025 in a single rulemaking action, so 
the MYs 2022-2025 standards were termed ``augural,'' meaning ``that 
they represent[ed] the agency's current judgment, based on the 
information available to the agency [then], of what levels of 
stringency would be maximum feasible in those model years.'' \2555\ 
That said, NHTSA also repeatedly clarified that the augural standards 
were in no way final standards and that a future de novo rulemaking 
would be necessary in order both to propose and to promulgate final 
standards for MYs 2022-2025.
---------------------------------------------------------------------------

    \2554\ 49 U.S.C. 32902(b)(3)(B).
    \2555\ 77 FR 62623, 62630 (Oct. 15, 2012).
---------------------------------------------------------------------------

    In the NPRM, NHTSA proposed to establish new standards for MYs 
2022-2026 and to revise the previously-established final standards for 
MY 2021. NHTSA explained that legislative history suggests that 
Congress included the five year maximum limitation so NHTSA would issue 
standards for a period of time where it would have reasonably realistic 
estimates of market conditions, technologies, and economic 
practicability (i.e., not set standards too far into the future).\2556\ 
However, NHTSA suggested that the concerns Congress sought to address 
by imposing those limitations are not present for nearer model years 
where NHTSA already has existing standards, and noted that revisiting 
existing standards is contemplated by both 49 U.S.C. 32902(c) and 
32902(g). NHTSA stated that the agency therefore believed that it is 
reasonable to interpret section 32902(b)(3)(B) as applying only to the 
establishment of new standards rather than to the combined action of 
establishing new standards and amending existing standards.
---------------------------------------------------------------------------

    \2556\ See 153 Cong. Rec. 2665 (Dec. 28, 2007).
---------------------------------------------------------------------------

    Moreover, NHTSA argued, it would be an absurd result if the five 
year maximum limitation were interpreted to prevent NHTSA from revising 
a previously-established standard that the agency had determined to be 
beyond maximum feasible, while concurrently setting five years of 
standards not so distant from today. The concerns Congress sought to 
address are much starker when NHTSA is trying to determine what 
standards would be maximum feasible 10 years from now as compared to 
three years from now.
    NADA commented that NHTSA has discretion and authority to set 
standards for MY 2026 and that the ``statutory five-year rule is not a 
barrier to doing so,'' \2557\ while the environmental group coalition 
argued that NHTSA ``is limited to prescribing fuel economy standards 
for only five model years at a time,'' but ``[h]ere, NHTSA is setting 
standards for six model years, 2021 through 2026. This exceeds NHTSA's 
statutory authority.'' \2558\ Consumers Union argued that ``[i]f 
Congress had intended the statute to only apply to the establishment of 
new standards, as the agencies contend, it certainly could have stated 
as such. But Congress did not include any language even hinting at this 
interpretation.'' \2559\
---------------------------------------------------------------------------

    \2557\ NADA, NHTSA-2018-0067-12064, at 9.
    \2558\ Environmental group coalition, NHTSA-2018-0067-12000, at 
66.
    \2559\ Consumers Union, NHTSA-2018-0067-12068, Attachment A, at 
24.
---------------------------------------------------------------------------

    NHTSA continues to believe, consistent with the legislative 
history, that the five year limitation was intended to prevent NHTSA 
from setting standards too far into the future, recognizing that 
predicting the future is difficult. Consumers Union is correct that 
nothing in the statute compels the interpretation that the five year 
limitation applies only to the setting of new standards rather than to 
the combined action of establishing new standards and amending existing 
standards, but NHTSA does not believe that the statute precludes this 
interpretation, either. The statute allows NHTSA to revisit existing 
standards; the statute separately allows NHTSA to prescribe new 
standards for at least 1, but not more than 5, model years when it 
``issues regulations.'' It is not clear whether the statute precludes 
multiple concurrent or quickly-sequential rulemakings ``issuing 
regulations'' for different periods of time. If this approach were 
used, for example, to try to set ten years' worth of CAFE standards 
essentially at once, this would appear directly contrary to the 
statute. If this approach were used to revisit an existing standard and 
then (in a separate rulemaking) set five years' worth of standards for 
the immediately ensuing model years, this would seem consistent with 
Congressional intent, but an unnecessary use of tax dollars that could 
be saved by consolidating agency (and commenter) work into a single 
rulemaking action. NHTSA does not believe that Congress intended to 
force the agency to waste resources, and continues to believe that the 
current interpretation is reasonable and appropriate.
(e) Maximum Feasible Standards
    As discussed above, EPCA requires NHTSA to consider four factors in 
determining what levels of CAFE standards would be maximum feasible, 
and NHTSA presents in the sections below its understanding of the 
meaning of those four factors. All factors should be considered, in the 
manner appropriate, and then the maximum feasible standards should be 
determined.
(1) Technological Feasibility
    ``Technological feasibility'' refers to whether a particular method 
of improving fuel economy is available for deployment in commercial 
application in the model year for which a standard is being 
established. Thus, NHTSA is not limited in determining the level of new 
standards to technology that is already being commercially applied at 
the time of the rulemaking. For the proposal, NHTSA explained that it 
had considered a wide range of technologies that improve fuel economy, 
subject to the constraints of EPCA regarding how to treat alternative 
fueled vehicles, such as battery-electric vehicles, in determining 
maximum feasible standards, and considering the need to account for 
which technologies have already been applied to which vehicle model/
configuration, and the need to realistically estimate the cost and fuel 
economy impacts of each technology.

[[Page 25130]]

NHTSA explained that it had not attempted to account for every 
technology that might conceivably be applied to improve fuel economy 
and considered it unnecessary to do so given that many technologies 
address fuel economy in similar ways.\2560\ NHTSA noted that 
technological feasibility and economic practicability are often 
conflated, trying to explain that the question of whether a fuel-
economy-improving technology does or will exist (technological 
feasibility) is a different question from what economic consequences 
could ensue if NHTSA effectively requires that technology to become 
widespread in the fleet and the economic consequences of the absence of 
consumer demand for technology that are projected to be required 
(economic practicability). NHTSA explained that it is therefore 
possible for standards to be technologically feasible but still beyond 
the level that NHTSA determines to be maximum feasible due to 
consideration of the other relevant factors.
---------------------------------------------------------------------------

    \2560\ For example, NHTSA has not considered high-speed 
flywheels as potential energy storage devices for hybrid vehicles; 
while such flywheels have been demonstrated in the laboratory and 
even tested in concept vehicles, commercially available hybrid 
vehicles currently known to NHTSA use chemical batteries as energy 
storage devices, and the agency has considered a range of hybrid 
vehicle technologies that do so.
---------------------------------------------------------------------------

    The States and Cities commenters argued that NHTSA's interpretation 
of the technological feasibility factor was unreasonable, stating that 
``. . . fuel economy standards under EPCA are 'intended to be 
technology forcing' because Congress recognized 'that 'market forces . 
. . may not be strong enough to bring about the necessary fuel 
conservation which a national energy policy demands.' '' \2561\ The 
States and Cities commenters thus argued that all alternatives less 
stringent than the baseline/augural standards alternative were 
unacceptable because they would not force technologies to be developed 
and applied, and NHTSA had ``conce[ded] that the technology already 
exists that could meet the more stringent augural standards.'' \2562\ 
These commenters stated that ``NHTSA is therefore impermissibly and 
unreasonably (and even implicitly) re-interpreting this factor in a 
manner contrary to the plain meaning of 'feasibility' and ignoring 
EPCA's technology-forcing purpose. See Chevron, 467 U.S. at 843; Fox 
Television, 556 U.S. at 515 (`An agency may not . . . depart from a 
prior policy sub silentio.').'' CARB \2563\ and CBD et al.\2564\ also 
argued that EPCA was intended to be technology forcing.
---------------------------------------------------------------------------

    \2561\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 66, citing CAS, 793 F.2d at 1339 (citing S. Rep. No. 
179, 94th Cong., 1st Sess. 2 (1975) at 9).
    \2562\ Id. at 66.
    \2563\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84 
(``Since market inefficiencies may preclude sufficient improvement 
without regulatory incentives, EPCA requires standards that advance 
technology. (Citing CAS v. NHTSA, 793 F.2d 1322, 1339, citing S. 
Rep. No. 179, 94th Cong., 1st Sess. 2 (1975), U.S.C.C.A.N. 1975 at 
9)'').
    \2564\ CBD et al., NHTSA-2018-0067-12057, at 2.
---------------------------------------------------------------------------

    The States and Cities commenters also argued that NHTSA had 
previously stated in rulemakings that it considered ``all types of 
technologies that improve real-world fuel economy,'' but in the NPRM 
NHTSA stated instead that it had ``not attempted to account for every 
technology that might conceivably be applied to improve fuel economy 
and consider[ed] it unnecessary to do so given that many technologies 
address fuel economy in similar ways.'' \2565\ The States and Cities 
commenters stated that ``[t]his is an unexplained departure from the 
agency's past practice and prior interpretation of `technological 
feasibility,A' citing Fox Television, and argued that NHTSA had not 
explained ``1) what `similar ways' means, or 2) why the fact that a 
technology that might improve fuel economy `in similar ways' to another 
technology obviates NHTSA's obligation to consider its availability, 
particularly given the differences in costs between different 
technologies.'' \2566\ The States and Cities commenters pointed to the 
examples of HCR1 and HCR2 as technologies ``already widely available in 
the market'' that should have been considered, and claimed that NHTSA 
had ``failed to even consult with EPA regarding which technologies the 
agency considered,'' ``result[ing] in fundamentally flawed predictions 
of what technology can be applied in model years 2021-2026.'' \2567\
---------------------------------------------------------------------------

    \2565\ Id. at 67, referring to 83 FR at 43208.
    \2566\ Id.
    \2567\ Id.
---------------------------------------------------------------------------

    Mazda, in contrast, stated that it agreed that ``mere development 
and introduction of advanced fuel efficient technologies is not 
sufficient for manufacturers to comply with established GHG and fuel 
efficiency standards. The technologies must be widely adopted by 
consumers for them to provide the expected environmental benefit.'' 
\2568\ Mr. Kreucher stated that manufacturers have been applying 
``unprecedented levels of technology'' but are still falling short of 
their compliance obligations, pointing in particular to light truck 
compliance in MY 2016. Kreucher argued that ``[t]his indicates a 
serious overestimation of technological feasibility in the prior [2012] 
analysis that must be corrected.'' \2569\
---------------------------------------------------------------------------

    \2568\ Mazda, NHTSA-2018-0067-11727, at 2.
    \2569\ Kreucher, NHTSA-2018-0067-0444, at 7.
---------------------------------------------------------------------------

    UCS stated that the NPRM analysis ``undermined'' an assessment of 
``technical feasibility,'' by ``paint[ing] fuel-saving technologies as 
less effective and more costly than real-world data indicate,'' through 
several mechanisms.\2570\ First, UCS argued that the analysis had 
underestimated ICE efficiency possibilities, ``frequently ignoring 
technology that is already commercialized or is widely anticipated to 
be readily available within the timeframe of the standards.'' \2571\ 
Second, UCS suggested that the NPRM analysis had ``overstate[d] the 
degree to which manufacturers have deployed some of the most cost-
effective technologies, while errors in full vehicle simulation and 
rampant disregard for the current state of technology underestimates 
the potential for future improvement.'' \2572\ UCS claimed that 
``[f]requently the agencies have departed from past precedence in 
specific ways in order to increase technology costs associated with 
technology deployment, sometimes failing to provide even a glimmer of 
reasonable justification for such decisions.'' \2573\ (emphasis added) 
Third, UCS argued that the model had been deliberately constructed to 
avoid choosing the most cost-effective technology pathways, showing 
higher costs and more future overcompliance than UCS analysis 
showed.\2574\ Finally, UCS argued that better modeling of credit 
trading and use would further reduce technology costs. UCS concluded 
that ``The mischaracterization of technology and unrealistic model 
construction lead to an inaccurate assessment of technological 
feasibility, effectively undermining this factor's weight in 
considering maximum feasible standards.'' \2575\
---------------------------------------------------------------------------

    \2570\ UCS, NHTSA-2018-0067-12039, at 4.
    \2571\ Id.
    \2572\ Id.
    \2573\ Id.
    \2574\ Id.
    \2575\ Id.
---------------------------------------------------------------------------

    Contrary to the assertion by several commenters that NHTSA has 
historically claimed that it must set technology-forcing standards, 
NHTSA has previously described the technological feasibility factor as 
allowing the agency to set standards that force the development and 
application of new fuel-efficient technologies.\2576\ In the same 
preamble section in which that description was set forth, NHTSA stated

[[Page 25131]]

that ``[i]t is important to remember that technological feasibility 
must also be balanced with the other of the four statutory factors. 
Thus, while 'technological feasibility' can drive standards higher by 
assuming the use of technologies that are not yet commercial, 'maximum 
feasible' is also defined in terms of economic practicability, for 
example, which might caution the agency against basing standards (even 
fairly distant standards) entirely on such technologies.'' \2577\ NHTSA 
further stated that ``. . . as the `maximum feasible' balancing may 
vary depending on the circumstances at hand for the model year in which 
the standards are set, the extent to which technological feasibility is 
simply met or plays a more dynamic role may also shift.'' \2578\
---------------------------------------------------------------------------

    \2576\ See, e.g., 77 FR at 63015 (Oct. 15, 2012).
    \2577\ Id.
    \2578\ Id.
---------------------------------------------------------------------------

    NHTSA continues to believe that, for purposes of this rulemaking 
covering standards for MYs 2021-2026, the crucial question is not 
whether technologies exist to meet the standards--they do. The question 
is rather, given that the technology exists, how much of it should be 
required to be added to new cars and trucks in order to conserve more 
energy, and how to appropriately balance additional energy conserved 
and additional cost for new vehicles. Regardless of whether 
technological feasibility allows the agency to set technology-forcing 
standards, technological feasibility does not require, by itself, NHTSA 
to set technology-forcing standards if other statutory factors would 
point the agency in a different direction. NHTSA has expressed this 
interpretation of technological feasibility over the course of multiple 
rulemakings.\2579\ The States and Cities commenters appear, at the 
root, to be contesting the agency's determination of maximum feasible 
standards, by way of arguing that NHTSA must interpret the 
technological feasibility factor as necessarily driving greater energy 
conservation. The balancing of factors to determine maximum feasible 
standards is a separate issue, for which EPCA/EISA gives NHTSA 
considerable discretion.
---------------------------------------------------------------------------

    \2579\ Id., see also 75 FR at 25605 (May 7, 2010).
---------------------------------------------------------------------------

    The States and Cities commenters focus on previous rulemaking 
language when they suggest that the agency was arbitrary and capricious 
for not explaining more fully why it need not expressly evaluate every 
single technology that does or could exist in MYs 2021-2026. While 
NHTSA stated in 2012 that it had ``considered all types of technologies 
that improve real-world fuel economy, including air-conditioner 
efficiency and other off-cycle technology, PHEVs, EVs, and highly-
advanced internal combustion engines not yet in production,'' \2580\ 
that statement was only one in a larger discussion. The 2012 final rule 
also stated expressly that ``[t]here are a number of other potential 
technologies available to manufacturers in meeting the 2017-2025 
standards that the agencies have evaluated but have not considered in 
our final analyses. These include HCCI, 'multi-air', and camless valve 
actuation, and other advanced engines currently under development.'' 
\2581\ (emphasis added) Thus, even under the prior analysis that some 
commenters appear to prefer, it is not entirely correct to say that 
NHTSA had considered all technologies in existence or that could exist, 
because some technologies were clearly and purposely left out of the 
prior rule's analysis. In response to commenters' apparent confusion 
regarding NHTSA's statement that it did not consider technologies that 
improved fuel economy in ``similar ways'' as other technologies 
discussed in the NPRM, the meaning behind that statement was discussed 
at greater length in the section of the NPRM that substantively covered 
those technologies. For example, in discussing the ``HCR2'' technology, 
the agencies explained that while the agencies were not modeling HCR2 
expressly due to concerns that it remained ``entirely speculative,'' 
``[t]he CAFE model allows for incremental improvement over existing 
HCR1 technologies with the addition of improved accessory devices 
(IACC), a technology that is available to be applied on many baseline 
MY 2016 vehicles with HCR1 engines and may be applied as part of a 
pathway of compliance to further improve the effectiveness of existing 
HCR1 engines.'' \2582\ In this and in other instances, technologies 
included in the analysis improved fuel economy in similar ways to other 
technologies not included. Here, HCR1, when combined with IACC, results 
in ``a step past'' HCR1, which is similar to the unproven HCR2. As in 
the 2012 rule, the agencies explained in the NPRM why certain 
technologies were not considered, and sought comment. In response to 
comments received, some technologies have been added to the analysis 
for the final rule. See Section VI for more information.
---------------------------------------------------------------------------

    \2580\ 77 FR at 63037 (Oct. 15, 2012).
    \2581\ 77 FR at 62706 (Oct. 15, 2012).
    \2582\ 83 FR at 43038 (Aug. 24, 2018).
---------------------------------------------------------------------------

    While the agencies respond to many of UCS's analytical concerns in 
Sections IV and VI (which include extensive discussion of changes made 
in response to comments), NHTSA recognizes that some commenters believe 
that more technologies are ``available for deployment'' more widely, 
and sooner, than the final rule's analysis reflects. This question has 
long been a topic of debate in CAFE and CO2 rulemakings--the 
agencies consider which technologies can be applied to which vehicles 
in which model years in order to assess the costs and benefits of 
pushing the industry to reach different levels of standards, which in 
turn helps to inform stringency determinations. In response to 
comments, the agencies have expanded the number of technologies and the 
vehicles to which they may be applied for this final rule, but continue 
to disagree that certain technologies can be applied widely in the 
rulemaking timeframe. NHTSA does not believe, for example, that HCCI 
will be unavailable for widespread application in the rulemaking 
timeframe because it wishes to believe this prediction--NHTSA believes 
it based on the fact that HCCI has been in the research phase for 
several decades, and the only production applications to date use a 
highly-limited version that restricts HCCI combustion to a very narrow 
range of engine operating conditions. Section VI contains further 
discussion of these issues.
(2) Economic Practicability
    ``Economic practicability'' has traditionally referred to whether a 
standard is one ``within the financial capability of the industry, but 
not so stringent as to'' lead to ``adverse economic consequences, such 
as a significant loss of jobs or unreasonable elimination of consumer 
choice.'' \2583\ In evaluating economic practicability, NHTSA considers 
the uncertainty surrounding future market conditions and consumer 
demand for fuel economy alongside consumer demand for other vehicle 
attributes. NHTSA has explained in the past that this factor can be 
especially important during rulemakings in which the auto industry is 
facing significantly adverse economic conditions (with corresponding 
risks to jobs). Consumer acceptability is also a major component of 
economic practicability,\2584\ which can involve

[[Page 25132]]

consideration of anticipated consumer responses not just to increased 
vehicle cost, but also to the way manufacturers may change vehicle 
models and vehicle sales mix in response to CAFE standards. In 
attempting to determine the economic practicability of attribute-based 
standards, NHTSA considers a wide variety of elements, including the 
annual rate at which manufacturers can increase the percentage of their 
fleet that employs a particular type of fuel-saving technology,\2585\ 
and manufacturer fleet mixes. NHTSA also considers the effects on 
consumer affordability resulting from costs to comply with the 
standards, and consumers' valuation of fuel economy, among other 
things.
---------------------------------------------------------------------------

    \2583\ 67 FR 77015, 77021 (Dec. 16, 2002).
    \2584\ See, e.g., Center for Auto Safety v. NHTSA (CAS), 793 
F.2d 1322 (DC Cir. 1986) (Administrator's consideration of market 
demand as component of economic practicability found to be 
reasonable); see also Public Citizen v. NHTSA, 848 F.2d 256 
(Congress established broad guidelines in the fuel economy statute; 
agency's decision to set lower standards was a reasonable 
accommodation of conflicting policies).
    \2585\ For example, if standards effectively require 
manufacturers to make technologies widely available that consumers 
do not want, or to make technologies widely available before they 
are ready to be widespread, NHTSA believes that these standards 
could potentially be beyond economically practicable.
---------------------------------------------------------------------------

    Prior to the MYs 2005-2007 rulemaking under the non-attribute-based 
(fixed value) CAFE standards, NHTSA generally sought to ensure the 
economic practicability of standards in part by setting them at or near 
the capability of the ``least capable manufacturer'' with a significant 
share of the market, i.e., typically the manufacturer whose fleet mix 
was, on average, the largest and heaviest, generally having the highest 
capacity and capability so as not to limit the availability of those 
types of vehicles to consumers. In the first several rulemakings 
establishing attribute-based standards, NHTSA applied marginal cost-
benefit analysis, considering both overall societal impacts and overall 
consumer impacts. Whether the standards maximize net benefits has thus 
been a significant, but not dispositive, factor in the past for NHTSA's 
consideration of economic practicability. Executive Order 12866, as 
amended by Executive Order 13563, states that agencies should ``select, 
in choosing among alternative regulatory approaches, those approaches 
that maximize net benefits . . .'' In practice, however, agencies, 
including NHTSA, must consider that the modeling of net benefits does 
not capture all considerations relevant to economic practicability. 
Therefore, as in past rulemakings, NHTSA explained in the NPRM that it 
was considering net societal impacts, net consumer impacts, and other 
related elements in the consideration of economic practicability.
    NHTSA's consideration of economic practicability depends on a 
number of elements. Expected availability of capital to make 
investments in new technologies matters; manufacturers' expected 
ability to sell vehicles with certain technologies matters; likely 
consumer choices matter; and so forth. NHTSA explained in the NPRM that 
NHTSA's analysis of the impacts of the proposal incorporated 
assumptions to capture aspects of consumer preferences, vehicle 
attributes, safety, and other elements relevant to an impacts estimate; 
but stated that it is difficult to capture every such constraint. 
Therefore, NHTSA explained, it is well within the agency's discretion 
to deviate from the level at which modeled net benefits are maximized 
if the agency concludes that that level would not represent the maximum 
feasible level for future CAFE standards. Economic practicability is 
complex, and like the other factors must also be considered in the 
context of the overall balancing and EPCA's overarching purpose of 
energy conservation. Depending on the conditions of the industry and 
the assumptions used in the agency's analysis of alternative standards, 
NHTSA stated that it could well find that standards that maximize net 
benefits, or that are higher or lower, could be at the limits of 
economic practicability, and thus potentially the maximum feasible 
level, depending on how the other factors are balanced.
    NHTSA also stated in the NPRM that while the agency would discuss 
safety as a separate consideration, NHTSA also considered safety as 
closely related to, and in some circumstances a subcomponent of, 
economic practicability. On a broad level, manufacturers have finite 
resources to invest in research and development. Investment into the 
development and implementation of fuel saving technology necessarily 
comes at the expense of investing in other areas such as safety 
technology. On a more direct level, when making decisions on how to 
equip vehicles, manufacturers must balance cost considerations to avoid 
pricing further consumers out of the market. As manufacturers add 
technology to increase fuel efficiency, they may decide against 
installing additional safety equipment to reduce cost increases. And as 
the price of vehicles increase beyond the reach of more consumers, such 
consumers continue to drive or purchase older, less safe vehicles. In 
assessing practicability, NHTSA also considers the harm to the Nation's 
economy caused by highway fatalities and injuries.
    CARB, the States and Cities commenters, and UCS all commented that 
the NPRM analysis, as the States and Cities put it, had ``inexplicably 
inflat[ed] technology costs and rel[ied] on flawed models to predict 
impacts on vehicle sales.'' \2586\ Both CBD et al. and UCS suggested 
that it was incorrect to assume that manufacturers would pass on 100 
percent of cost increases as price increases to consumers.\2587\ UCS 
further stated that ``The agencies have then strategically excluded 
well-established academic literature to limit the assumptions used to 
define a consumer's willingness to pay in ways that further increase 
costs to consumers and/or decrease the consumer benefits of fuel 
economy and greenhouse gas emissions.'' \2588\ UCS argued that assuming 
full pass-through of cost increases as price increases and assuming 
that consumers may not fully value improvements in fuel economy 
``arbitrar[ily] . . . depress the sales of highly fuel-efficient 
vehicles in the model by systematically negating consumer benefits of 
these vehicles.'' \2589\ The States and Cities further argued that 
NHTSA had not ``substantiated its concern that an increase in new 
vehicle prices would place a particular burden on `low-income 
purchasers,' '' and stated that NHTSA had ``assume[d], without 
explanation, that'' less-stringent fuel economy standards resulted in 
greater net savings for consumers, which NHTSA ``acknowledge[d], 
without justification, `is a significantly different analytical result 
from the 2012 final rule.' '' \2590\ The States and Cities commenters 
implied that this different result and NHTSA's ``failure to acknowledge 
it'' was impermissible under the standard set forth in Fox 
Television.\2591\
---------------------------------------------------------------------------

    \2586\ CARB, NHTSA-2018-0067-11873, at 79-80; States and Cities, 
NHTSA-2018-0067-11735, at 69-70; UCS, NHTSA-2018-0067-12039, at 4.
    \2587\ CBD et al., NHTSA-2018-0067-12057, at 4; UCS, NHTSA-2018-
0067-12039, at 4.
    \2588\ UCS, NHTSA-2018-0067-12039, at 5.
    \2589\ Id.
    \2590\ States and Cities, NHTSA-2018-0067-11735, at 70.
    \2591\ Id.
---------------------------------------------------------------------------

    A number of commenters stated that the NPRM's estimates of job 
losses associated with the proposal conflicted with NHTSA's concerns 
about job losses if more stringent standards were promulgated. CBD et 
al. argued that NHTSA could not reasonably conclude that job losses 
make less-stringent standards more economically practicable than more-
stringent

[[Page 25133]]

standards.\2592\ The States and Cities commenters stated that ``[b]y 
declining to address its own findings of significant job losses in the 
auto sector, NHTSA has ignored an important aspect of the problem and 
failed to propose a `rational connection between the facts found and 
the choice made.' '' \2593\ The States and Cities commenters also 
argued that ``the agency failed to acknowledge or explain its break 
with its own interpretation and practice of considering whether 
standards would cause a `significant loss of jobs.' '' \2594\ Some 
commenters argued that more-stringent standards would create more jobs 
(and conversely, that less-stringent standards would result in job 
losses), primarily for supplier companies,\2595\ and some noted that 
other studies had concluded that more-stringent standards would 
increase employment, citing, for example, the report by Synapse Energy 
Economics, Inc. on ``Cleaner Cars and Job Creation.'' \2596\ Some 
commenters further argued that less-stringent standards would hurt U.S. 
GDP,\2597\ and some argued that they would hurt U.S. industry's 
international competitiveness because other countries/regions have more 
stringent standards, and investment may shift to those countries if 
U.S. standards do not continue to compel it.\2598\ The States and 
Cities commenters stated that failing to address fully ``the negative 
employment and GDP impacts of the Proposed Rollback is an abdication of 
NHTSA's clear statutory duty to consider the economic practicability of 
its proposed standards, and an impermissible interpretation of the 
statutory text.'' \2599\
---------------------------------------------------------------------------

    \2592\ CBD et al., NHTSA-2018-0067-12057, at 4.
    \2593\ States and Cities, NHTSA-2018-0067-11735, at 68 (citing 
State Farm, 463 U.S. at 42).
    \2594\ Id. (citing 83 FR at 43208; Fox Television, 556 U.S. at 
515).
    \2595\ CBD et al., NHTSA-2018-0067-12057; Alliance for Vehicle 
Efficiency, NHTSA-2018-0067-11696, at 3-4; NESCAUM, NHTSA-2018-0067-
11691, at 5.
    \2596\ States and Cities, NHTSA-2018-0067-11735, at 68; UCS, 
NHTSA-2018-0067-12039, at 4.
    \2597\ States and Cities, NHTSA-2018-0067-11735, at 68; UCS, 
NHTSA-2018-0067-12039, at 4.
    \2598\ NESCAUM, NHTSA-2018-0067-11691, at 5; Alliance for 
Vehicle Efficiency, NHTSA-2018-0067-11696, at 4.
    \2599\ States and Cities, NHTSA-2018-0067-11735, at 68 (citing 
49 U.S.C. 32902(f); Chevron, 467 U.S. at 843).
---------------------------------------------------------------------------

    Commenters disagreed on whether and how NHTSA should consider 
consumer demand. Mr. Kreucher, the Texas Congressional 
Delegation,\2600\ and Senator Inhofe,\2601\ among others, all argued 
that considering consumer demand for fuel economy was important, while 
other commenters argued that while it may be permissible for NHTSA to 
consider consumer demand, NHTSA could not elevate that consideration 
above others. CARB and the States and Cities commenters both cited 
language from CAS v. NHTSA for the premise that ``Congress intended 
energy conservation to be a long-term effort that would continue 
through temporary improvements in energy availability. Thus, it would 
clearly be impermissible for NHTSA to rely on consumer demand to such 
an extent that it ignored the overarching goal of fuel conservation.'' 
\2602\ The Minnesota agencies stated that ``making sweeping assumptions 
about consumer preferences should not trump the clear public benefit to 
reducing GHG emissions through these standards.'' \2603\ Mr. Kreucher 
commented, in contrast, that consumer preferences are driven entirely 
by ``[l]ong term fuel price expectations and fuel price alone,'' and 
disagreed with the historical ``implicit assumption that if you build 
it customers will come.'' \2604\
---------------------------------------------------------------------------

    \2600\ Texas Congressional Delegation, NHTSA-2018-0067-1421, at 
1.
    \2601\ Senator Inhofe, NHTSA-2018-0067-1422, at 1.
    \2602\ CAS v. NHTSA, 793 F.2d 1322, 1340 (D.C. Cir. 1986), cited 
by CARB, NHTSA-2018-0067-11873, at 79, and by States and Cities, 
NHTSA-2018-0067-11735, at 69.
    \2603\ Minnesota agencies, NHTSA-2018-0067-11706, at 4.
    \2604\ Kreucher, NHTSA-2018-0067-0444, at 11-12.
---------------------------------------------------------------------------

    The Minnesota agencies argued that focusing on consumer preferences 
represented an ``unreasonable and unprecedented shift in 
interpretation.'' \2605\ The States and Cities commenters stated 
similarly that NHTSA had ``redefined `economically practicable' to 
categorically exclude standards that, based on some unspecified metric, 
`widely apply technologies that consumers do not want,' '' and argued 
that ``NHTSA has offered no explanation for how it would define `wide 
application,' much less how it would supposedly determine what 
consumers do or do not want.'' \2606\ The States and Cities commenters 
argued that it was internally inconsistent (and therefore arbitrary and 
capricious) for NHTSA to rely in its justification on concerns about 
consumer acceptance of technologies, while concurrently ``acknowledging 
the `extensive debate over how much consumers do (and/or should) value 
fuel savings and fuel economy as an attribute in new vehicles.' '' 
\2607\ The States and Cities commenters stated that the NPRM's modeling 
``assume[ed] that consumers assign no value to fuel savings 
whatsoever,'' and that ``This assumption is not only implausible but 
also flies in the face of the Agency's own statements that consumers 
likely value between half of and all future fuel savings.'' \2608\
---------------------------------------------------------------------------

    \2605\ Minnesota agencies, NHTSA-2018-0067-11706, at 4.
    \2606\ States and Cities, NHTSA-2018-0067-11735, at 69 (citing 
State Farm, 463 U.S. at 42-43).
    \2607\ Id. (citing NPRM at 43216; Fox Television, 556 U.S. at 
515, and United States Sugar Corp., 830 F.3d at 650).
    \2608\ Id. at 70 (citing NPRM at 43073).
---------------------------------------------------------------------------

    With regard to whether consumers do want more fuel economy, NESCAUM 
stated that ``the most recent surveys indicate that consumers continue 
to place a high value on fuel efficient vehicles of all types,'' \2609\ 
while Alliance for Vehicle Efficiency stated that ``Consumers have 
adopted incremental changes to new vehicles that increase fuel economy 
that don't compromise on power, size or safety.'' \2610\ The States and 
Cities commenters argued that ``consumer choice is, in fact, enhanced 
by providing consumers with the option of purchasing higher-efficiency 
vehicles.'' \2611\ CBD et al. and the States and Cities commenters 
stated that NHTSA had simply made assertions about consumer demands 
without supporting evidence,\2612\ with the States and Cities 
commenters also arguing that the fuel price assumptions in the NPRM 
were ``unsupported'' and ``contradicted by recent evidence,'' despite 
NHTSA's arguments that low fuel prices made ``fuel efficiency less 
attractive to consumers.'' \2613\ Somewhat in contrast, NESCAUM stated 
that ``[g]iven recent consumer preferences for larger vehicles, 
maximizing fuel efficiency and GHG emission reductions in larger 
footprint vehicles is even more important,'' noting that footprint 
based standards ``are intentionally flexible to accommodate industry 
and consumer preferences.'' \2614\ NESCAUM also stated that many HEV/
PHEV/EV models are now available and that their sales ``reflect[ ] 
growing consumer acceptance of the technology, . . . despite the low 
availability of electric vehicle models in the Northeast Section 177 
States and the auto industry's continuing failure to actively market 
[them].'' \2615\
---------------------------------------------------------------------------

    \2609\ NESCAUM, NHTSA-2018-0067-11691, at 2.
    \2610\ Alliance for Vehicle Efficiency, NHTSA-2018-0067-11696, 
at 2.
    \2611\ States and Cities, NHTSA-2018-0067-11735, at 70.
    \2612\ CBD et al., NHTSA-2018-0067-12057, at 4; States and 
Cities, NHTSA-2018-0067-11735, at 70.
    \2613\ Id.
    \2614\ NESCAUM, NHTSA-2018-0067-11691, at 2.
    \2615\ Id. at 3.

---------------------------------------------------------------------------

[[Page 25134]]

    Regarding the NPRM's statement that safety could be a subcomponent 
of economic practicability, the States and Cities commenters stated 
that this was ``an unreasonable interpretation of this factor, given 
that safety concerns are not discussed in EPCA and have no direct 
correlation to whether a standard is economically practicable.'' \2616\ 
The States and Cities commenters further stated that ``NHTSA has never 
before analyzed safety considerations as falling under this factor, and 
fails to explain its reason for doing so now,'' \2617\ and said that it 
was ``unmoored from reality'' for NHTSA to state without support that 
``[i]nvestment into the development and implementation of fuel saving 
technology necessarily comes at the expense of investing in other areas 
such as safety technology.'' \2618\ The States and Cities commenters 
argued that investment in fuel economy rather than safety ``does not 
explain why safety should be folded into a consideration of whether 
standards are economically practicable.'' \2619\ IPI argued that ``[i]t 
is arbitrary for NHTSA to count alleged safety costs as support for its 
propose [sic] rollback both under the economic practicability factor 
and as its own separate `bolster[ing] factor,' and yet never fully 
monetize climate- and pollution-related deaths and other welfare 
impacts under either the need to conserve energy factor nor under the 
economic practicability factor.'' \2620\
---------------------------------------------------------------------------

    \2616\ States and Cities, NHTSA-2018-0067-11735, at 70 
(``arbitrary and capricious for agency to rely on factors `which 
Congress has not intended it to consider' '') (citing Chevron, 467 
U.S. at 843; State Farm, 463 U.S. at 43).
    \2617\ Id. (citing Fox Television, 556 U.S. at 515).
    \2618\ Id.
    \2619\ Id.
    \2620\ NYU IPI, NHTSA-2018-0067-12213, Appendix, at 6-7.
---------------------------------------------------------------------------

    In response to these comments, NHTSA continues to believe that it 
is reasonable to interpret ``economic practicability'' as the agency 
has long interpreted it: As a question of whether a standard is one 
``within the financial capability of the industry, but not so stringent 
as to'' lead to ``adverse economic consequences, such as a significant 
loss of jobs or the unreasonable elimination of consumer choice.'' 
\2621\ NHTSA disagrees that this interpretation is new or divergent 
from past interpretations of economic practicability--this is, to the 
word, the same interpretation set forth in the 2010 and 2012 final 
rules, and in multiple earlier rules. Commenters disagreeing with the 
NPRM's assessment of economic practicability seem, fundamentally, to be 
disagreeing with how NHTSA applied this interpreted definition of 
economic practicability to the information then before the agency, and 
also with the agency's conclusion of how economic practicability 
weighed against the other statutory factors.
---------------------------------------------------------------------------

    \2621\ 67 FR 77015, 77021 (Dec. 16, 2002).
---------------------------------------------------------------------------

    The following text explains why NHTSA continues to believe that the 
pieces of the analysis it categorizes as relevant to economic 
practicability fit within the long-standing definition of that factor. 
Section VIII.B.4 below will explain how the agency has considered those 
pieces of the analysis in balancing economic practicability with the 
other statutory factors.
    NHTSA has consistently described the manner in which it applies the 
``economic practicability'' factor, and has given considerable weight 
to the phrasing of this description. Parsing the words of this 
description can be useful:
    The core of the description is the phrase ``within the financial 
capability of the industry,'' but not so stringent as to lead to 
``adverse economic consequences.'' The following clause ``such as a 
significant loss of jobs or the unreasonable elimination of consumer 
choice'' is set off by a comma from ``consequences,'' and use of the 
phrase ``such as'' indicates that it is a nonrestrictive clause.\2622\ 
A nonrestrictive clause means that ``significant loss of jobs'' and 
``unreasonable elimination of consumer choice'' are examples of 
``adverse economic consequences,'' but are not an exclusive list of the 
possible adverse economic consequences that NHTSA may consider. Further 
evidence that this clause was intended simply to offer examples comes 
from the 1977 final rule establishing passenger car standards for MYs 
1981-1984, in which NHTSA examined the potential meaning of ``economic 
practicability'' at length and concluded that it should be interpreted 
as ``requiring the standards to be within the financial capability of 
the industry, but not so stringent as to threaten substantial economic 
hardship for the industry,'' i.e., lacking the final clause.\2623\
---------------------------------------------------------------------------

    \2622\ See Strunk, William and E.B. White, The Elements of 
Style, Fourth Edition (2000), Rule 3, at 2-7.
    \2623\ 42 FR 33534, 33537 (Jun. 30, 1977). It is worth noting 
that the agency considered and rejected an interpretation of 
economic practicability at that time based solely on cost-benefit 
analysis, stating ``A cost-benefit analysis would be useful in 
considering these factors [of economic practicability], but sole 
reliance on such an analysis would be contrary to the mandate of the 
act.'' Id.
---------------------------------------------------------------------------

    A number of commenters took issue with NHTSA's consideration of 
consumer demand, citing the 1986 D.C. Circuit decision CAS v. NHTSA for 
the proposition that consumer demand cannot drive the balancing of 
factors in determining maximum feasible standards. In that case, the 
D.C. Circuit stated that ``[i]t is axiomatic that Congress intended 
energy conservation to be a long term effort that would continue 
through temporary improvements in energy availability. Thus, it would 
clearly be impermissible for NHTSA to rely on consumer demand to such 
an extent that it ignored the overarching goal of fuel conservation.'' 
\2624\ NHTSA agrees that the CAS decision makes this point, and that 
the 9th Circuit decision in CBD v. NHTSA also underscored that the 
overarching purpose of EPCA is energy conservation. That said, the CAS 
decision also contains a number of other points that are relevant both 
to the facts at hand in this rulemaking and NHTSA's current use of 
consumer demand as an aspect of economic practicability and as a 
consideration in determining maximum feasible standards. NHTSA will 
discuss CAS more extensively below in Section VIII.B.4, but this 
section will cover it briefly, specifically with respect to NHTSA's 
interpretation of economic practicability.
---------------------------------------------------------------------------

    \2624\ CAS, 793 F.2d 1322, 1340 (D.C Cir. 1986).
---------------------------------------------------------------------------

    As noted in the NPRM and in the 2012 final rule, the CAS decision 
found NHTSA's consideration of market demand as a component of economic 
practicability reasonable.\2625\ In CAS, petitioners the Center for 
Auto Safety, Public Citizen, Union of Concerned Scientists, and 
Environmental Policy Institute sued NHTSA over CAFE standards for MY 
1986, arguing that NHTSA could not determine stringency on the basis of 
low expected consumer demand for fuel economy, and ``that technology 
permitted greater fuel savings and that the statutorily required 
`maximum feasible' level of fuel economy is higher than the standard'' 
determined by NHTSA.\2626\ The court followed Chevron in evaluating 
whether NHTSA could consider consumer demand, and found that Congress 
had not directly spoken to the consideration of consumer demand. The 
court then assessed whether NHTSA's interpretation of the statute 
``represents a reasonable accommodation of conflicting policies that 
were committed to the agency's care by statute,'' stating that ``The 
agency's interpretation of the statutory requirements is due 
considerable deference and must be

[[Page 25135]]

found adequate if it falls within the range of permissible 
constructions.'' \2627\
---------------------------------------------------------------------------

    \2625\ 83 FR at 43208, fn. 402; 77 FR at 62668, fn. 111 (both 
citing CAS, 793 F.2d 1322, 1338 (D.C. Cir. 1986)).
    \2626\ CAS, at 1328.
    \2627\ CAS, at 1338.
---------------------------------------------------------------------------

    In assessing NHTSA's interpretation, the court stated that 
``Consumer demand is not specifically designated as a factor, but 
neither is it excluded from consideration; the factors of 
`technological feasibility' and `economic practicability' are each 
broad enough to encompass the concept. Thus, the unadorned language of 
the statute does not indicate a congressional intent concerning the 
precise objections raised by the petitioners.'' The court then examined 
EPCA's legislative history and concluded that ``this language neither 
precludes nor requires lower standards when consumer demand for heavy 
vehicles is strong. The agency is directed to weigh the `difficulties 
of individual automobile manufacturers;' there is no reason to conclude 
that difficulties due to consumer demand for a certain mix of vehicles 
should be excluded.'' \2628\ The court even noted that ``the 
petitioners [did] not challenge the consideration of consumer demand 
per se, but rather the weight the agency has given the factor in 
downgrading standards . . . .'' \2629\
---------------------------------------------------------------------------

    \2628\ CAS, at 1338-1339.
    \2629\ CAS, at 1340.
---------------------------------------------------------------------------

    NHTSA continues to believe that it is reasonable to consider 
consumer demand as an element of economic practicability, as the CAS 
court recognized. Comments objecting to the consideration of consumer 
demand appear to focus more, like the petitioners in CAS, on the 
agency's focus on consumer demand in the overall balancing of factors 
to determine what CAFE standards would be maximum feasible, insofar as 
they are expressing concern about consumer demand undermining energy 
conservation. Again, this question will be addressed further in Section 
VIII.B.4 below. To the extent that commenters dispute any consideration 
of consumer demand, the D.C. Circuit put that question to rest decades 
ago.
    Related to the agency's consideration of consumer demand, a number 
of commenters took issue with the agencies' estimates of the cost of 
meeting higher fuel economy standards, arguing essentially that the 
analysis was deliberately constructed to inflate costs and minimize 
consumer willingness to pay for fuel economy improvements in order to 
arrive at a policy conclusion that higher fuel economy standards would 
not be economically practicable. NHTSA does not believe that commenters 
mean to argue with the agency's legal interpretation (i.e., the 
consideration of cost as an aspect of economic practicability), but 
rather with the agencies' analytical findings which inform that 
consideration. Comments on those analytical findings, and the agencies' 
responses and changes to the analysis in response to those comments, 
are discussed in Sections VI and VII above. Consumer willingness to pay 
for additional fuel economy in their new vehicles, in particular, is 
represented throughout the final rule analysis as 2.5 years--that is, 
that consumers value, and manufacturers will voluntarily add, fuel 
economy-improving technology that pays for itself in fuel savings 
within 2.5 years.
    More generally, NHTSA believes that the cost of meeting CAFE 
standards is inherently relevant to assessing whether those standards 
are ``within the financial capability of the industry but not so 
stringent as to lead to adverse economic consequences,'' for two 
primary reasons. First, vehicle manufacturers tend to have relatively 
fixed budgets for R&D and production, which are tied to overall 
revenues. If more of those budgets are spent on improving fuel economy, 
less of those budgets are available to spend on other vehicle 
characteristics (such as advanced safety features, or better 
performance or utility) that might improve sales. Offering less of 
those other vehicle characteristics in a market where many consumers 
are not particularly focused on fuel economy could lead to adverse 
economic consequences for those manufacturers. Manufacturers cannot 
simply increase budgets or turn limited resources toward supplying more 
of vehicle characteristics that do not motivate most sales. To the 
extent that more stringent standards drive manufacturing costs higher 
and those costs are passed forward to consumers in the form of price 
increases, those price increases can affect vehicle sales to some 
extent. NHTSA understands that some commenters disagree that higher 
manufacturing costs are necessarily passed forward to consumers in the 
way that the agencies have modeled them being passed forward, but the 
agencies do not have adequate information on which to base a different 
approach. Commenters disagreeing with this approach generally object on 
two fronts: First, because they believe that automakers cross-subsidize 
cost increases by raising the prices of certain models rather than all 
models, and second, because they believe that automakers could absorb 
regulatory costs and reduce profits. The agencies do not have enough 
information to model either of those issues in a meaningful way. Some 
amount of cross-subsidization no doubt occurs, but automakers closely 
hold pricing strategy information. The agencies do not attempt to model 
automakers voluntarily reducing profits in response to standards, again 
in part because the agencies do not have sufficient information, but 
also because these companies are publicly-traded and taking losses is 
not a long-term solution for companies whose success is measured by 
profitability. NHTSA believes that the analytical approach used today 
is reasonable given the information available to the agencies. While 
today's analysis does not show large sales effects due to price 
increases, and even accounting for fuel economy differences in this 
final rule still does not show large sales effects, it seems reasonable 
to call negative sales effects ``adverse economic consequences.''
    Also related to consumer demand, NHTSA has previously considered 
manufacturer ``shortfalls'' as an aspect of economic 
practicability.\2630\ The CAFE standards are corporate average 
standards, by definition, giving manufacturers the flexibility to 
decide how to distribute fuel economy-improving technologies throughout 
their fleet. In other words, no given vehicle need, itself, meet a 
standard or even its ``target'' on the target curve, as long as the 
fleet as a whole meets the standard. However, CAFE compliance is 
measured on a sales-weighted basis, so if a manufacturer ultimately 
sells more vehicles that perform poorly relative to their targets than 
it sells vehicles that beat their targets, the manufacturer may fall 
short of its compliance obligation despite having applied fuel economy-
improving technologies in amounts that the manufacturer originally 
anticipated would result in compliance. Recent compliance trends have 
illustrated this phenomenon, as discussed in Section IV above. When 
fuel is relatively inexpensive, Americans tend to be less interested in 
saving money on fuel, and thus less interested in fuel economy as 
compared to other vehicle attributes. Compliance shortfalls represent 
this consumer decision-making playing out in the market, and can thus 
be evidence of economic impracticability if sufficiently 
widespread.\2631\
---------------------------------------------------------------------------

    \2630\ See 77 FR at 63040-43 (Oct. 15, 2012).
    \2631\ See, e.g., Alliance comments (Full Comment Set) at 25-29, 
describing automaker shortfalls in terms of fleet fuel economy 
increases required by augural and prior standards.
---------------------------------------------------------------------------

    As with the above-discussed aspects of economic practicability, 
commenters who objected to NHTSA's consideration

[[Page 25136]]

of employment impacts disagreed less with the principle of considering 
employment impacts, and more with how NHTSA discussed employment 
impacts in the proposal's justification given the NPRM's findings on 
employment. Namely, the NPRM included a simplistic analysis that 
converted reduced technology costs under the preferred alternative 
relative to the augural standards into ``job years'' metric and 
estimated U.S. auto sector labor would be slightly reduced under the 
proposal as compared to under the augural standards (reflecting those 
reduced technology costs). Although new vehicle sales increased 
slightly under the NPRM's preferred alternative, this was offset 
because ``manufacturing, integrating, and selling less technology means 
using less labor to do so.'' \2632\ However, NHTSA expressed concern in 
the proposal justification section that ``there could be potential for 
. . . loss of U.S. jobs . . . under nearly all if not all of the 
regulatory alternatives considered . . . .'' \2633\ A number of 
commenters argued that if more stringent standards led to higher 
employment, as the NPRM (and also outside analyses) appeared to show, 
there was no way that less stringent standards could be more 
economically practicable.
---------------------------------------------------------------------------

    \2632\ 83 FR at 43436 (Aug. 24, 2018).
    \2633\ Id. at 43216.
---------------------------------------------------------------------------

    As in the NPRM, NHTSA recognizes that the employment analysis for 
this final rule does not capture certain potential effects that may be 
important. NHTSA explained in the NPRM that the NPRM's employment 
analysis did not account for the risks that vehicle sales may be facing 
a bubble situation, or that manufacturers facing higher production 
costs might choose to move production overseas.\2634\ This topic is 
discussed at greater length in Section VIII.B.4 below.
---------------------------------------------------------------------------

    \2634\ Id. at 43224-25.
---------------------------------------------------------------------------

    Commenters addressing NHTSA's consideration of safety as an aspect 
of economic practicability argued generally that EPCA did not call for 
discussion of safety concerns, and that it was unreasonable to assume 
that requiring higher levels of fuel economy might preclude investment 
in further vehicle safety improvements. NHTSA has already explained 
above that the long-standing definition of ``economic practicability'' 
lists example ``adverse economic consequences'' in a nonrestrictive 
clause format, meaning that other things besides employment and 
consumer choice impacts could cause economic consequences and be 
relevant to economic practicability. NHTSA believes that it is 
reasonable and appropriate to consider some aspects of safety as part 
of its consideration of economic practicability, because NHTSA 
continues to believe that vehicle manufacturers have finite budgets for 
R&D and production that may be spent on fuel economy improvements when 
they may otherwise be spent on safety improvements, among other things 
that consumer value. Some commenters said that that was not a 
reasonable assumption, but it is supported by statements from vehicle 
manufacturers,\2635\ and NHTSA does not have a reason to disbelieve 
that companies have limited budgets. Moreover, case law does not object 
to consideration of safety as an aspect of economic 
practicability.\2636\ With regard to IPI's comment about monetization 
of climate and pollution-related deaths and other welfare impacts, the 
social cost of carbon and criteria pollutant damages estimates are 
intended to account for these impacts, and are considered both as part 
of the cost-benefit analysis and under the environmental implications 
aspect of the need of the U.S. to conserve energy. Given that the 
decision about what standards are ``maximum feasible'' is made by 
considering all of the factors, it is therefore less relevant under 
which factor a given issue is considered, so long as it is 
appropriately considered. To the extent that IPI disagrees with those 
estimated valuations, Section VI discusses comments on those topics and 
the agencies' responses.
---------------------------------------------------------------------------

    \2635\ See, e.g., Toyota comments at 6, NHTSA-2018-0067-12098 
(``There are now more realistic limits placed on the number of 
engines and transmissions in a powertrain portfolio which better 
recognizes manufacturers must manage limited engineering resources 
and control supplier, production, and service costs.'').
    \2636\ Competitive Enterprise Institute v. NHTSA, 901 F.2d 107, 
120, n. 11 (``Petitioners have never clearly identified the precise 
statutory basis on which safety concerns should be factored into the 
CAFE scheme, although they alluded to occupant safety as part of the 
`economic practicability' criterion in their MY 1989 petition to 
NHTSA and at oral argument. We do not find this failure fatal, 
however, because NHTSA has always examined the safety consequences 
of the CAFE standards in its overall consideration of relevant 
factors since its earliest rulemaking under the CAFE program, 
(citations omitted). Moreover, NHTSA itself believes Congress was 
cognizant of safety issues when it enacted the CAFE program. As 
evidence, NHTSA discusses a congressional report that dealt with the 
safety consequences of a downsized fleet of cars which had been 
considered by Congress during its enactment of the CAFE program.'').
---------------------------------------------------------------------------

    Based on the above, NHTSA continues to believe that its 
interpretation of economic practicability is reasonable. Section 
VIII.B.4 will discuss how NHTSA has considered and balanced economic 
practicability for this final rule, and also respond to comments that 
addressed the NPRM's application of economic practicability to the 
information before the agency at that time.
(3) The Effect of Other Motor Vehicle Standards of the Government on 
Fuel Economy
    ``The effect of other motor vehicle standards of the Government on 
fuel economy'' involves analysis of the effects of compliance with 
emission, safety, noise, or damageability standards on fuel economy 
capability and thus on average fuel economy. In many past CAFE 
rulemakings, NHTSA has said that it considers the adverse effects of 
other motor vehicle standards on fuel economy. It said so because, from 
the CAFE program's earliest years \2637\ until recently, the effects of 
such compliance on fuel economy capability over the history of the CAFE 
program have been negative ones. For example, safety standards that 
have the effect of increasing vehicle weight thereby lower fuel economy 
capability, thus decreasing the level of average fuel economy that 
NHTSA can determine to be feasible. In the analyses for both the NPRM 
and this final rule, NHTSA has considered the additional weight that it 
estimates would be added in response to new safety standards during the 
rulemaking timeframe.\2638\ NHTSA has also accounted for EPA's ``Tier 
3'' standards for criteria pollutants in its estimates of technology 
effectiveness in both the NPRM and final rule analyses.\2639\
---------------------------------------------------------------------------

    \2637\ 42 FR 63184, 63188 (Dec. 15, 1977). See also 42 FR 33534, 
33537 (Jun. 30, 1977).
    \2638\ PRIA, Chapter 5; FRIA, Section 5.
    \2639\ PRIA, Chapter 6; FRIA, Section 6.
---------------------------------------------------------------------------

    NHTSA discussed in the NPRM whether to consider EPA's 
CO2 standards as an ``other motor vehicle standard of the 
Government'' among the other regulations typically considered, and if 
so, how. NHTSA explained that in the 2012 final rule establishing CAFE 
standards for MYs 2017-2021, NHTSA recognized that ``To the extent the 
GHG standards result in increases in fuel economy, they would do so 
almost exclusively as a result of inducing manufacturers to install the 
same types of technologies used by manufacturers in complying with the 
CAFE standards.'' \2640\ NHTSA concluded in 2012 that ``no further 
action was needed'' because ``the agency had already considered EPA's 
[action] and the harmonization benefits of the National Program in 
developing its own [action].'' \2641\
---------------------------------------------------------------------------

    \2640\ 77 FR 62624, 62669 (Oct. 15, 2012).
    \2641\ Id.

---------------------------------------------------------------------------

[[Page 25137]]

    In the NPRM, NHTSA considered the issue afresh, and determined that 
it was clear based on a purely textual analysis of the statutory 
language that EPA's CO2 standards applicable to light-duty 
vehicles are literally ``other motor vehicle standards of the 
Government,'' in that they are standards set by a Federal agency that 
apply to motor vehicles. Basic chemistry makes fuel economy and 
tailpipe CO2 emissions two sides of the same coin, as 
discussed at length above, and when two agencies functionally regulate 
both (because when regulating fuel economy, CO2 emissions 
are necessarily also regulated, and vice versa), it would be absurd not 
to link the standards.\2642\ The global warming potential of 
N2O, CH4, and HFC emissions are not closely 
linked with fuel economy, but neither do they affect fuel economy 
capabilities. Simply concluding that EPA's CO2 standards 
were ``other motor vehicle standards of the Government,'' however, did 
not answer how should NHTSA should consider them.
---------------------------------------------------------------------------

    \2642\ In fact, EPA includes tailpipe CH4, CO, and 
CO2 in the measurement of tailpipe CO2 for 
CO2 compliance using a carbon balance equation so that 
the measurement of tailpipe CO2 exactly aligns with the 
measurement of fuel economy for the CAFE compliance.
---------------------------------------------------------------------------

    NHTSA acknowledged in the NPRM that some stakeholders had 
previously suggested that NHTSA should implement this statutory factor 
by letting EPA decide what CO2 standards are appropriate and 
reasonable under the CAA and then simply setting CAFE standards with 
reference to CO2 stringency. NHTSA disagreed that such an 
approach would be a reasonable interpretation of EPCA, explaining that 
while EPA and NHTSA consider some similar factors under the CAA and 
EPCA/EISA, respectively, they are not identical, and standards that are 
appropriate under the CAA may not be ``maximum feasible'' under EPCA/
EISA, and vice versa. Moreover, NHTSA explained, considering EPCA's 
language in the context in which it was written, it seemed unreasonable 
to conclude that Congress intended EPA to dictate CAFE stringency. In 
fact, Congress clearly separated NHTSA's and EPA's responsibilities for 
CAFE under EPCA by giving NHTSA authority to set standards and EPA 
authority to measure and calculate fuel economy. If Congress had wanted 
EPA to set CAFE standards, it could have given that authority to EPA in 
EPCA or at any point since Congress amended EPCA.\2643\
---------------------------------------------------------------------------

    \2643\ The NPRM noted, for instance, that EISA was passed after 
the Massachusetts v. EPA decision by the Supreme Court. If Congress 
had wanted to amend EPCA in light of that decision, it would have 
done so at that time, but did not.
---------------------------------------------------------------------------

    NHTSA explained that NHTSA and EPA are obligated by Congress to 
exercise their own independent judgment in fulfilling their statutory 
missions, even though both agencies' regulations affect both fuel 
economy and CO2 emissions. Because of this relationship, it 
is incumbent on both agencies to coordinate and look to one another's 
actions to avoid unreasonably burdening industry through inconsistent 
regulations,\2644\ but both agencies' programs must stand on their own 
merits. As with other recent CAFE and CO2 rulemakings, NHTSA 
explained that the agencies were continuing do all of these things in 
the proposal.
---------------------------------------------------------------------------

    \2644\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007) (``[T]here 
is no reason to think the two agencies cannot both administer their 
obligations and yet avoid inconsistency.'').
---------------------------------------------------------------------------

    With regard to standards issued by the State of California, the 
NPRM explained that State tailpipe standards (whether for 
CO2 or for other pollutants) do not qualify as ``other motor 
vehicle standards of the Government'' under 49 U.S.C. 32902(f), and 
that therefore, NHTSA would not consider them as such in proposing 
maximum feasible average fuel economy standards. NHTSA explained that 
States may not adopt or enforce standards related to fuel economy 
standards, which are preempted under EPCA, regardless of whether EPA 
granted any waivers under the Clean Air Act (CAA).
    NHTSA and EPA agreed in the NPRM that State tailpipe CO2 
emissions standards do not become Federal standards and qualify as 
``other motor vehicle standards of the Government,'' when subject to a 
CAA preemption waiver. NHTSA stated that EPCA's legislative history 
supports that position, as follows:
    EPCA, as initially passed in 1975, mandated average fuel economy 
standards for passenger cars beginning with model year 1978. The law 
required the Secretary of Transportation to establish, through 
regulation, maximum feasible fuel economy standards \2645\ for model 
years 1981 through 1984 with the intent to provide steady increases to 
achieve the standard established for 1985 and thereafter authorized the 
Secretary to adjust that standard.
---------------------------------------------------------------------------

    \2645\ As is the case today, EPCA required the Secretary to 
determine ``maximum feasible average fuel economy'' after 
considering technological feasibility, economic practicability, the 
effect of other Federal motor vehicle standards on fuel economy, and 
the need of the Nation to conserve energy. 15 U.S.C. 2002(e) 
(recodified July 5, 1994).
---------------------------------------------------------------------------

    For the statutorily-established standards for model years 1978-
1980, EPCA provided each manufacturer with the right to petition for 
changes in the standards applicable to that manufacturer. A petitioning 
manufacturer had the burden of demonstrating a ``Federal fuel economy 
standards reduction'' was likely to exist for that manufacturer in one 
or more of those model years and that it had made reasonable technology 
choices. ``Federal standards,'' for that limited purpose, included not 
only safety standards, noise emission standards, property loss 
reduction standards, and emission standards issued under various 
Federal statutes, but also ``emissions standards applicable by reason 
of section 209(b) of [the CAA].'' \2646\ (Emphasis added). Critically, 
all definitions, processes, and required findings regarding a Federal 
fuel economy standards reduction were located within a single self-
contained subsection of 15 U.S.C. 2002 that applied only to model years 
1978-1980.\2647\
---------------------------------------------------------------------------

    \2646\ Section 202 of the CAA (42 U.S.C. 7521) requires EPA to 
prescribe air pollutant emission standards for new vehicles; Section 
209 of the CAA (42 U.S.C. 7543) preempts state emissions standards 
but allows California to apply for a waiver of such preemption.
    \2647\ As originally enacted as part of Public Law 94-163, that 
subsection was designated as section 502(d) of the Motor Vehicle 
Information and Cost Savings Act.
---------------------------------------------------------------------------

    In 1994, Congress recodified EPCA. As part of this recodification, 
the CAFE provisions were moved to Title 49 of the United States Code. 
In doing so, unnecessary provisions were deleted. Specifically, the 
recodification eliminated subsection (d). The House report on the 
recodification declared that the subdivision was ``executed,'' and 
described its purpose as ``[p]rovid[ing] for modification of average 
fuel economy standards for model years 1978, 1979, and 1980.'' \2648\ 
It is generally presumed, when Congress includes text in one section 
and not in another, that Congress knew what it was doing and made the 
decision deliberately.
---------------------------------------------------------------------------

    \2648\ H.R. Rep. No. 103-180, at 583-584, tbl. 2A.
---------------------------------------------------------------------------

    NHTSA stated in the NPRM that it had previously considered the 
impact of California's Low Emission Vehicle standards in establishing 
fuel economy standards and occasionally has done so under the ``other 
standards'' sections.\2649\ During the 2012 rulemaking, NHTSA sought 
comment on the appropriateness of considering California's tailpipe 
CO2 emission standards in this section and concluded that 
doing so was unnecessary.\2650\ In light of the legislative history 
discussed above, however, NHTSA stated in the NPRM that such 
consideration would be inappropriate, and confirms that consideration 
of California's LEV

[[Page 25138]]

standards as among the ``other standards of the Government'' was 
inappropriate.
---------------------------------------------------------------------------

    \2649\ See, e.g., 68 FR 16896, 71 FR 17643.
    \2650\ See 77 FR 62669.
---------------------------------------------------------------------------

    Commenters addressing criteria pollutant standards generally 
supported NHTSA's approach in the NPRM. AFPM commented that NHTSA 
``must consider the effect on fuel economy of EPA's Title II standards, 
including the use of catalytic converters, PM traps and other 
technologies that address emissions and have a fuel economy impact.'' 
\2651\ Ford also stated that previous analyses ``did not assess the 
impact of the criteria pollutant emission standards that were adopted 
subsequent to the [2012 final rule],'' which Ford said ``increased the 
challenge of meeting the fuel economy and GHG targets and should be 
taken into consideration.'' \2652\ Ford stated that the NPRM 
appropriately included ``updat[ed] core engine maps using correct, 
regular-grade octane test fuel,'' and that it accounts for ``ultra-low 
2025 MY Tier 3 and LEVIII emissions standards [which] will require 
aggressive cold start strategies [that] consume additional fuel at 
start-up in order to rapidly heat the catalyst to an effective 
operating temperature, which degrades CO2 and fuel economy 
performance on the FTP test [and] was not considered previously. . . 
.'' \2653\
---------------------------------------------------------------------------

    \2651\ AFPM, NHTSA-2018-0067-12078, at 52.
    \2652\ Ford, NHTSA-2018-0067-11928, at 7.
    \2653\ Id.
---------------------------------------------------------------------------

    Regarding how NHTSA should consider EPA's CO2 standards 
as ``other motor vehicle standards of the Government,'' ACEEE suggested 
amongst its comments that, in considering EPA's CO2 
standards, ``NHTSA should not weaken its program . . . to compensate 
for . . . inevitable, modest differences'' between EPA's and NHTSA's 
programs.\2654\ ``Indeed, to the extent that differences in the 
requirements of the two programs remain, it is clear that the more 
stringent requirement in any given respect should govern the 
obligations of the manufacturer.'' \2655\ AFPM commented similarly that 
``Although NHTSA must consider the effect of other governmental 
regulations, Congress intended that NHTSA would have exclusive 
authority over a single set of national fuel economy standards.'' 
\2656\ Mr. Dotson expressed his belief that ``Congress was cognizant of 
the relationship between EPCA and the Clean Air Act when crafting 
EISA'' and cited and discussed various types of legislative history for 
the proposition that EISA had not limited EPA's CAA authority, and that 
various legislative efforts to do so had been put forth in some fashion 
and had failed.\2657\
---------------------------------------------------------------------------

    \2654\ ACEEE, NHTSA-2018-0067-12122, joint NGO comment to 
Alliance/Global petition for flexibilities, at 3.
    \2655\ Id.
    \2656\ AFPM, NHTSA-2018-0067-12078, at 52.
    \2657\ Dotson, EPA-HQ-OAR-2018-0283-4132, Appendix A, at A2-A23. 
NHTSA disagrees with the persuasiveness of the legislative history 
cited by Mr. Dotson, which includes floor debates, colloquies, and 
other similar information that does not reflect the agreement of the 
Congress as a whole. NHTSA looks to the language Congress actually 
passed and the President signed into law.
---------------------------------------------------------------------------

    NHTSA agrees that while it is appropriate for NHTSA to coordinate 
with and look to EPA's actions to avoid unreasonably burdening industry 
through inconsistent regulations, it would not be appropriate for NHTSA 
to reduce stringency below levels it believes to be maximum feasible 
solely for purposes of accommodating differences between programmatic 
flexibilities. The 2012 final rule clearly stated that while the 
agencies had made efforts to align their standards, programmatic 
differences existed, and how manufacturers chose to rely on compliance 
flexibilities could affect the relative stringency of NHTSA's and EPA's 
standards:

    We note, however, that the alignment is based on the assumption 
that manufacturers implement the same level of direct A/C system 
improvements as EPA currently forecasts for those model years, and 
on the assumption of PHEV, EV, and FCV penetration at specific 
levels. If a manufacturer implements a higher level of direct A/C 
improvement technology (although EPA predicts 100% of manufacturers 
will use substitute refrigerants by MY 2021, and the GHG standards 
assume this rate of substitution) and/or a higher penetration of 
PHEVs, EVs and FCVs, then NHTSA's standards would effectively be 
more stringent than EPA's. Conversely, if a manufacturer implements 
a lower level of direct A/C improvement technology and/or a lower 
penetration of PHEVs, EVs and FCVs, then EPA's standards would 
effectively be more stringent than NHTSA's. Several manufacturers 
commented on this point and suggested that this meant the standards 
were not aligned, because NHTSA's standards might be more stringent 
in some years than EPA's. This reflects a misunderstanding of the 
agencies' purpose. The agencies have sought to craft harmonized 
standards such that manufacturers may build a single fleet of 
vehicles to meet both agencies' requirements. That is the case for 
these final standards. Manufacturers will have to plan their 
compliance strategies considering both the NHTSA standards and the 
EPA standards and assure that they are in compliance with both, but 
they can still build a single fleet of vehicles to accomplish that 
goal.\2658\
---------------------------------------------------------------------------

    \2658\ 77 FR at 63054-55 (Oct. 15, 2012) (emphasis added).

Thus, NHTSA has been consistent in its position that CO2 
stringency does not and should not, by itself, dictate CAFE stringency. 
That said, consideration of EPA's standards was inherent in development 
of this final rule, given that the same technologies improve fuel 
economy and reduce CO2 emissions, and given that 
CO2 emissions represent the majority of GHGs produced by 
light-duty vehicles, and given that the agencies have conducted the 
analysis for this rulemaking jointly. NHTSA believes that EPA's 
standards have been fully and appropriately considered as part of its 
decision on these final standards. To be clear, NHTSA did not assert in 
the NPRM that EISA constrained EPA's authorities under the CAA and do 
---------------------------------------------------------------------------
not disagree with that aspect of Mr. Dotson's comment.

    Chemours argued that, contrary to the NPRM's statements about 
having considered EPA's GHG standards in developing the proposal, NHTSA 
had not adequately considered EPA's GHG standards because only the no-
action alternative reflected EPA regulation of the non-CO2 
GHGs, and the analysis did not otherwise account for the non-
CO2 GHG standards.\2659\ Chemours stated that those 
standards were ``required, pursuant to CAA section 202(a), to address 
`air pollution' from mobile sources,'' and that ``No assessment was 
done as to whether such standards could be made less stringent in order 
to avoid the various issues identified (e.g., changes in technology 
since the 2012 final rule, costs to consumers, the effect of 
`diminishing returns,' a changed petroleum market and other factors.'' 
\2660\
---------------------------------------------------------------------------

    \2659\ Chemours, NHTSA-2018-0067-12018, at 25.
    \2660\ Id. at 25-26.
---------------------------------------------------------------------------

    NHTSA disagrees that it was necessary for NHTSA to consider EPA's 
standards for non-CO2 GHG emissions any further than as 
discussed above. Regulation of CH4, N2O, and HFCs 
affects fuel economy only indirectly, if at all. As explained above and 
in the 2012 final rule, while NHTSA recognizes that some manufacturers 
may choose paths to compliance with EPA's GHG standards that make their 
compliance with CAFE standards more challenging, the agencies previewed 
this possibility and stated their expectation that manufacturers could 
make these decisions for themselves. To the extent that Chemours is 
asking NHTSA to examine regulatory alternatives reflecting less 
stringent CAFE standards in light of changed conditions since the 2012 
final rule, that is exactly what the NPRM and final rule analyses have 
done.
    A number of commenters disagreed with NHTSA's explanation of how 
State standards need not be considered under this factor. The States 
and Cities

[[Page 25139]]

commenters stated that NHTSA was required to consider State tailpipe 
standards because 49 U.S.C. 32902(f) does not specify that 
``Government'' refers only to ``Federal'' government; because NHTSA had 
not offered compelling evidence or arguments that Congress did not 
intend NHTSA to consider State tailpipe standards; and because ``case 
law . . . states unequivocally that California's standards must be 
considered by NHTSA under this factor [citing Green Mountain Chrysler's 
``federalizing'' language].'' \2661\ The States and Cities commenters 
further argued that NHTSA was trying to argue simultaneously that it 
could not consider State standards under the ``other standards'' factor 
but could consider State standards ``under other EPCA factors, if and 
when it sees fit'' (citing NPRM language that technological feasibility 
and economic practicability are broad factors allowing NHTSA to 
consider elements not specifically designated by Congress).\2662\ The 
States and Cities commenters further argued, citing Fox Television, 
that NHTSA was deviating from past practice without a reasoned 
explanation by not specifically requesting comment in the NPRM on the 
fact that it was not considering California's standards as ``other 
motor vehicle standards of the Government.'' \2663\
---------------------------------------------------------------------------

    \2661\ States and Cities, NHTSA-2018-0067-12018, at 71.
    \2662\ Id. at 71-72.
    \2663\ Id. at 72. Fox Television did not involve a rulemaking, 
and does not require agencies to specifically seek public comment 
when they deviate from past practice. In any event, by articulating 
in the NPRM that NHTSA was not considering California's standards as 
``other motor vehicle standards of the Government'' the public had 
ample opportunity to provide comment on this issue, and commenters 
in fact did so as discussed above.
---------------------------------------------------------------------------

    With regard to NHTSA's analysis of EPCA's original language for MYs 
1978-80 and the 1994 positive law recodification, the States and Cities 
commenters stated that ``NHTSA's statutory and legislative history 
arguments related to standards for model years 1978-1980 lack merit, as 
NHTSA has provided no reasonable argument that Congress meant NHTSA to 
consider a wider range of standards for those years than for others,'' 
and stated that the section in question ``was removed from the statute 
because it expired, not because Congress took issue with NHTSA's 
consideration of California's waiver standards.'' \2664\ Mr. Dotson 
commented similarly that NHTSA could not rely on the 1994 positive law 
codification as basis to conclude that State tailpipe standards 
(whether for GHGs or other emissions) do not qualify as ``other motor 
vehicle standards of the Government,'' because it said ``without 
substantive change. . . .'' \2665\
---------------------------------------------------------------------------

    \2664\ Id. at 71.
    \2665\ Dotson, EPA-HQ-OAR-2018-0283-4132, Appendix A, at A23-
A24.
---------------------------------------------------------------------------

    Additionally, the States and Cities commenters stated that NHTSA 
could not argue that California's emissions standards are not ``other 
motor vehicle standards of the Government'' because they are preempted, 
because NHTSA ``has no authority to decide whether or not California's 
standards are preempted,'' and ``one of the reasons California's 
Advanced Clean Cars program is not preempted by EPCA is because those 
standards are `other motor vehicle standards of the Government' within 
the meaning of EPCA.'' \2666\ Besides this comment, a number of 
comments were submitted regarding NHTSA's statements in the NPRM about 
EPCA's preemption provision and how it applied to California's 
standards. Those comments have been addressed \2667\ as part of the 
separate final rule published on September 27, 2019,\2668\ and will not 
be discussed further as part of this action.
---------------------------------------------------------------------------

    \2666\ States and Cities, NHTSA-2018-0067-12018, at 71.
    \2667\ To the extent that any individual comment was not 
specifically addressed, NHTSA believes that the substance and themes 
of all substantive comments on EPCA preemption were addressed as 
part of that final rule.
    \2668\ 84 FR 51310.
---------------------------------------------------------------------------

    NHTSA affirms that its interpretation set forth in the NPRM that 
``other motor vehicle standards of the Government'' does not apply to 
State emissions standards that relate to fuel economy. NHTSA does not 
understand how 49 U.S.C. 32919 could be given effect if the purpose of 
the ``other motor vehicle standards of the Government'' provision is to 
compel their inclusion in NHTSA's decision-making. NHTSA continues to 
disagree with the two district court cases suggesting that the ``other 
motor vehicle standards of the Government'' provision obviates 49 
U.S.C. 32919, as explained at some length in the ``One National 
Program'' final rule preceding this regulatory action.\2669\ NHTSA 
refers readers to that document for more detail on this topic.
---------------------------------------------------------------------------

    \2669\ See, e.g., 84 FR at 51323 (Sep. 27, 2019).
---------------------------------------------------------------------------

    With regard to State tailpipe standards that do not directly relate 
to fuel economy, NHTSA continues to believe that Congress's original 
direction to consider ``emissions standards applicable by reason of 
section 209(b) of [the CAA]'' applied only to CAFE standards for MYs 
1978-1980, as discussed in the NPRM. NHTSA agrees that the 1994 
positive law recodification was not intended to make substantive 
changes to EPCA; the NPRM explained that, in dropping Section 502(d), 
Congress made clear that that provision was executed, and that 
provision expressly directed NHTSA to consider State standards that had 
been granted preemption waivers under CAA 209(b). In order for States 
even to have their own emissions standards for motor vehicles, 
California must be granted a waiver of preemption under CAA section 
209(b). If Congress had intended for NHTSA to continue to consider 
State tailpipe standards post-MY 1980, the direction to consider 
emissions standards that had been granted Section 209 waivers could 
have been placed elsewhere in the statute. Congress did not do 
so.\2670\ While NHTSA may have considered State tailpipe standards in 
the past, it is not bound to do so, and NHTSA does not believe that it 
is unreasonable to consider those standards under technological 
feasibility or economic practicability if they are to be considered.
---------------------------------------------------------------------------

    \2670\ The negative inference canon is logically and reasonably 
employed here, particularly given that, as a factual matter and as 
discussed further below, considering EPA's Tier 3 standards (which 
are clearly ``other motor vehicle standards of the Government'') 
effectively accounts for the technological implications of 
California's LEVIII standards.
---------------------------------------------------------------------------

    State tailpipe standards primarily affect fuel economy by requiring 
gasoline ICE vehicles to burn additional fuel when the engine first 
starts. For most gasoline engines on the road today, the majority of 
tailpipe NOX, NMOG, and CO emissions occur during ``cold 
start,'' before the three-way catalyst has reached the very high 
temperature (e.g., 900-1000 [deg]F), at which point it is able to 
convert (through oxidation and reduction reactions) those emissions 
into less harmful derivatives. By strictly limiting the amount of those 
emissions, tailpipe smog standards require the catalyst to be brought 
to temperature extremely quickly, so modern vehicles employ cold start 
strategies that intentionally release fuel energy into the engine 
exhaust to heat the catalyst to the relevant temperature as quickly as 
possible. The additional fuel that must be used to heat the catalyst is 
typically referred to as a ``cold-start penalty,'' meaning that 
vehicle's fuel economy (over a test cycle) is reduced because the fuel 
consumed to heat the catalyst did not go toward the goal of moving the 
vehicle forward.\2671\ The Autonomie

[[Page 25140]]

work employed to develop technology effectiveness estimates for this 
final rule does, in fact, account for cold-start penalties.\2672\ The 
Autonomie model documentation discusses the fact that cold-start 
penalties were derived from an EPA database of MY 2016 vehicles, which 
would have met both EPA and California smog standards. Moreover, EPA 
regulations allow manufacturers to employ LEVIII data for Tier 3 
compliance. Based on all of these factors, NHTSA believes that the 
negative fuel economy effects of California's tailpipe standards for 
smog-related emissions are reasonably represented in the analysis for 
the final rule, regardless of whether NHTSA was obligated by law to 
consider them expressly.
---------------------------------------------------------------------------

    \2671\ For more information on this, see, e.g., Pihl, Josh A., 
et al., ``Development of a Cold Start Fuel Penalty Metric for 
Evaluating the Impact of Fuel Composition Changes on SI Engine 
Emissions Control,'' Oak Ridge National Laboratory, 2018. Available 
at https://www.osti.gov/biblio/1462896-development-cold-start-fuel-penalty-metric-evaluating-impact-fuel-composition-changes-si-engine-emissions-control.
    \2672\ See ANL Model Documentation, Section 6.1.5, available in 
Docket No. NHTSA-2018-0067.
---------------------------------------------------------------------------

    Ultimately, it would be illogical for NHTSA to consider legally 
unenforceable standards to be ``other motor vehicle standards of the 
Government.'' That is the case for State standards preempted by EPCA. 
While NHTSA understands that certain commenters disagree with a 
separate final rule that NHTSA issued concerning EPCA preemption, and 
the particular State standards that NHTSA considers preempted by EPCA, 
those issues are outside the scope of this final rule.
(4) The Need of the United States To Conserve Energy
    NHTSA has historically interpreted ``the need of the United States 
to conserve energy'' to mean ``the consumer cost, national balance of 
payments, environmental, and foreign policy implications of our need 
for large quantities of petroleum, especially imported petroleum.'' 
\2673\
---------------------------------------------------------------------------

    \2673\ 42 FR 63184, 63188 (Dec. 15, 1977).
---------------------------------------------------------------------------

(a) Consumer Costs and Fuel Prices:
    NHTSA explained in the NPRM that fuel for vehicles costs money for 
vehicle owners and operators. All else equal--a critical caveat--
consumers benefit from vehicles that need less fuel to perform the same 
amount of work. Future fuel prices are a critical input into the 
economic analysis of potential CAFE standards because they determine 
the value of fuel savings both to new vehicle buyers and to society, 
the amount of fuel economy that the new vehicle market is likely to 
demand in the absence of new standards, and they inform NHTSA about the 
``consumer cost . . . of our need for large quantities of petroleum.'' 
In the proposal, NHTSA's analysis relied on fuel price projections from 
the U.S. Energy Information Administration's (EIA) Annual Energy 
Outlook (AEO) for 2017; in the final rule, on fuel price projections 
derived from the version of NEMS used to produce AEO 2019. Federal 
government agencies generally use EIA's price projections in their 
assessment of future energy-related policies.
    Several commenters stated that consumer costs for fuel were an 
important consideration. ACEEE stated that ``The average U.S. household 
still spent nearly $2,000 on gasoline and motor oil (directly) in 2017, 
making oil savings very relevant for consumers,'' and argued that ``Oil 
price volatility remains a threat to U.S. consumers and businesses--the 
price of crude oil has more than doubled since 2016, belying the 
theoretical suggestion in the notice that conditions for oil price 
shocks no longer exist,'' suggesting that further fuel efficiency 
improvements were necessary to protect consumers.\2674\ NESCAUM 
commented that prior analyses had suggested that consumers would save 
$6,000 on net, after paying more for their vehicles upfront, and that 
the proposal would cost consumers more in fuel.\2675\ Both NESCAUM and 
the States and Cities commenters stated that higher fuel costs would 
disproportionately affect low-income consumers, who spend a higher 
share of their income on fuel costs.\2676\ The Congressional Tri-Caucus 
commented that ``As we see oil prices rising again, it makes no sense 
for DOT to roll back these standards.'' \2677\ The States and Cities 
commenters argued that increased gas expenditures would result ``in 
negative economy-wide effects'' for many years ``given that cars sold 
in the model years for which NHTSA proposes to freeze standards will, 
according to the Agencies, be on the road for decades,'' and stated 
that ``NHTSA's analysis is arbitrary and capricious because it entirely 
fails to consider how the Proposed Rollback would impact consumers and 
the economy as a whole due to increased gasoline expenditures.'' \2678\ 
The States and Cities commenters further argued that NHTSA was 
incorrect in the NPRM when it interpreted ``the relevant question for 
the need of the U.S. to conserve energy is not whether there will be 
any movement in prices but whether that movement will be sudden and 
large,'' \2679\ and cited State Farm to say that NHTSA had ``failed to 
consider an important aspect of the problem'' by ``failing to analyze 
the likely impact of even moderate future increases and volatility in 
fuel prices.'' \2680\
---------------------------------------------------------------------------

    \2674\ ACEEE, NHTSA-2018-0067-12122, at 2.
    \2675\ NESCAUM, NHTSA-2018-0067-11691, at 4.
    \2676\ NESCAUM, NHTSA-2018-0067-11691, at 5; States and Cities, 
NHTSA-2018-0067-11735, at 75, citing Synapse Report.
    \2677\ Congressional Tri-Caucus, NHTSA-2018-0067-1424, at 2.
    \2678\ States and Cities, NHTSA-2018-0067-11735, at 75.
    \2679\ 83 FR at 43214, n. 444.
    \2680\ States and Cities, NHTSA-2018-0067-11735, at 75.
---------------------------------------------------------------------------

    A number of commenters addressed consumer willingness to pay more 
money upfront in order to save money on fuel costs. Many of these 
comments are addressed in Section VI.C as part of the discussion of how 
sales are modeled. More specifically in the context of how NHTSA 
interprets the need of the U.S. to conserve energy, IPI commented that 
NHTSA was incorrect that ``consumers' need to save money is now `less 
urgent' and no longer supports a strong overall need to conserve 
energy. The agencies assert that past rulemakings were overly and 
paternalistically focused on `myopia.' This statement ignores all the 
other pathways through which the 2012 standards benefit consumers' need 
to save money, including by correcting informational asymmetries, 
attention costs, and other informational failures; positional 
externalities; and various other supply-side and demand-side 
explanations for consumers' inability to achieve in an unregulated 
market the level of fuel economy that they desire. These components of 
the national need to conserve energy are discussed at length throughout 
these comments, and were specifically considered by the agencies in the 
2012 rule.'' \2681\
---------------------------------------------------------------------------

    \2681\ IPI, NHTSA-2018-0067-12213, Appendix, at 5-6.
---------------------------------------------------------------------------

    Several commenters disagreed with NHTSA's suggestion in the NPRM 
that increasing U.S. production and exports reduced volatility in the 
oil market. Securing America's Energy Future stated that ``. . . recent 
events are an important validation of public policies that support 
long-term goals like efficiency and fuel diversity. Indeed, in the 
absence of fuel-efficiency standards, global oil price volatility would 
likely render the country even more exposed to oil price shocks than it 
is currently.'' \2682\ Mr. Bordoff, IPI, the States and Cities 
commenters, and UCS all commented that the oil market is global, so 
increasing U.S. production does not prevent price shocks that occur

[[Page 25141]]

due to non-U.S. events or circumstances. Mr. Bordoff stated that ``In a 
globalized oil market, the consequence of a supply disruption anywhere 
is a price increase everywhere--regardless of how much oil the U.S. 
imports.'' \2683\ UCS made similar comments.\2684\ Mr. Bordoff further 
commented that U.S. gasoline prices still follow the fluctuations in 
global crude oil prices regardless of the U.S. oil import/export 
balance,\2685\ and stated that ``Gasoline prices at the pump are 
especially sensitive to changes in the global crude oil price due to 
the relatively low level of fuel taxation [in the U.S.] compared to 
other OECD countries.'' \2686\ Mr. Bordoff stated that gas price spikes 
are still possible due to ongoing geopolitical challenges in major oil 
producing areas, and concluded that ``Continuing with planned fuel 
economy increases through CAFE standards is one effective way to reduce 
the oil intensity of the economy and mitigate the adverse impact of 
future oil price increases on American drivers.'' \2687\ The States and 
Cities commenters cited to and echoed Mr. Bordoff's comments on this 
point.\2688\ CARB commented that the proposal had relied on AEO 2017, 
which reflected fuel prices that still assumed the augural standards 
remained in place, but that AEO 2018 assumes ``no new fuel efficiency 
standard'' and held fuel economy flat after 2021, and showed fuel 
prices would be higher.\2689\
---------------------------------------------------------------------------

    \2682\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 7.
    \2683\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 6.
    \2684\ UCS, NHTSA-2018-0067-12039, at 7.
    \2685\ IPI cited and echoed these comments. IPI, NHTSA_2018-
0067-12213, Appendix, at 3.
    \2686\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 7.
    \2687\ Id. at 10-12.
    \2688\ States and Cities, NHTSA-2018-0067-11735, at 74-75.
    \2689\ CARB, NHTSA-2018-0067-11783, at 318.
---------------------------------------------------------------------------

    Mr. Bordoff also commented that the future of shale oil in the U.S. 
was uncertain, and therefore increased U.S. oil production was not a 
basis on which to assume future global price stability.\2690\ Mr. 
Bordoff argued that ``Although shale oil is more responsive to price 
changes than conventional supply, it cannot serve as a swing supplier 
to stabilize oil markets in the way true spare capacity (held by Saudi 
Arabia) can. It takes at least 6-12 months for U.S. shale to respond to 
price changes.'' \2691\ Bordoff continued, stating that ``For example, 
although shale oil is more responsive to oil prices, oil prices still 
plunged below $30 per barrel at the start of 2016 and soared to $80 per 
barrel earlier this year. Shale oil could not swing quickly enough to 
stabilize markets. This role fell to OPEC instead in both cases, first 
to put a floor under prices by cutting supply and, more recently, to 
provide relief by ramping up production.'' \2692\ Bordoff further 
commented that political or popular pressures due to environmental 
concerns may significantly increase the cost and/or difficulty of 
expanding shale infrastructure,\2693\ and that even disregarding 
uncertainty in supply, ongoing uncertainty in demand (both U.S. and 
abroad) also contributed to global price uncertainty.\2694\
---------------------------------------------------------------------------

    \2690\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 3.
    \2691\ Id., at 7.
    \2692\ Id., at 7-8.
    \2693\ Id., at 9-10.
    \2694\ Id., at 3.
---------------------------------------------------------------------------

    NHTSA agrees with commenters that consumer costs for fuel are 
relevant to the need of the U.S. to conserve energy. NHTSA also agrees 
that future fuel prices are uncertain, and that shale oil development 
in the U.S. is (1) still proceeding and subject to uncertainty, (2) 
very different from traditional sources like Saudi Arabia, and (3) not 
enough, by itself, to preclude any possibility of major swings in 
future global oil prices. That said, NHTSA continues to believe that 
U.S. shale development may reduce the negative price effects of global 
price swings due to events and situations outside of our borders. Shale 
represents a large, new, relatively-geopolitically-stable oil supply 
source, and traditional oil producers appear to understand that 
stabilizing prices below the price at which shale production starts to 
ramp up faster helps those traditional producers take market advantage 
of their lower cost of production.\2695\ The net effect of this, for 
American drivers, should be greater fuel price stability, at least at 
the upper end of fuel prices. NHTSA also continues to believe that, for 
purposes of considering consumer cost of fuel as part of the need of 
the U.S. to conserve energy, the fact that Americans' gasoline costs 
might be minutely lower under more stringent CAFE standards and 
minutely higher under comparatively less stringent CAFE standards is 
not dispositive by itself. There is some tolerance in the market for 
some amount of fluctuation in fuel prices, as evidenced by the 
discussion in Section VI. Slow increases in fuel prices are relatively 
easy for households to absorb; sharp increases are more difficult.
---------------------------------------------------------------------------

    \2695\ Since 1995, EIA data indicates that OPEC production 
roughly stabilized in late 2016 and has either remained steady or 
fallen since then. See https://www.eia.gov/opendata/qb.php?category=1039874&sdid=STEO.PAPR_OPEC.M. See also Ilya 
Arkhipov, Will Kennedy, Olga Tanas, and Grant Smith, ``Putin Dumps 
MBS to Start a War on America's Shale Oil Industry,'' March 7, 2020, 
Bloomberg News, available at https://www.bloomberg.com/news/articles/2020-03-07/putin-dumps-mbs-to-start-a-war-on-america-s-shale-oil-industry (describing the collapse of the OPEC+ coalition); 
EIA, ``This Week in Petroleum--OPEC shift to maintain market share 
will result in global inventory increases and lower prices,'' March 
11, 2020, https://www.eia.gov/petroleum/weekly/; DOE, ``DOE Responds 
to Recent Oil Market Activity,'' March 9, 2020, https://www.energy.gov/articles/doe-responds-recent-oil-market-activity.
---------------------------------------------------------------------------

    Increases in CAFE stringency reduce the effects of all types of 
increases in fuel prices, at least to the extent that people can buy 
new cars and trucks, but as discussed below in Section VIII.B.4, fuel 
costs and per-vehicle costs balance against one another for many 
buyers. With respect to relatively low U.S. gasoline taxes creating 
more pass-through effects of global oil price fluctuations, that would 
be true regardless of stringency. Broadly speaking, while consumer fuel 
costs are an important consideration of the need of the U.S. to 
conserve energy, at this time NHTSA believes, as discussed in Section 
VI, that American consumers generally understand fuel costs and their 
tolerance for fluctuations, and tend to purchase vehicles accordingly. 
Requiring consumers to save more fuel over the longer term by spending 
more money upfront on new vehicle purchases may involve more tradeoffs 
than suggested in prior rulemakings, and this rulemaking seeks to keep 
these possible tradeoffs in mind.
(b) National Balance of Payments:
    As the NPRM explained, the need of the United States to conserve 
energy has historically included consideration of the ``national 
balance of payments'' because of concerns that importing large amounts 
of oil created a significant wealth transfer to oil-exporting countries 
and left the U.S. economically vulnerable.\2696\ As recently as 2009, 
nearly half the U.S. trade deficit was driven by petroleum,\2697\ yet 
this concern has largely laid fallow in more recent CAFE actions, 
arguably in part because other factors besides petroleum consumption 
have since played a bigger role in the U.S. trade deficit. Given

[[Page 25142]]

recent significant increases in U.S. oil production and corresponding 
decreases in oil imports, this concern seems likely to remain fallow 
for the foreseeable future.\2698\ Increasingly, changes in the price of 
fuel have come to represent transfers between domestic consumers of 
fuel and domestic producers of petroleum rather than gains or losses to 
foreign entities. NHTSA explained in the NPRM that some commenters have 
lately raised concerns about potential economic consequences for 
automaker and supplier operations in the U.S. due to disparities 
between CAFE standards at home and their counterpart fuel economy/
efficiency and CO2 standards abroad. NHTSA finds these 
concerns more relevant to technological feasibility and economic 
practicability than to the national balance of payments. Moreover, to 
the extent that an automaker decides to globalize a vehicle platform to 
meet more stringent standards in other countries, that automaker would 
comply with United States' standards and additionally generate 
overcompliance credits that it can save for future years if facing 
compliance concerns, or sell to other automakers. While CAFE standards 
are set at maximum feasible rates, efforts of manufacturers to exceed 
those standards are rewarded not only with additional credits but a 
market advantage in that those consumers who place a large weight on 
fuel savings will find such vehicles that much more attractive.
---------------------------------------------------------------------------

    \2696\ See 42 FR 63184, 63192 (Dec. 15, 1977) (``A major reason 
for this need [to reduce petroleum consumption] is that the 
importation of large quantities of petroleum creates serious balance 
of payments and foreign policy problems. The United States currently 
spends approximately $45 billion annually for imported petroleum. 
But for this large expenditure, the current large U.S. trade deficit 
would be a surplus.'').
    \2697\ See Today in Energy: Recent improvements in petroleum 
trade balance mitigate U.S. trade deficit, U.S. Energy Information 
Administration (July 21, 2014), https://www.eia.gov/todayinenergy/detail.php?id=17191.
    \2698\ For an illustration of recent increases in U.S. 
production, see, e.g., U.S. crude oil and liquid fuels production, 
Short-Term Energy Outlook, U.S. Energy Information Administration 
(June 2018), https://www.eia.gov/outlooks/steo/images/fig13.png. 
While it could be argued that reducing oil consumption frees up more 
domestically-produced oil for exports, and thereby raises U.S. GDP, 
that is neither the focus of the CAFE program nor consistent with 
Congress' original intent in EPCA. EIA's Annual Energy Outlook (AEO) 
series provides midterm forecasts of production, exports, and 
imports of petroleum products, and is available at https://www.eia.gov/outlooks/aeo/.
---------------------------------------------------------------------------

    Several commenters addressed how much oil the U.S. imports, and the 
assumptions about imports in the NPRM analysis. Securing America's 
Energy Future commented that ``Because there are no readily available 
substitutes to oil in the U.S. transportation sector, volatile crude 
oil and petroleum product prices represent an enduring threat to the 
U.S. economy.'' \2699\ ACEEE commented that overall U.S. oil imports 
are higher now than they were in 1975, and nearly as high as they were 
in 2012, and also stated that compared to a small overall trade surplus 
in 1975, ``the U.S. now runs a large overall trade deficit.'' \2700\ 
The States and Cities commenters made a similar point, arguing that the 
U.S. still imports large amounts of petroleum; that imports made up 
about 25 percent of total U.S. oil consumption in 2017; and that EIA 
indicates that ``imports as a share of oil consumption in the United 
States are only about 10% lower today as compared to 1975, and we are 
producing the same amount of crude oil domestically today as we were in 
1970.'' \2701\ IPI stated that EIA analysis shows that the ``U.S. will 
continue to import crude oil through 2050 and `remains a net importer 
of petroleum and other liquids on an energy basis.' '' \2702\ CARB 
disagreed that the U.S. was projected to become a net petroleum 
exporter, and stated that even if it were, the rollback would have 
negative effects on the U.S., because (1) it ignores short-run damages 
caused by increased oil consumption and imports; (2) relies on 
projections of net imports of oil which also do not take account of the 
effects of the proposed rule; and (3) is not supported by the 
evidence.\2703\
---------------------------------------------------------------------------

    \2699\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 6.
    \2700\ ACEEE, NHTSA-2018-0067-12122, at 2.
    \2701\ States and Cities, NHTSA-2018-0067-11735, at 76.
    \2702\ IPI, NHTSA-2018-0067-12213, Appendix, at 3.
    \2703\ CARB, NHTSA-2018-0067-11873, at 317.
---------------------------------------------------------------------------

    Regarding assumptions about oil imports in the NPRM analysis, the 
States and Cities commented that in 2016 the agencies had assumed that 
``90% of fuel savings from existing standards would lead directly to a 
reduction in imported oil,'' and argued that the NPRM analysis had 
ignored that previous assumption and ``la[id] great emphasis on the 
fact that `oil imports have declined while exports have increased' 
since 2005.'' \2704\ IPI argued that the NPRM analysis was internally 
inconsistent, assuming in NHTSA's need of the nation discussion that 
``additional gasoline consumption will be entirely domestic,'' while 
``upstream emissions calculations assume that 95% of increased 
consumption will either be from foreign refining or from foreign crude 
imports,'' and suggested that this inconsistency was purposeful to make 
the NPRM analysis look more favorable to the proposal.\2705\ ACEEE 
commented that ``The EIA AEO side cases suggest that reduced oil demand 
will primarily reduce oil imports, thus improving the overall balance 
of trade regardless of the narrow balance of trade in petroleum.'' 
\2706\
---------------------------------------------------------------------------

    \2704\ States and Cities, NHTSA-2018-0067-11735, at 75.
    \2705\ IPI, NHTSA-2018-0067-12213, Appendix, at 3-4.
    \2706\ ACEEE, NHTSA-2018-0067-12122, at 2.
---------------------------------------------------------------------------

    Regarding the effects on the U.S. economy of increasing U.S. oil 
production, Mr. Morris agreed with the NPRM's suggestion that U.S. 
self-sufficiency in petroleum supply meant that higher consumer 
payments for fuel under less-stringent CAFE standards would be 
transfers within the U.S. economy, and stated that ``[a]t that point, 
the initial purpose of EPCA is entirely obviated.'' \2707\ The States 
and Cities commenters, in contrast, argued that focusing on this effect 
meant that NHTSA essentially claims that increasing revenues of oil 
companies--which report annual profits in the billions--is an even 
trade-off for adding cost pressures and oil-price shock exposure to 
American households.'' \2708\ The States and Cities commenters stated 
that ``. . .this assertion ignores the negative economic impacts that 
would result from increasing the cost burden on oil consumers,'' and 
was ``. . .so implausible that it could not be ascribed to a difference 
of view or the product of agency expertise,' citing State Farm, 463 
U.S. at 43.\2709\
---------------------------------------------------------------------------

    \2707\ Morris (GWU RSC), EPA-HQ-OAR-2018-0283-4028, at 15.
    \2708\ States and Cities, NHTSA-2018-0067-11735, at 76.
    \2709\ Id.
---------------------------------------------------------------------------

    As discussed above, NHTSA agrees that oil is a global commodity. 
Living in a globalized economy necessarily means that supply 
disruptions (and thus, price effects) can come from a great variety of 
sources--this was why the CAFE program was created, in recognition of 
this risk. Increasing U.S. energy independence reduces this risk. There 
are two ways to increase petroleum independence: To use less petroleum, 
and to produce more of our own petroleum and use less petroleum 
purchased from abroad. Both approaches work, and both are being 
followed today.
    NHTSA also agrees that the Draft TAR text describes the analytical 
assumption that for every gallon of fuel not consumed as a result of 
more stringent standards, imported crude would be reduced by 0.9 
gallons. The Draft TAR stated that this assumption was based on 
``changes in U.S. crude oil imports and net petroleum products in the 
AEO 2015 Reference Case in comparison [sic] the Low (i.e., Economic 
Growth) Demand Case,'' and also on a 2013 paper by Paul Leiby which 
``suggests that `Given a particular reduction in oil demand stemming 
from a policy or significant technology change, the fraction of oil use 
savings that shows up as reduced U.S. imports, rather than reduced 
U.S., supply, is actually quite

[[Page 25143]]

close to 90 percent, and probably close to 95 percent.' '' \2710\
---------------------------------------------------------------------------

    \2710\ Draft TAR, 2016, Chapter 10, Endnote 39, p. 10-59.
---------------------------------------------------------------------------

    EIA data clearly states that while the U.S. still relies on oil 
imports, it is producing an increasingly large share of the petroleum 
it consumes.\2711\ In 2018, domestic petroleum production made up 86 
percent of domestic consumption, while imports made up 11 percent. EIA 
data also clearly states that U.S. reliance on petroleum imports peaked 
in 2005 and has declined since then, and that the import-percentage-of-
consumption in 2018 was the lowest it has been since 1957--this despite 
the fact that overall U.S. petroleum consumption has increased 
significantly over that time period as the on-road fleet has grown and 
VMT (both individual and collective) has increased. Of the 11 percent 
of oil consumed that was imported, 43 percent came from Canada, and 16 
percent came from Persian Gulf countries. AEO 2019 states that under 
its Reference case assumptions, which it describes as a ``best 
assessment'' and ``a reasonable baseline case,'' \2712\ the U.S. 
remains projected to become a net exporter of petroleum liquids by 
2020.\2713\ During several weeks in 2019, the U.S. also exported more 
oil than it imported.\2714\
---------------------------------------------------------------------------

    \2711\ EIA, ``Oil: Crude and Petroleum Products Explained, Oil 
Imports and Exports,'' updated May 29, 2019, available at https://www.eia.gov/energyexplained/oil-and-petroleum-products/imports-and-exports.php.
    \2712\ AEO 2019, at 5.
    \2713\ AEO 2019, at 14.
    \2714\ See https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=wttntus2&f=4.
---------------------------------------------------------------------------

    U.S. Census data indicate that the U.S. balance of trade has 
generally grown over time, although it has fluctuated since peaking in 
2006.\2715\ U.S. Census data further indicate that the U.S. petroleum 
balance of trade, in particular, has fluctuated over time, peaking in 
2008 at roughly -$386 million and decreasing to -$50 million in 2018. 
2019 trends demonstrate further decreases. In percentage terms, 
petroleum trade as a percentage of total trade went from roughly 52 
percent in 1992 (the earliest year for which Census appears to have 
data online), to 47 percent in 2008, to less than 6 percent in 2018. In 
terms of national balance of payments, this is fairly clear evidence 
that petroleum has decreased rapidly as part of the problem. Part of 
this is due to improvements in fleet fuel economy over time, and part 
is due to increases in U.S. production, particularly in the last 
several years.
---------------------------------------------------------------------------

    \2715\ ``U.S. Trade in Goods and Services--Balance of Payments 
(BOP) Basis,'' June 6, 2019, available at https://www.census.gov/foreign-trade/statistics/historical/gands.pdf.
---------------------------------------------------------------------------

    NHTSA notes also that the Draft TAR previewed the possibility of 
this outcome, discussing the ``Shale Oil Revolution'' and the fact that 
``[t]he recent economics literature on whether oil shocks are the 
threat to economic stability that they once were is mixed.'' \2716\ The 
Draft TAR stated that because of increased U.S. shale oil production, 
``The resulting decrease in foreign imports . . . effectively permits 
U.S. supply to act as a buffer against artificial or other supply 
restrictions (the latter due to conflict or a natural disaster, for 
example).'' \2717\
---------------------------------------------------------------------------

    \2716\ See Draft TAR at 10-30--10-33.
    \2717\ Draft TAR at 10-31.
---------------------------------------------------------------------------

    Since the Draft TAR was issued, U.S. shale production has developed 
even further, and U.S. petroleum imports have continued to fall. If 
more oil is being produced in the U.S., and more of domestic 
consumption comes from domestic production, then even though oil is a 
global commodity and thus subject to price changes resulting from non-
U.S. events, the U.S. economy is inherently better off. When money 
moves around within the U.S. instead of having to leave the U.S., and 
everyone's needs are being met, U.S. citizens are better off when 
things outside the U.S. go wrong--this is what NHTSA means when it 
refers to within-U.S. transfers not being a bad thing as compared to 
greater reliance on imports for consumption needs. To the extent that 
some commenters find within-U.S. transfers problematic because they 
increase U.S. oil company revenues without reducing fuel cost burdens 
on consumers, NHTSA notes that, as discussed above, consumers seem 
willing and able to tolerate some amount of fuel price increases and 
fluctuation risk, as evidenced by their purchasing decisions. Prices 
may still fluctuate, but shortages may foreseeably be reduced.
    The Draft TAR stated that ``despite continuing uncertainty about 
oil market behavior and outcomes and the sensitivity of the U.S. 
economy to oil shocks, it is generally agreed that it is beneficial to 
reduce petroleum fuel consumption from an energy security standpoint. 
It is not just imports alone, but both imports and consumption of 
petroleum from all sources and their role in economic activity, that 
may expose the U.S. to risk from price shocks in the world oil price. 
Reducing fuel consumption reduces the amount of domestic economic 
activity associated with a commodity whose price depends on volatile 
international markets.'' NHTSA continues to agree with these 
statements, but cannot ignore the fact that increased U.S. petroleum 
production represents the other side of the coin. Again, both national 
balance of payments and energy security can be improved on both the 
supply side and the demand side. While today's final rule continues to 
improve on the demand side by setting standards that continue to push 
CAFE levels upward, it also recognizes that supply side improvements 
are playing a role.
(c) Environmental Implications
    The NPRM explained that higher fleet fuel economy can reduce U.S. 
emissions of CO2 as well as various other pollutants by 
reducing the amount of oil that is produced and refined for the U.S. 
vehicle fleet, but can also increase emissions by reducing the cost of 
driving, which can result in increased vehicle miles traveled (i.e., 
the rebound effect). Thus, the net effect of more stringent CAFE 
standards on emissions of each pollutant depends on the relative 
magnitudes of its reduced emissions in fuel refining and distribution 
and increases in its emissions from vehicle use. Fuel savings from CAFE 
standards also necessarily result in lower emissions of CO2, 
the main gas emitted as a result of refining, distribution, and use of 
transportation fuels. Reducing fuel consumption directly reduces 
CO2 emissions because the primary source of transportation-
related CO2 emissions is fuel combustion in internal 
combustion engines.
    NHTSA has considered environmental issues, both within the context 
of EPCA and the context of the National Environmental Policy Act 
(NEPA), in making decisions about the setting of standards since the 
earliest days of the CAFE program. As courts of appeal have noted in 
three decisions stretching over the last 20 years,\2718\ NHTSA defined 
``the need of the United States to conserve energy'' in the late 1970s 
as including, among other things, environmental implications. In 1988, 
NHTSA included climate change concepts in its CAFE notices and prepared 
its first environmental assessment addressing that subject.\2719\ It 
cited concerns about climate change as one of its reasons for limiting 
the extent of its reduction of the CAFE standard for MY 1989 passenger

[[Page 25144]]

cars.\2720\ Since then, NHTSA has considered the effects of reducing 
tailpipe emissions of CO2 in its fuel economy rulemakings 
pursuant to the need of the United States to conserve energy by 
reducing petroleum consumption.
---------------------------------------------------------------------------

    \2718\ CAS, 793 F.2d 1322, 1325 n. 12 (D.C. Cir. 1986); Public 
Citizen, 848 F.2d 256, 262-63 n. 27 (D.C. Cir. 1988) (noting that 
``NHTSA itself has interpreted the factors it must consider in 
setting CAFE standards as including environmental effects''); CBD, 
538 F.3d 1172 (9th Cir. 2007).
    \2719\ 53 FR 33080, 33096 (Aug. 29, 1988).
    \2720\ 53 FR 39275, 39302 (Oct. 6, 1988).
---------------------------------------------------------------------------

    Many commenters addressed the environmental implications of CAFE 
standards and the proposal. ACEEE stated that ``The environmental need 
to save energy is much greater than we realized in 1975,'' and that 
``The notice argues that since improved standards will not by 
themselves solve global warming, they are not necessary. That logic 
would equally suggest that since no one soldier would win a war, we 
should never deploy any troops. No one measure will solve global 
warming. . . . vehicle standards have been the most important.'' \2721\ 
The Harvard environmental law clinic commenters similarly stated that 
``It is illogical to argue against taking a single step on the basis 
that a single step is insufficient to reach one's goal,'' and commented 
that it was unreasonable for the DEIS to state that ``[t]he emission 
reductions necessary to keep global emissions within this carbon budget 
could not be achieved solely with drastic reductions in emissions from 
the U.S. passenger car and light truck fleet.'' \2722\ UCS also argued 
that with respect to the environmental implications of the standards, 
NHTSA's ``argument that the augural standards would only limit global 
warming by 0.02 degrees C in 2100 actually supports the need to 
maintain the standards. That a single U.S. policy could make that much 
difference in limiting global warming is, in fact, quite significant.'' 
\2723\
---------------------------------------------------------------------------

    \2721\ ACEEE, NHTSA-2018-0067-12122, main comments, at 2.
    \2722\ Harvard environmental law clinic, EPA-HQ-OAR-2018-0283-
5486, at 13.
    \2723\ UCS, NHTSA-2018-0067-12039, at 7.
---------------------------------------------------------------------------

    The States and Cities commenters objected to NHTSA's consideration 
in the NPRM of ``whether rapid ongoing increases in CAFE stringency . . 
. can sufficiently address climate change to merit their costs,'' 
arguing that NHTSA had ``completely disregard[ed] environmental costs'' 
contrary to NHTSA's own long-standing approach to CAFE standards.\2724\ 
The States and Cities commenters then framed the CO2 impacts 
of the proposal in tons (specifically, 7,400 million metric tons 
additional CO2 emitted by 2100 as compared to the augural 
standards) and argued that ``the agency effectively ignores its own 
findings, in a sharp and unexplained break with the agency's past 
practice of considering climate impacts,'' citing Fox Television, 556 
U.S. at 515 and the 2010 and 2012 final CAFE rules which discussed 
reduced economic damages from lower climate impacts for those standards 
compared to their baselines.\2725\ IPI also argued that if NHTSA had 
focused on economic damages rather than fractions of degrees Celsius, 
``Once climate damages are fully monetized (as the agencies are 
required to do), it will become apparent that the proposed rollback 
will cause billions of dollars in climate damages. Billions of dollars 
lost to avoidable climate damages is not a small effect, and it very 
clearly is a `destructive and wasteful' effect.'' \2726\ CARB also 
argued that the NPRM had ``wholly fail[ed] to analyze the economic 
effects of the climate change and public health implications of the 
rollback,'' stating that [t]he Agencies assert these are insignificant, 
but that is only because the Agencies' projections of climate change 
are so extreme. An appropriate analysis of a proposal that speeds 
progress toward such a calamitous condition must acknowledge and 
analyze the expected effects.'' \2727\
---------------------------------------------------------------------------

    \2724\ States and Cities, NHTSA-2018-0067-11735, at 73.
    \2725\ Id.
    \2726\ IPI, NHTSA-2018-0067-12213, Appendix, at 4-5.
    \2727\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84.
---------------------------------------------------------------------------

    The States and Cities commenters also argued that NHTSA had not 
explained what the NPRM's definition of ``conservation'' as meaning 
``avoid[ing] wasteful or destructive use'' ``actually means and how it 
changes the agency's past practice of considering environmental 
impacts,'' citing State Farm, 463 U.S. at 43, and Fox Television, 556 
U.S. at 515.\2728\
---------------------------------------------------------------------------

    \2728\ States and Cities, NHTSA-2018-0067-11735, at 73.
---------------------------------------------------------------------------

    Regarding non-climate impacts, IPI commented that the NPRM ``only 
briefly mention[ed] the possible effects on other emissions without 
detailing any of the myriad non-climate public health and welfare 
consequences from pollution associated with petroleum production and 
combustion for motor vehicles.'' \2729\ The States and Cities 
commenters similarly stated that ``NHTSA's evaluation of this factor 
fails to include any analysis of environmental costs related to air 
quality,'' and that the NPRM/DEIS analysis substantially understates 
the actual impacts of the Proposed Rollback on criteria air pollutants 
(such as NOX and PM) and air toxics (such as benzene), 
making it inappropriate to rely upon.'' \2730\
---------------------------------------------------------------------------

    \2729\ IPI, NHTSA-2018-0067-12213, Appendix, at 5.
    \2730\ States and Cities, NHTSA-2018-0067-11735, at 73-74.
---------------------------------------------------------------------------

    NHTSA agrees that the NPRM considered environmental implications of 
the standards somewhat differently from past rulemaking discussions. 
The 2012 final rule, for example, stated that ``[t]he need of the 
nation to conserve energy has long operated to push the balancing 
toward more stringent standards,'' and asked ``[i]n this final rule, 
then, the question raised by this factor, combined with technological 
feasibility, becomes `how stringent can NHTSA set standards before 
economic practicability considerations intercede?' '' \2731\ The NPRM 
discussed the dictionary definition of ``to conserve,'' tentatively 
concluded that thousandths of a degree centigrade in 2100 did not rise 
to the level of being ``wasteful,'' and suggested that ultimately ``we 
no longer view the need of the U.S. to conserve energy as nearly 
infinite.'' \2732\ This is an evolution in interpretation that was 
expressly acknowledged in the NPRM--the words ``we no longer view'' 
clearly indicate acknowledgement of a change in view, i.e., 
interpretation. The NPRM's climate findings were not ignored, they were 
directly examined and discussed at 83 FR 43215-16 in the context of 
NHTSA's interpretation of their significance. The NPRM also discussed 
overall costs and benefits and net benefits in the context of the 
proposed maximum feasible determination, and the cost of carbon 
emissions was included in those values. This final rule similarly 
directly examines and discusses the analytical findings below.
---------------------------------------------------------------------------

    \2731\ 77 FR at 63038-39.
    \2732\ 83 FR at 43215-16.
---------------------------------------------------------------------------

    Moreover, contrary to commenters' statements that NHTSA did not 
acknowledge that its interpretation of the effect of the ``need of the 
U.S. to conserve energy'' factor was changing, or that the balancing of 
factors was different, the NPRM directly stated that:

    NHTSA well recognizes that the decision it proposes to make in 
today's NPRM is different from the one made in the 2012 final rule 
that established standards for MY 2021 and identified `augural' 
standard levels for MYs 2022-2025. Not only do we believe that the 
facts before us have changed, but we believe that those facts have 
changed sufficiently that the balancing of the EPCA factors and the 
other considerations must also change.
The standards that we are proposing today reflect that 
balancing.\2733\
---------------------------------------------------------------------------

    \2733\ 83 FR at 43213. See also 83 FR at 43226 (``In the 2012 
final rule . . . , NHTSA stated that `maximum feasible standards 
would be represented by the mpg levels that we could require of the 
industry before we reach a tipping point that presents risk of 
seriously adverse economic consequences.' [citation omitted] 
However, the context of that rulemaking was meaningfully different 
from the current context. At that time, NHTSA understood the need of 
the U.S. to conserve energy as necessarily pushing the agency toward 
setting stricter and stricter standards. Combining a then-paramount 
need of the U.S. to conserve energy with the perception that 
technological feasibility should no longer be seen as a limiting 
factor, NHTSA then concluded that only significant economic harm 
would be the basis for controlling the pace at which CAFE stringency 
increased over time. Today, the relative importance of the need of 
the U.S. to conserve energy has changed . . . a great deal even 
since the 2012 rulemaking. [T]he need of the U.S. to conserve energy 
may no longer disproportionately outweigh other statutorily-mandated 
considerations such as economic practicability--even when 
considering fuel savings from potentially more-stringent 
standards.'').


[[Page 25145]]


---------------------------------------------------------------------------

NHTSA believes that this is clear acknowledgement of the differences in 
interpretation and the effect of those differences on policy decisions.

[[Page 25146]]

    That said, NHTSA agrees (indeed, has always agreed) with commenters 
that environmental implications exist as a result of changes in CAFE 
stringency. While CO2 emissions will be higher under this 
final rule than if NHTSA had determined that the augural standards were 
maximum feasible, they will be lower than they would have been under 
the proposal--for the ``standard setting'' runs, which are what NHTSA 
looks at for assistance in determining maximum feasible standards, 
NHTSA estimates that, accounting for both tailpile and upstream 
emissions, CO2 emissions in 2050 under the final standards 
will total 1,134 mmt, as compared to 1,149 mmt under the proposed 
standards, or 1,020 mmt under the augural standards. According to the 
Final EIS, which uses a ``real-world'' analysis that incorporates 
models and modeling approaches that permit the agency to take a hard 
look at the potential environmental impacts of the rule,\2734\ NHTSA 
estimates that these amounts of CO2 emissions would lead to 
the following global temperature, sea level, and ocean acidification 
effects: \2735\
---------------------------------------------------------------------------

    \2734\ See Kleppe v. Sierra Club, 427 U.S. 390, 410, n. 21 
(1976).
    \2735\ As discussed in Section 5.3.1 of the FEIS, NHTSA used the 
Global Change Assessment Model (GCAM) Reference scenario to 
represent the No Action Alterantive (Alternative 0) in the modeling 
runs used to create Table I-1. The GCAM Reference Scenario is based 
on a set of assumptions about drivers such as population, 
technology, and socioeconomic changes, in the absence of global 
action to mitigate climate change. It can be described as a 
``business-as-usual'' scenario. NHTSA also conducted an analysis in 
Chapter 8 of the FEIS using the GCAM6.0 scenario, which assumes a 
moderate level of global GHG reductions and corresponds to 
stabilization, by 2100, of total radiative forcing and associated 
CO2 concentrations at roughly 678 ppm. Several commenters 
argued that NHTSA presented climate results in the NPRM/DEIS in the 
context of a ``doomsday scenario,'' in which no actions at all are 
taken to mitigate carbon emissions, but NHTSA emphasizes that this 
is simply the GCAM Reference Scenario, which is a reasonable 
scenario to run given that GCAM is a widely accepted climate model. 
Running the analysis using the GCAM Reference Scenario and GCAM6.0 
Scenario results in different absolute values for the climate 
variables presented in this table and Table 8.6.4-1 of the FEIS, but 
again, this is because of the underlying scenarios, which reflect 
very different levels of global action. When the differences in 
levels of global action are accounted for, the relative impact of 
each action alternative as compared to the No Action Alternative is 
very similar. Thus, regardless of what GCAM scenario the agencies 
consider regarding global action to mitigate climate change, it is 
still meaningful to draw conclusions about the relative impacts of 
the alternatives, because the alternatives are what is within the 
agencies' authority to affect.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 25147]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.733


[[Page 25148]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.734

BILLING CODE 4910-59-C
    NHTSA understands that some commenters view climate change as an 
imminent existential threat. NHTSA does not agree, however, that 
Congress

[[Page 25149]]

intended for NHTSA to set aside other statutory factors in determining 
what CAFE standards would be maximum feasible. Even the maximum 
feasible discussion for the 2012 final rule stated that

    We recognize that higher standards would help the need of the 
nation to conserve more energy . . ., but based on our analysis and 
the evidence presented by the industry, we conclude that higher 
standards would not represent the proper balancing for MYs 2017-2025 
cars and trucks. [footnote omitted] We conclude that the correct 
balancing recognizes economic practicability concerns as discussed 
above, and sets standards at the levels that the agency is 
promulgating in this final rule for MYs 2017-2021 and presenting for 
MYs 2022-2025.\2736\
---------------------------------------------------------------------------

    \2736\ 77 FR at 63055.

The footnote following the last sentence quoted above further stated 
that ``We underscore that the agency's decision regarding what 
standards would be maximum feasible for MYs 2017-2025 is made with 
reference to the rulemaking time frame and the circumstances of this 
final rule. Each CAFE rulemaking (indeed, each stage of any given CAFE 
rulemaking) presents the agency with new information that may affect 
how the agencies we balance the relevant factors.'' \2737\ NHTSA has 
been consistent over time, despite commenters' suggestions to the 
contrary, that maximum feasible is a balancing of factors; that all 
factors must be considered; and that information before the agency may 
change how the agency both understands and balances the statutory 
factors.
---------------------------------------------------------------------------

    \2737\ Id at fn. 1275.
---------------------------------------------------------------------------

    With regard to criteria and toxic air pollutant emissions, NHTSA 
agrees with commenters that the NPRM discussion of environmental 
implications did not specifically identify these emissions, but notes 
that air quality issues were discussed in a variety of places in the 
NPRM, DEIS, and PRIA, and that the monetized effects of air quality 
impacts were included in the overall cost-benefit analysis which 
informed NHTSA's balancing of factors, as discussed above. To the 
extent that commenters disagreed with the values or the agency's air 
quality analyses, those topics will be addressed in Section VII and 
VIII and in the FEIS. NHTSA has considered all of these findings along 
with other factors, as discussed below.
(d) Foreign Policy Implications
    In the NPRM, NHTSA explained that U.S. consumption and imports of 
petroleum products impose costs on the domestic economy that are not 
reflected in the market price for crude petroleum or in the prices paid 
by consumers for petroleum products such as gasoline. These costs 
include (1) higher prices for petroleum products resulting from the 
effect of U.S. oil demand on world oil prices, (2) the risk of 
disruptions to the U.S. economy caused by sudden increases in the 
global price of oil and its resulting impact of fuel prices faced by 
U.S. consumers, and (3) expenses for maintaining the strategic 
petroleum reserve (SPR) to provide a response option should a 
disruption in commercial oil supplies threaten the U.S. economy, to 
allow the U.S. to meet part of its International Energy Agency 
obligation to maintain emergency oil stocks, and to provide a national 
defense fuel reserve.\2738\ Higher U.S. consumption of crude oil or 
refined petroleum products increases the magnitude of these external 
economic costs, thus increasing the true economic cost of supplying 
transportation fuels above the resource costs of producing them. 
Conversely, reducing U.S. consumption of crude oil or refined petroleum 
products (by reducing motor fuel use) can reduce these external costs.
---------------------------------------------------------------------------

    \2738\ While the U.S. maintains a military presence in certain 
parts of the world to help secure global access to petroleum 
supplies, that is neither the primary nor the sole mission of U.S. 
forces overseas. Moreover, the scale of oil consumption reductions 
associated with CAFE standards would be insufficient to alter any 
existing military missions focused on ensuring the safe and 
expedient production and transportation of oil around the globe. 
Chapter 7 of the PRIA discussed this topic in more detail.
---------------------------------------------------------------------------

    The NPRM stated that while these costs are considerations, the 
United States has significantly increased oil production capabilities 
in recent years to the extent that the U.S. is currently producing 
enough oil to satisfy nearly all of its energy needs and is projected 
to continue to do so or become a net energy exporter. This has added 
new stable supply to the global oil market and reduced the urgency of 
the U.S. to conserve energy. The NPRM referred readers to the balancing 
discussion for more detail on this issue.
    Securing America's Energy Future commented that continuing to raise 
stringency would be good for energy security, spur innovation, and 
``advance the administration's energy dominance agenda.'' \2739\ CARB 
argued that the proposal would ``significantly diminish U.S. energy 
security,'' ``. . . contrary to the President's recent executive order 
to promote national security, and contrary to the intent of Congress in 
EPCA.'' \2740\
---------------------------------------------------------------------------

    \2739\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 6.
    \2740\ CARB, NHTSA-2018-0067-11783, at 316.
---------------------------------------------------------------------------

    Several commenters disagreed with the NPRM's suggestion that 
increases in U.S. oil production reduced the foreign policy 
implications relevant to the need of the U.S. to conserve energy. ACEEE 
commented that because the market for oil is global, ``. . . regardless 
of actual imports, the nation is still affected by what happens to oil 
worldwide, and oil remains a foreign policy concern . . . .'' \2741\ 
Securing America's Energy Future commented that increased U.S. 
production ``. . . has reduced some of the negative consequences of oil 
dependence, energy security is primarily a function of consumption, not 
production.'' \2742\ IPI argued that ``. . . the agencies falsely and 
inconsistently argue that the need to conserve energy has diminished 
because U.S. reliance on foreign oil has decreased,'' disagreeing with 
the NPRM's assumption that monopsony and military security costs 
resulting from the proposal would be zero.\2743\ The States and Cities 
commenters raised similar points, stating that ``U.S. military and 
foreign policy institutes'' place emphasis on ``global oil market 
stability and the stability of major oil-exporting nations,'' which the 
States and Cities argued had not changed as

[[Page 25150]]

U.S. exports have risen.\2744\ The States and Cities commenters further 
argued that if a quarter of U.S. oil consumed is still imported, then 
increases in consumption would necessarily raise imports, and thus also 
monopsony and military security costs associated with those 
imports.\2745\
---------------------------------------------------------------------------

    \2741\ ACEEE, NHTSA-2018-0067-12122, main comments, at 2.
    \2742\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 6.
    \2743\ IPI, NHTSA-2018-0067-12213, Appendix, at 2-3.
    \2744\ States and Cities, NHTSA-2018-0067-11735, at 76-77.
    \2745\ Id.
---------------------------------------------------------------------------

    CARB questioned whether it was accurate to assume that the U.S. 
would ever reach net exporter status, and commented that even if 
becoming a net exporter helped to insulate the Nation from the effects 
of reducing CAFE stringency, it would not lead to greater energy 
security until at least 2029, the first year for which AEO 2018 
forecasts that the U.S. will stop being a net importer.\2746\ CARB 
further argued that increased domestic oil production did not insulate 
the U.S. from risk, and that in fact ``. . . current conditions are 
more prone to risk due to lower available spare oil production capacity 
in major oil producing countries, meaning that a supply disruption is 
more likely to have a more pronounced effect on oil prices and U.S. 
energy security.'' \2747\
---------------------------------------------------------------------------

    \2746\ CARB, NHTSA-2018-0067-11783, at 317.
    \2747\ Id., at 319.
---------------------------------------------------------------------------

    Mr. Bordoff commented that geopolitical risk can still affect 
global oil prices, citing U.S. withdrawal from the Iran nuclear 
agreement and the reimposition of sanctions on Iranian oil sales; the 
collapse of Libyan oil production following conflict there; ongoing 
problems in Venezuela; a variety of short-term production outages in 
other producing areas; and even situations where geopolitics can result 
in lower prices rather than higher prices.\2748\
---------------------------------------------------------------------------

    \2748\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 3-4.
---------------------------------------------------------------------------

    IPI stated that ``. . . the protective value that the SPR offers 
given its size does automatically change as total U.S. petroleum 
consumption changes,'' and argued that it was not sufficient to 
consider only ``the budgetary costs for maintaining [the size of] the 
SPR.'' IPI thus argued that ``The agencies have failed to assess how 
much the relative protective value of the SPR will change as total U.S. 
consumption rises following the proposed rollback, and therefore have 
failed entirely to consider one important element of the national need 
to conserve energy.'' \2749\
---------------------------------------------------------------------------

    \2749\ IPI, NHTSA-2018-0067-12213, Appendix, at 4.
---------------------------------------------------------------------------

    Total energy independence for any country is only possible if it 
does not participate in the global energy markets, either because it 
consumes no energy (which is unrealistic) or because it produces enough 
energy to meet all of its energy needs and uses only energy that is 
produced domestically. As discussed above, NHTSA agrees with commenters 
that the oil market is global, and that events and situations abroad 
can affect oil prices even as U.S. oil production increases. The fact 
that the U.S. became a net oil exporter, at least on a weekly basis, in 
November 2019, and the evidence indicates that it will become a net oil 
exporter on a longer-term basis in MY 2020 does not change geopolitics 
in many parts of the world. Striving for energy independence in a 
global market necessarily means reducing risks, because even if the 
U.S. consumed only domestically-produced petroleum and continued to 
export, the U.S. economy would still be subject to oil price 
fluctuations due to external events and situations. The NPRM was clear 
on all of these points.\2750\ The NPRM and PRIA repeatedly emphasized 
that changes in the oil market meant that the risk of damage to the 
U.S. economy and of additional pain for U.S. drivers is lower than it 
was at the beginning of the CAFE program, not that it was eliminated 
entirely. NHTSA agrees with commenters that risk still exists, and that 
both production and consumption of oil are relevant to how big that 
risk might be. NHTSA simply believes, as explained in the NPRM and as 
explained again below, that the risk is lower than it would have been 
in the absence of the rapid growth in U.S. oil production, and that the 
lower risk means that the need of the U.S. to conserve energy, from 
this perspective, is less dire than it was at earlier points in the 
program.
---------------------------------------------------------------------------

    \2750\ See 83 FR at 43213-15.
---------------------------------------------------------------------------

    The analyses for both the NPRM and the final rule account for the 
ongoing economic risk of participating in the global oil market by 
placing a value on energy security. The energy security value is made 
of several components. While commenters are correct that neither the 
NPRM nor the final rule analyses attributed a positive cost to the 
monopsony or military security components, the agencies do employ a 
cost for macroeconomic shock risk as part of energy security. Section 
VI discusses these estimates in more detail; for purposes of this 
discussion, NHTSA only notes that these issues are accounted for in the 
agencies' cost-benefit analysis, and to the extent that zero values are 
used for some elements, the reason for that is explained at length in 
those sections and public comments received on these issues did not 
present new information to change the agencies' minds on those values.
    With regard to the comment that NHTSA should be accounting for the 
``protective value'' of the SPR along with the literal cost of 
maintaining it, NHTSA is not in a position at this time to attempt to 
estimate such a value, and notes that the commenter provided no 
suggestions as to how to do so. The Department of Energy's website 
states that the maximum number of days of import protection provided by 
the SPR is 143 days, and that it takes 13 days from Presidential 
decision for SPR fuel to enter the market.\2751\ The 1973 OPEC oil 
embargo lasted from October 1973 to March 1974, roughly 150 days. As 
explained, NHTSA continues to believe that the effect of increased U.S. 
oil production is to stabilize, broadly, global oil markets. The longer 
a sustained spike in prices due to geopolitical events continues, the 
greater incentive U.S. shale production has to respond. NHTSA believes 
that it is foreseeable that the SPR could be utilized to help mitigate 
a price shock in the interim, for the majority of foreseeable shock 
situations.
---------------------------------------------------------------------------

    \2751\ See https://www.energy.gov/fe/services/petroleum-reserves/strategic-petroleum-reserve/spr-quick-facts-and-faqs.
---------------------------------------------------------------------------

(5) Factors That NHTSA Is Prohibited From Considering
    The NPRM explained that EPCA also provides that in determining the 
level at which it should set CAFE standards for a particular model 
year, NHTSA may not consider the ability of manufacturers to take 
advantage of several EPCA provisions that facilitate compliance with 
CAFE standards and thereby reduce the costs of compliance.\2752\ As 
discussed further in Section IX below, NHTSA cannot consider compliance 
credits that manufacturers earn by exceeding the CAFE standards and 
then use to achieve compliance in years in which their measured average 
fuel economy falls below the standards. NHTSA also cannot consider the 
use of alternative fuels by dual fuel vehicles nor the availability of 
dedicated alternative fuel vehicles--including battery-electric 
vehicles--in any model year. EPCA encourages the production of 
alternative fuel vehicles by specifying that their fuel economy is to 
be determined using a special calculation procedure that results in 
those vehicles being assigned a higher equivalent fuel economy level 
than they actually achieve.
---------------------------------------------------------------------------

    \2752\ 49 U.S.C. 32902(h).
---------------------------------------------------------------------------

    The NPRM further explained that the effect of the prohibitions 
against

[[Page 25151]]

considering these statutory flexibilities in setting the CAFE standards 
is that the flexibilities remain voluntarily-employed measures. If 
NHTSA were instead to assume manufacturer use of those flexibilities in 
setting new standards, higher standards would appear less costly and 
therefore more feasible, which would thus effectively require 
manufacturers to use those flexibilities in order to meet higher 
standards. By keeping NHTSA from including them in our stringency 
determination, the provision ensures that these statutory credits 
remain true compliance flexibilities.
    Additionally, for the non-statutory fuel economy improvement value 
program that NHTSA developed by regulation, the NPRM stated that NHTSA 
does not consider these subject to the EPCA prohibition on considering 
flexibilities. EPCA is very clear as to which flexibilities are not to 
be considered. When the agency has introduced additional flexibilities 
such as A/C efficiency and ``off-cycle'' technology fuel economy 
improvement values, NHTSA has considered those technologies as 
available in the analysis. Thus, today's analysis includes assumptions 
about manufacturers' use of those technologies, as detailed in Section 
VI.
    Michalek and Whitefoot commented that ``[w]e find [the statutory 
prohibition on considering certain flexibilities in determining maximum 
feasible CAFE standards] problematic because the automakers use these 
flexibilities as a common means of complying with the regulation, and 
ignoring them will bias the cost-benefit analysis to overestimate 
costs.'' \2753\ IPI commented that ``it is not clear that the statutory 
prohibition on considering credit availability was intended to apply to 
banked credits,'' because 49 U.S.C. 32902(h)(3) was
---------------------------------------------------------------------------

    \2753\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 10-11.

added . . . as a `conforming amendment' to EISA, which was the 
statute that gave NHTSA authority to allow credit trading and 
transferring; meanwhile, banking and borrowing have been part of 
NHTSA's authority since EPCA in 1975. In 1989, e.g., NHTSA 
explicitly relied on the availability of `credit banks' to justify 
maintaining the MY 1990 standard at 27.5 mpg instead of lowering its 
stringency. NHTSA has not explained why it now believes it may not 
more fully consider banking.\2754\
---------------------------------------------------------------------------

    \2754\ IPI, NHTSA-2018-0067-12213, Appendix, at 19.

    NHTSA agrees, as explained in the NPRM, that if the agency was able 
to consider the compliance flexibilities in determining maximum 
feasible standards, more-stringent standards would appear less costly 
and therefore more feasible. NHTSA is nevertheless bound by the 
statutory prohibition on considering the above-mentioned flexibilities. 
As for IPI's disagreement that 32902(h)(3) should apply to banked 
credits because it was labeled a ``conforming amendment,'' NHTSA looks 
to the specific statutory language provided, which prohibits 
``[consideration], when prescribing a fuel economy standard, [of] the 
trading, transferring or availability of credits . . . .'' (Emphasis 
added.) IPI's suggested interpretation would render ``availability'' as 
surplusage. If Congress had meant the prohibition to apply only to 
traded and transferred credits, it would have said so. Instead, 
Congress also prohibited consideration of the ``availability of 
credits,'' which must be read reasonably to refer to ``what credits are 
available,'' i.e., banked credits. The fact that NHTSA considered the 
availability of banked credits in 1989, prior to establishment of this 
statutory prohibition, has no bearing in a post-EISA world.
    Nonetheless, NHTSA notes that it is informed by the ``real-world'' 
analysis presented in the FRIA, which accounts for credit availability 
and usage, and manufacturers' ability to employ alternative fueled 
vehicles--for purpose of conformance with E.O. 12866. Under the real-
world analysis, compliance does, in fact, appear less costly. For 
example, today's ``real world'' analysis shows manufacturers' costs 
averaging about $1,420 in MY 2029 under the final standards, as 
compared to the $1,640 shown by the ``standard setting'' analysis. 
However, for purposes of determining maximum feasible CAFE levels, 
NHTSA considers only the ``standard-setting'' analysis shown in the 
NPRM, consistent with Congress's direction.
(f) EPCA/EISA Requirements That No Longer Apply Post-2020
    The NPRM explained that Congress amended EPCA through EISA to add 
two requirements not yet discussed in this section relevant to 
determination of CAFE standards during the years between MY 2011 and MY 
2020 but not beyond. First, Congress stated that, regardless of NHTSA's 
determination of what levels of standards would be maximum feasible, 
standards must be set at levels high enough to ensure that the combined 
U.S. passenger car and light truck fleet achieves an average fuel 
economy level of not less than 35 mpg no later than MY 2020.\2755\ And 
second, between MYs 2011 and 2020, the standards must ``increase 
ratably'' in each model year.\2756\ Neither of these requirements apply 
after MY 2020, so given that this rulemaking concerns the standards for 
MY 2021 and after, the NPRM stated that they are not relevant to this 
rulemaking.
---------------------------------------------------------------------------

    \2755\ 49 U.S.C. 32902(b)(2)(A).
    \2756\ 49 U.S.C. 32902(b)(2)(C).
---------------------------------------------------------------------------

    CARB commented that because the proposal did not ``provide for 
improved efficiency of motor vehicles'' over the long term, 
``Stagnating the standards violates Congressional direction to ratably 
increase fuel economy when the technology for doing so has been 
demonstrated to exist (which it does . . .) or could be developed in 
the necessary time.'' \2757\
---------------------------------------------------------------------------

    \2757\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84.
---------------------------------------------------------------------------

    NHTSA notes, again, that the statutory language is clear that 
Congress only directed ratable increases in stringency through MY 2020. 
After MY 2020, the statutory language is clear that standards simply 
need be ``maximum feasible, as determined by the Secretary.'' Some 
commenters may have disagreed that the proposal represented maximum 
feasible levels, but there is no statutory basis for arguing that the 
``ratable increase'' requirement extends beyond MY 2020.
(g) Other Considerations in Determining Maximum Feasible Standards
    The NPRM explained that NHTSA has historically considered the 
potential for adverse safety consequences in setting CAFE standards. 
This practice has been consistently approved in case law. As courts 
have recognized, ``NHTSA has always examined the safety consequences of 
the CAFE standards in its overall consideration of relevant factors 
since its earliest rulemaking under the CAFE program.'' Competitive 
Enterprise Institute v. NHTSA, 901 F.2d 107, 120 n. 11 (D.C. Cir. 1990) 
(``CEI-I'') (citing 42 FR 33534, 33551 (June 30, 1977)). The courts 
have consistently upheld NHTSA's implementation of EPCA in this manner. 
See, e.g., Competitive Enterprise Institute v. NHTSA, 956 F.2d 321, 322 
(D.C. Cir. 1992) (``CEI-II'') (in determining the maximum feasible fuel 
economy standard, ``NHTSA has always taken passenger safety into 
account'') (citing CEI-I, 901 F.2d at 120 n. 11); Competitive 
Enterprise Institute v. NHTSA, 45 F.3d 481, 482-83 (D.C. Cir. 1995) 
(``CEI-III'') (same); Center for Biological Diversity v. NHTSA, 538 
F.3d 1172, 1203-04 (9th Cir. 2008) (upholding NHTSA's analysis of 
vehicle safety issues associated with weight in connection with the MYs 
2008-2011

[[Page 25152]]

light truck CAFE rulemaking). Thus, NHTSA explained that in evaluating 
what levels of stringency would result in maximum feasible standards, 
NHTSA assesses the potential safety impacts and considers them in 
balancing the statutory considerations and to determine the maximum 
feasible level of the standards.
    The attribute-based standards that Congress requires NHTSA to set 
help to mitigate the negative safety effects of the historical single 
number standards originally required in EPCA, and in past rulemakings, 
NHTSA constrained its modeling so as not to consider possible mass 
reduction in lower weight vehicles in its analysis, which affected the 
resulting assessment of potential adverse safety impacts. That 
analytical approach did not reflect, however, the likelihood that 
automakers may pursue the most cost-effective means of improving fuel 
efficiency to comply with CAFE requirements. For the NPRM, as for the 
final rule, the modeling did not limit the amount of mass reduction 
that is applied to any segment, but rather considered that automakers 
may apply mass reduction based upon cost-effectiveness, similar to most 
other technologies. NHTSA does not, of course, mandate the use of any 
particular technology by manufacturers in meeting the standards. The 
NPRM and today's final rule, like the Draft TAR, also considered the 
safety effect associated with the additional vehicle miles traveled due 
to the rebound effect.
    NHTSA explained that the NPRM considered the safety effects of 
vehicle scrappage rates on the fleet as a whole. The NPRM also 
explained NHTSA's consideration of the effect of additional expenses in 
fuel savings technology on the affordability of vehicles--the 
likelihood that increased standards will result in consumers being 
priced out of the new vehicle market and choosing to keep their 
existing vehicle or purchase a used vehicle. Since new vehicles are 
significantly safer than used vehicles, slowing fleet turnover to newer 
vehicles results in older and less safe vehicles remaining on the roads 
longer. NHTSA stated that this significantly affects the safety of the 
United States light duty fleet, as described more fully in in the 
safety section of the NPRM and in Chapter 11 of the PRIA. Furthermore, 
as fuel economy standards become more stringent, and more fuel 
efficient vehicles are introduced into the fleet, fueling costs are 
reduced. This results in consumers driving more miles, which results in 
more crashes and increased highway fatalities.
    A number of commenters disagreed with a variety of aspects of the 
NPRM's analysis of safety, and several also disagreed with how NHTSA 
considered safety along with the other factors in the proposal. The 
States and Cities commenters, for example, agreed that ``NHTSA has 
historically considered safety impacts when setting maximum feasible 
standards,'' but argued that:

in the Proposed Rollback, NHTSA departs from its past practice by 
relying on completely novel and unsupported theories regarding the 
linkages between fuel economy and safety that do not reflect 
reality. In the past, NHTSA has considered the safety of the 
technologies that improve fuel economy. [citations omitted] In the 
Proposed Rollback, however, NHTSA has linked safety concerns with 
rebound and scrappage effects of more stringent fuel economy 
standards. [citations omitted] As discussed [elsewhere], these 
theories are unsupported, implausible, and contradicted by numerous 
experts--rendering them arbitrary and capricious. The agency has 
also failed to acknowledge or adequately justify its break with past 
analyses of safety. See Fox Television, 556 U.S. at 515.'' \2758\
---------------------------------------------------------------------------

    \2758\ States and Cities, NHTSA-2018-0067-11735, at 77.

    EDF commented that NHTSA cannot ``. . . lawfully rely upon the 
repercussions of increased driving as a justification. . . . The fact 
that the standards do not `compel' this driving prevents such reliance, 
and . . . [EPCA/EISA] nowhere indicate that [NHTSA] can refuse to 
comply with [its] statutory obligations by pointing to a projection 
that individuals might drive more and in doing so, some of them will 
get into traffic accidents.\2759\ EDF further argued that:
---------------------------------------------------------------------------

    \2759\ EDF, NHTSA-2018-0067-12137, Supplemental Safety Comments, 
at 3.

    It is especially unlikely that Congress intended for NHTSA to 
consider potential increases in driving (or . . . `VMT'). Under 
basic economic theory and under the Agency's traditional analysis 
(including their analysis of this proposal), an improvement in fuel 
economy--which makes driving cheaper--would be expected to lead to 
some increase in driving for households that are sensitive to and 
conscious of that effect on their budgets. Thus, consideration of 
VMT impacts could be used to undermine any fuel economy standard. 
Because VMT is `a factor [that] is both so indirectly related to 
[fuel economy] and so full of potential for canceling the 
conclusions drawn from [a fuel economy analysis] . . . it would 
surely have been expressly mentioned in [the statute] had Congress 
meant it to be considered.' Whitman v. Am. Trucking Associations, 
531 U.S. 457, 469 (2001).'' \2760\
---------------------------------------------------------------------------

    \2760\ Id.

    Other comments on safety as part of the legal justification varied. 
NESCAUM claimed that NHTSA's safety justification ``is disputed by 
EPA's technical staff based on their identification of flaws in NHTSA's 
analysis,'' suggesting that it was therefore invalid and not a basis 
for decision-making.\2761\ Global commented that there was no policy 
reason for freezing the level of standards due to mass reduction 
concerns (i.e., safety), given footprint standards.\2762\ IPI argued 
that it was inappropriate to account for vehicle safety-related deaths 
and injuries ``without an adequate discussion of the health and safety 
impacts of the Proposed Rule's increased emissions or without an 
accurate estimate of the actual safety impact of the rollback versus 
the 2012 standards.'' \2763\
---------------------------------------------------------------------------

    \2761\ NESCAUM, NHTSA-2018-0067-11691, at 3.
    \2762\ Global, NHTSA-2018-0067-12032, Attachment A, at A-32.
    \2763\ IPI, NHTSA-2018-0067-12213, Appendix, at 11.
---------------------------------------------------------------------------

    NHTSA agrees with commenters that the safety analysis conducted to 
inform this rulemaking (both NPRM and final rule) is different from--
broader than--past safety analyses conducted to inform CAFE and 
CO2 rulemakings. NHTSA disagrees, however, that the agency 
failed to acknowledge or explain this fact. The NPRM directly 
acknowledges and explains the evolution of the safety analysis over 
time and why, specifically, the NPRM included the safety effects of 
rebound and scrappage phenomena.\2764\ The NPRM also expressly sought 
comment on these elements of the safety analysis and the safety 
analysis generally, before explaining how they worked and describing 
their tentative findings in considerable detail. It is inaccurate for 
commenters to claim that the agency did not acknowledge or explain 
these changes. Commenters' disagreement with the substance of the 
safety analysis does not create a valid process complaint here. Section 
VI discusses in detail the comments received on the substance of the 
safety analysis, including a number of comments citing deliberative 
feedback provided by some members of EPA staff during NPRM development, 
and contains the agencies' responses. With regard to the comment from 
EDF, as explained above, the premise that vehicles may be driven more 
or less in response to more or less stringent CAFE (or CO2) 
standards is called the rebound effect, and it is discussed at length 
in Section VI above. The rebound effect has been factored into 
rulemaking cost-benefit analyses and reduced CAFE and CO2 
standard benefits in such analyses for well over a decade,\2765\ and 
EPA and NHTSA have

[[Page 25153]]

written repeatedly about and considered the magnitude of this effect. 
NHTSA is aware that some commenters disagree that a rebound effect even 
exists for fuel economy, and understands how such commenters would 
correspondingly disagree that VMT-related safety effects could arise 
from differences in CAFE standards. But NHTSA does not agree that the 
rebound effect is zero, and correspondingly believes that safety 
effects from additional driving (due to exposure to crashes) exist and 
are capable of quantification for analytical purposes.
---------------------------------------------------------------------------

    \2764\ See 83 FR at 43106-07.
    \2765\ See, e.g., 68 FR 16868, 16878 (Apr. 7, 2003).
---------------------------------------------------------------------------

    Moreover, if EDF were correct that agencies may consider only the 
behavior that regulations directly ``compel,'' then CAFE analysis would 
be challenged to consider even fuel savings--the purpose of CAFE 
standards--because the standards do not compel Americans to drive, or 
to buy new vehicles, or to buy any vehicles at all. Reasonable 
assumptions about how much Americans drive (depending on how much it 
costs to drive, among other things), and what vehicles Americans buy 
and how often they buy them (depending on how much those vehicles cost, 
among other things), are useful and important for including in analyses 
that help decision-makers distinguish between different levels of 
potential CAFE standards. Circular A-4 additionally directs agencies to 
consider ancillary effects of rulemakings.\2766\ NHTSA believes that it 
is reasonable to consider these effects as part of the safety analysis, 
and to consider safety effects as part of its determination of maximum 
feasible standards.
---------------------------------------------------------------------------

    \2766\ See OIRA, ``Regulatory Impact Analysis: A Primer,'' at 7, 
https://www.reginfo.gov/public/jsp/Utilities/circular-a-4_regulatory-impact-analysis-a-primer.pdf (``In addition to the 
direct benefits and costs of each alternative, the list should 
include any important ancillary benefits and countervailing risks. 
An ancillary benefit is a favorable impact of the alternative under 
consideration that is typically unrelated or secondary to the 
purpose of the action (e.g., reduced refinery emissions due to more 
stringent fuel economy standards for light trucks). A countervailing 
risk is an adverse economic, health, safety, or environmental 
consequence that results from a regulatory action and is not already 
accounted for in the direct cost of the action (e.g., adverse safety 
impacts from more stringent fuel-economy standards for light 
trucks). As with other benefits and costs, an effort should be made 
to quantify and monetize both ancillary benefits and countervailing 
risks.'')
---------------------------------------------------------------------------

(2) Administrative Procedure Act
    To be upheld under the ``arbitrary and capricious'' standard of 
judicial review in the APA, an agency rule must be rational, based on 
consideration of the relevant factors, and within the scope of the 
authority delegated to the agency by the statute. The agency must 
examine the relevant data and articulate a satisfactory explanation for 
its action including a ``rational connection between the facts found 
and the choice made.'' \2767\
---------------------------------------------------------------------------

    \2767\ Burlington Truck Lines, Inc., v. United States, 371 U.S. 
156, 168 (1962).
---------------------------------------------------------------------------

    Statutory interpretations included in an agency's rule are subject 
to the two-step analysis of Chevron, U.S.A. v. Natural Resources 
Defense Council.\2768\ Under step one, where a statute ``has directly 
spoken to the precise question at issue,'' id. at 842, the court and 
the agency ``must give effect to the unambiguously expressed intent of 
Congress.'' \2769\ If the statute is silent or ambiguous regarding the 
specific question, the court proceeds to step two and asks ``whether 
the agency's answer is based on a permissible construction of the 
statute.'' \2770\
---------------------------------------------------------------------------

    \2768\ 467 U.S. 837 (1984).
    \2769\ Id. at 843.
    \2770\ Id.
---------------------------------------------------------------------------

    If an agency's interpretation differs from the one that it has 
previously adopted, the agency need not demonstrate that the prior 
position was wrong or even less desirable. Rather, the agency would 
need only to demonstrate that its new position is consistent with the 
statute and supported by the record and acknowledge that this is a 
departure from past positions. The Supreme Court emphasized this in FCC 
v. Fox Television.\2771\ When an agency changes course from earlier 
regulations, ``the requirement that an agency provide a reasoned 
explanation for its action would ordinarily demand that it display 
awareness that it is changing position,'' but ``need not demonstrate to 
a court's satisfaction that the reasons for the new policy are better 
than the reasons for the old one; it suffices that the new policy is 
permissible under the statute, that there are good reasons for it, and 
that the agency believes it to be better, which the conscious change of 
course adequately indicates.'' \2772\ The APA also requires that 
agencies provide notice and comment to the public when proposing 
regulations,\2773\ as the agencies did when publishing the NPRM for 
this rulemaking.
---------------------------------------------------------------------------

    \2771\ 556 U.S. 502 (2009).
    \2772\ Id., at 1181.
    \2773\ 5 U.S.C. 553.
---------------------------------------------------------------------------

a) Requests To Extend the Comment Period
    On August 2, 2018, the agencies published the NPRM on the agencies' 
respective websites, soliciting public comments.\2774\ On August 24, 
2018, the Federal Register published the NPRM, which began a 60-day 
public comment period.\2775\ The public comment period would have ended 
on October 23, 2018, but the agencies extended the comment period until 
October 26, 2018.\2776\ In the Federal Register notice extending the 
comment period, the agencies explained that they were denying requests 
for an extension of the comment period by at least 60 days, explaining 
that ``[a]utomakers will need maximum lead time to respond to the final 
rule[.]'' \2777\ Although the comment period ultimately closed on 
October 26, 2018, the agencies' dockets remained open, and the agencies 
continued to accept and consider comments, to the extent possible, for 
more than one year after the comment period began.\2778\
---------------------------------------------------------------------------

    \2774\ https://www.nhtsa.gov/corporate-average-fuel-economy/safe; https://www.epa.gov/newsreleases/us-epa-and-dot-propose-fuel-economy-standards-my-2021-2026-vehicles.
    \2775\ 83 FR 42986 (Aug. 24, 2018).
    \2776\ See 83 FR 48578 (Sept. 26, 2018) (extending comment 
period).
    \2777\ Id.
    \2778\ The agencies notified the public of this possibility in 
the NPRM, stating that: ``To the extent practicable, we will also 
consider comments received after'' the close of the comment period. 
83 FR 42986, 43471 (Aug. 24, 2018).
---------------------------------------------------------------------------

    After publishing the NPRM, the agencies received a number of 
requests to extend the comment period, generally for an additional 60 
days.\2779\ For example, seventeen States and the District of Columbia 
jointly requested a 60-day extension of the comment period.\2780\ That 
request cited the voluminous record, the complexity of the material, 
and the profound potential impact on human health and the environment, 
among other things.\2781\ The City of Los Angeles and New York State 
Department of Environmental Conservation also requested a 60-day 
extension, for similar reasons.\2782\ In addition, 32 United States 
Senators jointly requested a 60-day extension of the comment 
period.\2783\ The Senators argued that an extension was appropriate to 
ensure adequate public participation with such an important rule.\2784\ 
Several non-government organizations similarly requested a 60-day 
extension of the comment period due to the complexity of the issues and

[[Page 25154]]

the importance of the proposed rule.\2785\ Other organizations also 
requested a 60-day extension, stressing the complexity of the issues 
and the significance of the proposed rule's impact on the 
environment.\2786\ The American Lung Association also requested a 60-
day extension of the comment period, asserting that it needed more time 
to analyze the impact of the proposed rule on human health.\2787\ The 
California Air Resources Board (CARB) likewise requested a 60-day 
extension, in part, based on information that it asserted should have 
been included in the NPRM.\2788\ New York University School of Law's 
Institute for Policy Integrity similarly requested a 60-day extension 
based on information that it contended should have been included in the 
NPRM's ``sensitivity analysis table for the `Cumulative Changes in 
Fleet Size, Travel (VMT), Fatalities, Fuel Consumption and C02 
Emissions through MY2029.' '' \2789\
---------------------------------------------------------------------------

    \2779\ See 83 FR 48578 (Sept. 26, 2018).
    \2780\ See comments from the State of California et al., Request 
for an extension, Docket No. NHTSA-2018-0067-3458.
    \2781\ See id.
    \2782\ Also for similar reasons, the Minnesota Pollution Control 
Agency and the Minnesota Department of Transportation submitted a 
joint request for a 120-day extension of the comment period. See 
comments from the Minnesota Pollution Control Agency and Minnesota 
Department of Transportation, Docket No. NHTSA-2018-0067-3580.
    \2783\ See comments from 32 U.S. Senators (Kamala D. Harris et 
al.), Docket No. NHTSA-2018-0067-5643.
    \2784\ See id.
    \2785\ See, e.g., comments from the Alliance of Automobile 
Manufacturers, Docket No. NHTSA-2018-0067-3619; Communities for a 
Better Environment, Docket No. EPA-HQ-OAR-2018-0283-1095; Consumer 
Federation of America, NHTSA-2018-0067-3400; Edison Electric 
Institute, received by mail; and South Coast Air Quality Management 
District, Docket No. EPA-HQ-OAR-2018-0283-0885.
    \2786\ See, e.g., comments from the Environmental Law and Policy 
Center, NHTSA-2018-0067-2728; Georgetown Climate Center, Docket No. 
NHTSA-2018-0067-3610; Center for Biological Diversity, Conservation 
Law Foundation, Earthjustice, Environmental Defense Fund, Natural 
Resources Defense Council, Public Citizen,
    Sierra Club, and Union of Concerned Scientists, Docket No. 
NHTSA-2018-0067-3278; and National Governors Association, Docket No. 
EPA-HQ-OAR-2018-0283-0871.
    \2787\ See comments from American Lung Association, Docket No. 
NHTSA-2018-0067-3615.
    \2788\ See comments from California Air Resources Board, Docket 
No. NHTSA-2018-0067-4166.
    \2789\ See comments from New York University School of Law's 
Institute for Policy Integrity, NHTSA-2018-0067-5641.
---------------------------------------------------------------------------

    The agencies do not believe a further extension of the comment 
period was warranted under the circumstances.\2790\ The APA does not 
specify a minimum number of days for a comment period.\2791\ Two 
Executive Orders also provide direction to Federal agencies with 
respect to the length of a comment period for a proposed rule.\2792\ 
Executive Order 12,866 states that ``[e]ach agency shall (consistent 
with its own rules, regulations, or procedures) provide the public with 
meaningful participation in the regulatory process . . . . In addition, 
each agency should afford the public a meaningful opportunity to 
comment on any proposed regulation, which in most cases should include 
a comment period of not less than 60 days.'' \2793\ Additionally, 
Executive Order 13,563 reaffirmed Executive Order 12,866's directive 
that comment periods should generally not be less than 60 days, 
stating: ``To the extent feasible and permitted by law, each agency 
shall afford the public a meaningful opportunity to comment through the 
internet on any proposed regulation, with a comment period that should 
generally be at least 60 days.'' \2794\ More recently, in December of 
2018, the Department of Transportation implemented DOT Order 2100.6, 
which provides its operating administrations, including NHTSA, with 
direction on appropriate rulemaking processes and procedures.\2795\ 
While not yet effective at the time the proposal was published, the 
Order provides that ``the comment period for significant DOT rules 
should be at least 45 days.'' \2796\ The 63 day comment period for the 
proposal far exceeded this amount.
---------------------------------------------------------------------------

    \2790\ See 83 FR 48578 (Sept. 26, 2018) (extending comment 
period until October 26, 2018 and denying requests for longer 
extensions).
    \2791\ See 5 U.S.C. 553(c).
    \2792\ The Executive Orders do not create any enforceable right 
or benefit by a party against any federal agency. E.O. 12,866 Sec.  
10; E.O. 13,563 Sec.  7(d).
    \2793\ Executive Order 12,866 Sec.  6(a)(1).
    \2794\ Executive Order 13,563 Sec.  2(b).
    \2795\ DOT Order 2100.6, ``Policies and Procedures for 
Rulemakings,'' available at: https://www.transportation.gov/sites/dot.gov/files/docs/regulations/328561/dot-order-21006-rulemaking-process-signed-122018.pdf.
    \2796\ Id., at (11)(i)(3).
---------------------------------------------------------------------------

    Consistent with these principles, courts give broad discretion to 
agencies in determining the reasonableness of a comment period. Courts 
have frequently upheld comment periods that were significantly less 
than the 63-day comment period here. See Connecticut Light & Power Co. 
v. Nuclear Regulatory Comm'n, 673 F.2d 525, 534 (D.C. Cir. 1982) 
(upholding a 30-day comment period and stating that ``neither statute 
nor regulation mandates that the agency do more''); see also North 
American Van Lines v. ICC, 666 F.2d 1087, 1092 (7th Cir. 1981) 
(upholding a 45-day comment period).\2797\ In addition to the length of 
a comment period, courts consider the number of comments received and 
whether comments had an effect on an agency's final rule, in assessing 
whether the public had a meaningful opportunity to comment.\2798\
---------------------------------------------------------------------------

    \2797\ In certain circumstances, particularly urgent ones, 
courts have even upheld comment periods of less than 30 days. See 
Omnipoint Corp. v. FCC, 78 F.3d 620, 629-30 (D.C. Cir. 1996) 
(holding that a 14-day comment period was sufficient given the 
``urgent necessity for rapid administrative action under the 
circumstances''); see also Fla. Power & Light Co. v. United States, 
846 F.2d 765, 772 (D.C. Cir. 1988) (upholding a 15-day comment 
period given a deadline that Congress imposed on the Nuclear 
Regulatory Commission to finalize its rule).
    \2798\ See Florida Power & Light, Co. v. United States, 846 F.2d 
765, 772 (D.C. Cir. 1988); see also Conference of State Bank Sup'rs 
v. Office of Thrift Supervision, 792 F. Supp. 837, 844 (D.D.C. 
1992).
---------------------------------------------------------------------------

    These principles are easily satisfied here. Here, the agencies 
initially provided a 60-day comment period and then further extended it 
to ensure compliance with the Clean Air Act. The Clean Air Act requires 
that the record of proceedings allowing oral presentation of data, 
views, and arguments on a proposed rule be kept open for 30 days after 
completion of a proceeding to provide an opportunity for submission of 
rebuttal and supplementary information.\2799\ Because the final 
``proceeding allowing oral presentation of data, views, and arguments'' 
was expected to be on September 26, 2018, the comment period for the 
proposed rule was extended by three days to meet that 
requirement.\2800\
---------------------------------------------------------------------------

    \2799\ 42 U.S.C. 7607(d)(5).
    \2800\ See 83 FR 48578, 48581 (Sept. 26, 2018).
---------------------------------------------------------------------------

    The 63-day comment period was consistent with what the law 
requires.\2801\ While the agencies understand and agree with commenters 
about the importance and complexity of the issues here, the public 
docket demonstrates that the public had a meaningful opportunity to 
comment on the proposed rule.\2802\ The agencies received a total of 
more than 750,000 public comments, many of which commented on detailed, 
technical portions of the proposed rule. For instance, the California 
Air Resources Board provided 415 pages of detailed comments involving 
very specific aspects of the proposal,\2803\ and the Auto Alliance 
filed 202 pages of detailed comments, and commissioned a separate 
econometric study analyzing the effects of multiple alternatives.\2804\ 
This is clear evidence that the public had not only the opportunity to 
review and comment on the proposal, but to do so with an extraordinary 
level of detail.
---------------------------------------------------------------------------

    \2801\ In any event, the two Executive Orders explicitly state 
that they do not create any enforceable right or benefit by a party 
against any federal agency. See Executive Order 12,866 Sec.  10; see 
also Executive Order 13,563 Sec.  7(d).
    \2802\ See Rural Cellular Ass'n v. FCC, 588 F.3d 1095, 1101 
(D.C. Cir. 2009).
    \2803\ NHTSA-2018-0067-11873.
    \2804\ NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Finally, notwithstanding the sufficiency of the agencies' 63-day 
comment period, the agencies published their NPRM on their websites on 
August 2, 2018, more than three weeks before the comment period 
formally opened on August 24, and this effectively provided the public 
with 22 additional days in

[[Page 25155]]

which to review the proposal and draft comments.\2805\
---------------------------------------------------------------------------

    \2805\ The agencies' public dockets also remained open for more 
than one year after the start of the comment period, and the 
agencies considered some late comments received, to the extent 
practicable, although many late comments were simply too untimely to 
be considered.
---------------------------------------------------------------------------

b) Other Comments on Public Participation
    Several commenters objected to NHTSA's 15-page limit on primary 
comments, asserting that it impacted the public's ability to 
meaningfully participate in the rulemaking process.\2806\ However, as 
certain of the commenters acknowledged, the NPRM also explicitly stated 
that commenters could also submit attachments--without any page 
limit.\2807\ Thus, the page limit on primary comments did not prevent 
commenters from presenting any information they deemed relevant to the 
agencies. Both primary comments and their attachments are available in 
the agencies' public dockets, and were considered by the agencies in 
this rulemaking as demonstrated by the responses to comments discussed 
throughout this final rule.
---------------------------------------------------------------------------

    \2806\ See States of California et al., Attachment1_States and 
Cities Detailed Comments, Docket No. NHTSA-2018-0067-11735, at 46; 
Center for Biological Diversity, et al., NHTSA-2018-0067-12088; 
CARB, NHTSA-2018-0067-1187; Environmental Defense Fund, NHTSA-2018-
0067-12108; BlueGreen Alliance, NHTSA-2018-0067-12440; Connecticut 
Department of Energy and Environmental Protection (DEEP), EPA-HQ-
OAR-2018-0283-4202.
    \2807\ 83 FR 43470 (Aug. 24, 2018) (citing 49 CFR 553.21).
---------------------------------------------------------------------------

    NHTSA's 15-page limit simply prescribed the form that comments 
should take: A concise summary comment of up to 15 pages, with optional 
attachments with no page limit. Many commenters submitted extensive 
attachments to their comments, including commenters that objected to 
the 15-page limit for primary comments. For example, several States and 
cities that jointly commented submitted a 13-page primary comment, 
accompanied by 145 pages of ``detailed comments'' and three appendices 
totaling 101 additional pages.\2808\ The 15-page limit had the effect 
of creating executive summaries of otherwise voluminous comments, which 
increased efficiency during the rulemaking process. This was NHTSA's 
stated purpose for the 15-page limit. As explained in the NPRM: ``NHTSA 
established this limit to encourage you to write your primary comments 
in a concise fashion.'' \2809\ In any event, no commenter was prevented 
from submitting information to the agencies based on NHTSA's page 
limitation for primary comments. The agencies strongly disagree that 
public participation was impeded by NHTSA's specification that primary 
comments were limited to 15 pages.
---------------------------------------------------------------------------

    \2808\ States of California et al., NHTSA-2018-0067-11735.
    \2809\ 83 FR 43470 (Aug. 24, 2018).
---------------------------------------------------------------------------

    On August 2, 2018, the agencies published a joint Notice of 
Proposed Rulemaking (NPRM) on the agencies' respective websites, which 
solicited public comments on ``The Safer Affordable Fuel-Efficient 
(SAFE) Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light 
Trucks.'' \2810\ The NPRM indicated that the public may submit written 
comments by any of the following methods: Online through the Federal 
eRulemaking Portal at www.regulations.gov, by fax, by mail, or by hand 
delivery. The NPRM also notified the public that the agencies planned 
to hold three joint public hearings, and would accept oral and written 
comments at the hearings. The NPRM indicated that the agencies planned 
to hold the hearings in Washington, DC; the Detroit, Michigan area; and 
the Los Angeles, California area, but indicated that the specific 
addresses and dates for the hearings would be announced in a 
supplemental Federal Register notice.\2811\ On August 24, 2018, the 
agencies published a notice in the Federal Register, which provided new 
locations for two of the three hearings and added dates for each 
hearing.\2812\ That notice informed the public that the agencies 
planned to hold three joint public hearings during the comment period: 
(1) On September 24, 2018 in Fresno, California; (2) on September 25, 
2018 in Dearborn, Michigan; and (3) on September 26, 2018 in 
Pittsburgh, Pennsylvania.\2813\
---------------------------------------------------------------------------

    \2810\ https://www.nhtsa.gov/corporate-average-fuel-economy/safe; https://www.epa.gov/newsreleases/us-epa-and-dot-propose-fuel-economy-standards-my-2021-2026-vehicles. The Agencies subsequently 
published the NPRM in the Federal Register on August 24, 2018. 83 FR 
42986 (August 24, 2018).
    \2811\ 83 FR 42986 (August 24, 2018).
    \2812\ 83 FR 42817 (August 24, 2018).
    \2813\ Id.
---------------------------------------------------------------------------

    The agencies also received several comments with respect to the 
sufficiency of the agencies' public hearings during the comment period. 
For example, the South Coast Air Quality Management District asserted 
that EPA failed to meet its obligation to hold public hearings under 
the Clean Air Act, claiming that an EPA ``political appointee'' did not 
have the legal authority to change hearing locations.\2814\ The comment 
also claimed that holding certain of the hearings in smaller 
metropolitan areas than originally announced resulted in 15 million 
fewer potential participants in the hearings.\2815\ Additionally, the 
comment noted that the NPRM and the notice that set the new locations 
of two of the public hearings were both published in the Federal 
Register on the same day, yet those documents contained conflicting 
hearing locations (the NPRM listed the originally planned hearing 
locations).\2816\
---------------------------------------------------------------------------

    \2814\ See comments from the South Coast Air Quality Management 
District, Attachment 1--SCAQMD Combined NHTSA Waiver Comment (Oct. 
25, 2018), Docket No. NHTSA-2018-0067-11813, at 37-38.
    \2815\ See id. at 37.
    \2816\ See id.
---------------------------------------------------------------------------

    Similarly, seventeen States and the District of Columbia submitted 
a joint comment requesting that the agencies reinstate the hearing 
locations that were initially listed in the NPRM, with the stated goal 
of maximizing the number of public participants.\2817\ Similarly, a 
group of environmental organizations jointly submitted a comment 
stating that the new hearing locations failed to maximize the potential 
participants for the agencies' public hearings.\2818\ That group also 
asserted that the agencies failed to provide a reason for the agencies' 
denial of requests to hold more than three public hearings.\2819\
---------------------------------------------------------------------------

    \2817\ See comments from the State of California et al., Request 
for an extension, Docket No. NHTSA-2018-0067-3458.
    \2818\ See comments from the Center for Biological Diversity, 
Conservation Law Foundation, Environmental Defense Fund, 
Earthjustice, Environmental Law and Policy Center, Natural Resources 
Defense Council, Public Citizen, Inc., Sierra Club, and Union of 
Concerned Scientists, Appendix A--Coalition Comment Letter (10-26-
2018), Docket No. NHTSA-2018-0067-12000, at 213. A number of other 
commenters also requested that the Agencies hold additional public 
hearings. See, e.g., comments from the Georgetown Climate Center, 
20180906--GCC Comments to NHTSA and EPA, Docket No. NHTSA-2018-0067-
3610; The City of Los Angeles, Docket No. NHTSA-2018-0067-4159, at 
2-3; California Air Resources Board, 2018-09-11 SAFE Rule DEIS--CARB 
Req Add Info, Docket No. NHTSA-2018-0067-4166, at 1; Northeast 
States for Coordinated Air Use Management, NESCAUM SAFE rule request 
for comment extension and hearing_20180824, Docket No. NHTSA-2018-
0067-2158, at 1-2.
    \2819\ Id.
---------------------------------------------------------------------------

    The agencies more than satisfied their legal obligation with 
respect to holding public hearings, and the three hearings provided 
substantial additional opportunity for public participation. While the 
agencies understand that some commenters were disappointed with some 
aspects of the process, those commenters did not demonstrate that the 
agencies' process was legally deficient, nor that any party suffered 
prejudice from the changes the agencies made to their public hearing 
arrangement.

[[Page 25156]]

    The APA does not require agencies to hold public hearings during 
the rulemaking process, unless the opportunity for a public hearing is 
required by a governing statute.\2820\ NHTSA's governing fuel economy 
statute does not require a public hearing during the rulemaking 
process.\2821\ The Clean Air Act requires EPA to ``give interested 
persons an opportunity for the oral presentation of data, views, or 
arguments, in addition to an opportunity to make written submissions . 
. . .'' 42 U.S.C. 7607(d)(5)(ii). The agencies' three joint public 
hearings satisfied this statutory requirement.
---------------------------------------------------------------------------

    \2820\ See 5 U.S.C. 553(c). Absent a statutory requirement, the 
APA gives agencies the discretion whether or not to hold a public 
hearing, stating that ``the agency shall give interested persons an 
opportunity to participate in the rule making through submission of 
written data, views, or arguments with or without opportunity for 
oral presentation.'' Id.
    \2821\ See 49 U.S.C. 32902.
---------------------------------------------------------------------------

    The agencies note that it was clear from the NPRM that the hearings 
were not yet finalized. No addresses or dates were announced for the 
hearings, and the NPRM indicated that information on the hearings would 
be forthcoming in a supplemental Federal Register notice. The NPRM 
(signed by the EPA Administrator) indicated that three hearings would 
be held, and the fact that specific details about those hearings were 
announced in a later notice signed by a different political appointee 
does not itself make the hearings themselves invalid. The Clean Air Act 
does not mandate hearings in any particular location and the public was 
aware from the NPRM that additional information on the hearings would 
be forthcoming. To the extent that any individual person or group was 
inconvenienced by the change in location announced in the supplemental 
notice, they still had ample time to submit public comments through any 
of the multiple other available methods indicated in the NPRM.\2822\
---------------------------------------------------------------------------

    \2822\ Executive Order 13,563 offers guidance to agencies with 
respect to how to maximize public participation. The Executive Order 
states that agencies should ``afford the public a meaningful 
opportunity to comment through the internet on any proposed 
regulation . . . .'' The vast majority of the comments the agencies 
received in this rulemaking were submitted through the internet.
---------------------------------------------------------------------------

    The agencies regret any confusion that resulted from publication of 
the NPRM in the Federal Register on the same date as publication of the 
notice that updated the hearing locations and provided additional 
information, including hearing dates. However, because the NPRM did not 
include dates for the hearings, and the NPRM informed interested 
parties to look for an additional notice that would announce specific 
dates and addresses for the hearings, no one could have relied on the 
NPRM to the exclusion of the supplemental notice.\2823\
---------------------------------------------------------------------------

    \2823\ Additionally, as a matter of fairness, the agencies gave 
interested parties notice about the change in public hearing 
locations one month prior to the first public hearing. See 83 FR 
42817 (August 24, 2018).
---------------------------------------------------------------------------

    The agencies ultimately held three public hearings, as was 
originally announced. There is no Clean Air Act requirement for a 
particular number of hearings, and by holding the hearings in locations 
throughout the United States (including in California), the agencies 
offered a meaningful opportunity for participation. Moreover, the 
public docket remained open for two months subsequent to the 
announcement of the final hearing locations, providing any interested 
party who was unable to attend a public hearing ample opportunity to 
submit comments in writing. As evidence of this meaningful opportunity 
to comment on the proposed rule, the agencies received a total of more 
than 750,000 public comments.
    Several commenters also asserted that the agencies delayed posting 
the hearing transcripts to the public docket until October 25, which 
was one day before the close of the public comment period.\2824\ The 
Environmental Defense Fund claimed that this was inconsistent with the 
Clean Air Act's requirements that ```[t]he transcript of public 
hearings, if any, on the proposed rule shall also be included in the 
docket promptly upon receipt from the person who transcribed such 
hearings.' 42 U.S.C. 7607(d)(4)(B).'' \2825\ As one commenter 
acknowledged, the transcripts were certified by the reporters on 
September 26, 2018 (Pittsburgh hearing), September 27, 2018 (Dearborn 
hearing), and October 1, 2018 (Fresno hearing).\2826\ The agencies made 
the transcripts publicly available within a reasonable period. 
Moreover, it was reasonable for the agencies to have an opportunity to 
review the transcripts for errors prior to making them publicly 
available. While the concern expressed by these commenters was an 
inadequate ability to offer responsive comments to the transcripts, the 
rulemaking process would be never-ending if every commenter had an 
opportunity to respond to every other commenter. There is no such 
requirement in the APA, the Clean Air Act, or otherwise. The public had 
sufficient opportunity to comment on the agencies' proposals, as 
described above.
---------------------------------------------------------------------------

    \2824\ Environmental Defense Fund, NHTSA-2018-0067-12108, NHTSA-
2018-0067-12327, NHTSA-2018-0067-12371; State of California et al., 
NHTSA-2018-0067-11735.
    \2825\ Environmental Defense Fund, NHTSA-2018-0067-12371.
    \2826\ State of California et al., NHTSA-2018-0067-11735.
---------------------------------------------------------------------------

    A few commenters requested that the agencies host a workshop or 
webinar to help commenters better understand the agencies' modeling and 
analyses.\2827\ The commenters pointed to similar activities undertaken 
by EPA for other complex rulemakings. While the agencies did not 
conduct a live workshop or webinar regarding the proposal, they did 
make extensive information publicly available beyond the contents of 
the NPRM. To assist the public, NHTSA hosted a dedicated web page with 
information on the modeling.\2828\ The web page included a video 
introduction to the CAFE model.\2829\ The web page enabled members of 
the public to download the model software, its system documentation, 
source code, and input files.\2830\ Many commenters commented in detail 
on the modeling and analyses. However, the agencies recognize that 
public stakeholders vary in their experience and understanding of the 
modeling and analyses and will continue to consider ways to facilitate 
public participation in future rulemakings, which could include the use 
of workshops or webinars.
---------------------------------------------------------------------------

    \2827\ See Minnesota Pollution Control Agency (MPCA), NHTSA-
2017-0069-0528; Minnesota Pollution Control Agency (MPCA) et al., 
NHTSA-2018-0067-11706.
    \2828\ https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
    \2829\ Id.
    \2830\ Id.
---------------------------------------------------------------------------

    Some comments criticized the agencies for the agencies' 
untimeliness in adding materials to the rulemaking dockets, for 
example, identifying material ``that was not added to the rulemaking 
docket until the end of the original comment period or, in some cases, 
added either after that period already had closed or not at all.'' 
\2831\
---------------------------------------------------------------------------

    \2831\ CBD et. al, Supplemental Comments, Docket No. NHTSA-2018-
0067-12371, at 8.
---------------------------------------------------------------------------

    The critical question is ``whether the final rule changes 
critically from the proposed rule rather than on whether the agency 
relies on supporting material not published for comment.'' \2832\ In 
other words, ``[t]he question is typically whether the agency's final 
rule so departs from its proposed rule as to constitute more surprise 
than notice.'' \2833\ To that end, agencies are allowed--as the 
agencies here did--to

[[Page 25157]]

rely on supplemental data that clarified, expanded on, or confirmed 
information in the proposed rule, even if that supplemental data was 
not disclosed in the proposed rule.\2834\ In any event, the commenters 
have failed to show how they were prejudiced by any information posted 
later than they would have preferred.\2835\
---------------------------------------------------------------------------

    \2832\ Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7 (D.C. 
Cir. 1999).
    \2833\ Id. (citing Air Transp. Ass'n of Am., 732 F.2d 219, 225 
n.12 (D.C. Cir. 1984)).
    \2834\ See Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7 
(D.C. Cir. 1999) (citing Solite Corp. v. EPA, 952 F.2d 473, 485 
(D.C. Cir. 1991); Air Transp. Ass'n of Am. v. CAB, 732 F.2d 219, 224 
(D.C. Cir. 1984)).
    \2835\ See Solite Corp. v. U.S. E.P.A., 952 F.2d 473, 484 (D.C. 
Cir. 1991) (citing Cmty. Nutrition Inst. v. Block, 749 F.2d 50, 57-
58 (D.C. Cir. 1984)). Parties also could have submitted comments 
after the end of the comment period on any of these materials. See 
49 CFR 553.23 (NHTSA regulation providing that ``[l]ate filed 
comments will be considered to the extent practicable.'').
---------------------------------------------------------------------------

    Some commenters noted that certain aspects of the CAFE model used 
for the proposal were not previously subject to peer review.\2836\ 
Certain commenters asserted that the proposal was legally flawed 
because the full CAFE model was not peer reviewed prior to the 
proposal.\2837\ In support of this argument, commenters cited the 
Information Quality Act and related OMB guidance that states that 
``each agency shall have a peer review conducted on all influential 
scientific information that the agency intends to disseminate.'' \2838\ 
Commenters also cited EPA's Peer Review Handbook, which states: ``For 
highly influential scientific assessments, external peer review is the 
expected procedure.'' \2839\
---------------------------------------------------------------------------

    \2836\ See, e.g., Center for Biological Diversity et al., NHTSA-
2018-0067-12000; Environmental Defense Fund, NHTSA-2018-0067-12327; 
Environmental Defense Fund et al., NHTSA-2018-0067-12371; 
Environmental Defense Fund et al., NHTSA-2018-0067-12406; Center for 
Biological Diversity, Environment America, Environmental Defense 
Fund, Environmental Law Policy Center, Public Citizen, Inc., Sierra 
Club, and Union of Concerned Scientists, NHTSA-2018-0067-12439; 
States of California et al., NHTSA-2018-0067-11735.
    \2837\ See, e.g., Center for Biological Diversity et al., NHTSA-
2018-0067-12000.
    \2838\ See Center for Biological Diversity et al., NHTSA-2018-
0067-12000.
    \2839\ See Center for Biological Diversity et al., NHTSA-2018-
0067-12000.
---------------------------------------------------------------------------

    The agencies agree that peer review is appropriate for the CAFE 
model, and the CAFE model has been peer reviewed. As discussed in the 
NPRM, and as certain commenters acknowledged, the CAFE model was peer 
reviewed in 2017.\2840\ NHTSA included peer review materials in the 
public docket as well as on its web page regarding the model.\2841\ As 
described in those materials: ``In 2017, the Volpe Center arranged for 
a formal peer review of the version of the CAFE model released and 
documented in 2016 . . . . All of the peer reviewers supported much 
about the model's general approach, and supported many of the model's 
specific characteristics. Peer reviewers also provided a variety of 
general and specific recommendations regarding potential changes to the 
model, inputs, outputs, and documentation. NHTSA and Volpe Center staff 
agree with many of these recommendations and have either completed or 
begun work to implement many of them; implementing others would require 
further research, testing, and development not possible at this time, 
but we are considering them for future model versions.''\2842\
---------------------------------------------------------------------------

    \2840\ 83 FR 43000 (Aug. 24, 2018) (``A report available in the 
docket for this rulemaking presents peer reviewers' detailed 
comments and recommendations, and provides DOT's detailed 
responses.''); see Center for Biological Diversity et al., NHTSA-
2018-0067-12000.
    \2841\ NHTSA-2018-0067-0055; https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
    \2842\ NHTSA-2018-0067-0055.
---------------------------------------------------------------------------

    However, certain new elements of the CAFE model were not completed 
at the time of the 2017 peer review.\2843\ NHTSA subsequently obtained 
a peer review of significant new elements added to the model after the 
2017 peer review.\2844\ As described in the new peer review charge, 
included in a July 2019 report included in the rulemaking docket, NHTSA 
explained:
---------------------------------------------------------------------------

    \2843\ NHTSA-2018-0067-0055 (explaining, in responses to 2017 
peer review, that ``[t]he model has been updated to including 
procedures to estimate impacts on new vehicle sales, and on older 
vehicle scrappage'').
    \2844\ NHTSA-2018-0067-0055.

    To inform the proposed rule announced in August 2018, DOT staff 
introduced significant new elements to the model, including methods 
to estimate changes in vehicle sales volumes, vehicle scrappage, and 
automotive sector labor usage. Each of these regulatory actions 
involved consideration of and response to significant public comment 
on model results, as well as comments on the model itself. In 
addition to DOT staff's own observations, these comments led DOT 
staff to make a wide range of improvements to the model. Insofar as 
a formal peer review could identify additional potential 
opportunities to improve the model, DOT sponsored a review of the 
entire model in 2017. At this time, DOT seeks review of some of the 
---------------------------------------------------------------------------
significant new elements added to the model after that review.

    This subsequent peer review of the new elements was not complete at 
the time the proposal was published, and therefore materials concerning 
the peer reviewers' comments and NHTSA's responses were not available 
until later.\2845\ Although the comment period on the proposal had 
closed at that time, the agencies continued to receive comments on the 
new peer review materials, which they have considered in issuing this 
final rule.\2846\ Of course, the new elements of the modeling were also 
described in detail in the NPRM and commenters also directly commented 
on them in great detail. Thus, the public was fully apprised of all 
aspects of the modeling and had a robust opportunity to provide 
comment.
---------------------------------------------------------------------------

    \2845\ NHTSA-2018-0067-0055 (July 2019 report).
    \2846\ See, e.g., Center for Biological Diversity et al., NHTSA-
2018-0067-12439; Environment America et al., NHTSA-2018-0067-12441.
---------------------------------------------------------------------------

    To the extent commenters are suggesting the Information Quality Act 
required a full peer review of all aspects of the CAFE model prior to 
the proposal, the agencies disagree.\2847\ Peer review of the new 
elements of the CAFE model helped ensure that the model is 
scientifically sound, and the peer reviewers provided feedback that 
helped improve the model and may help develop additional improvements 
to the model in the future. In this sense, the peer review of the new 
elements of the model functioned similarly to public comments from 
commenters with specific scientific expertise. Much of the feedback 
from the peer reviewers were in fact similar in nature to comments 
received from public commenters on the model. By engaging in both peer 
review and notice-and-comment procedures, the agencies ensured that 
they had information from a wide variety of sources, including those 
with specific expertise, to validate and improve the model.\2848\ The 
technical aspects of the model, including improvements made to the 
model following the proposal, are described in detail in this final 
rule. Moreover, as the Center for Biological Diversity noted, the 
Information Quality Act does not create third-party rights.\2849\
---------------------------------------------------------------------------

    \2847\ See, e.g., Center for Biological Diversity et al., NHTSA-
2018-0067-12000; Environment America et al., NHTSA2018-0067-12441.
    \2848\ The timing of the peer review of new elements of the 
model also did not require a second cycle of notice and comment. 
See, e.g., Alto Dairy v. Veneman, 336 F.3d 560, 569-70 (7th Cir. 
2003) (``The law does not require that every alteration in a 
proposed rule be reissued for notice and comment. If that were the 
case, an agency could `learn from the comments on its proposals only 
at the peril of subjecting itself to rulemaking without end.''').
    \2849\ Center for Biological Diversity et al., NHTSA-2018-0067-
12000.
---------------------------------------------------------------------------

    The agencies also disagree that EPA needed to obtain a separate 
peer review of the CAFE model.\2850\ The peer review addressed aspects 
of the model relevant to the analysis by both agencies under their 
respective statutory schemes. The agencies have expertise in their

[[Page 25158]]

statutory requirements and discussed in detail both in the proposal and 
this final rule how the CAFE model was used to inform the decision-
making under both EPCA and the CAA.
---------------------------------------------------------------------------

    \2850\ Center for Biological Diversity et al., NHTSA-2018-0067-
12000.
---------------------------------------------------------------------------

(c) Other APA Comments
    Many commenters suggested that the record of evidence developed for 
the 2016 Draft TAR and EPA's Original Determination was a better basis 
for NHTSA to determine maximum feasible standards than the record of 
evidence for the current rulemaking. These commenters also argued that, 
in the NPRM, NHTSA ignored the findings and analysis in the TAR and the 
Technical Support Document and contradicted the pre-existing record 
without explanation. Lastly, these commenters argued that the NPRM did 
not have a reasoned basis under the APA, particularly in light of the 
agency's change in position and the reliance interests at stake.
    Agencies always have authority under the Administrative Procedure 
Act to revisit previous decisions in light of new facts, as long as 
they provide notice and an opportunity for comment--as the agencies did 
here. Indeed, it is the best practice to do so when changed 
circumstances so warrant.\2851\
---------------------------------------------------------------------------

    \2851\ See FCC v. Fox Television, 556 U.S. 502 (2009).
---------------------------------------------------------------------------

    ``Changing policy does not, on its own, trigger an especially 
`demanding burden of justification.' '' \2852\ ``Agencies are free to 
change their existing policies as long as they provide a reasoned 
explanation for the change.'' \2853\ Providing this explanation ``would 
ordinarily demand that [the agency] display awareness that it is 
changing position.'' \2854\ Beyond that, however, ``[w]hen an agency 
changes its existing position, it `need not always provide a more 
detailed justification than what would suffice for a new policy created 
on a blank slate.' '' \2855\ The agency ``need not demonstrate to a 
court's satisfaction that the reasons for the new policy are better 
than the reasons for the old one.'' \2856\ For instance, ``evolving 
notions'' about the appropriate balance of varying policy 
considerations constitute sufficiently good reasons for a change in 
position.\2857\ A change in policy is ``well within an agency's 
discretion:'' Agencies are permitted to conduct a ``reevaluation of 
which policy would be better in light of the facts,'' without 
``rely[ing] on new facts.'' \2858\
---------------------------------------------------------------------------

    \2852\ Mingo Logan Coal Co. v. Envtl. Prot. Agency, 829 F.3d 
710, 718 (DC Cir. 2016) (quoting Ark Initiative v. Tidwell, 816 F.3d 
119, 127 (DC Cir. 2016)).
    \2853\ Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 
(2016) (citations omitted).
    \2854\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 
(2009) (emphasis in original) (``An agency may not, for example, 
depart from a prior policy sub silentio or simply disregard rules 
that are still on the books.'').
    \2855\ Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125-
26 (2016) (quoting FCC v. Fox Television Stations, Inc., 556 U.S. 
502, 515 (2009)).
    \2856\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 
(2009) (emphasis in original).
    \2857\ N. Am.'s Bldg. Trades Unions v. Occupational Safety & 
Health Admin., 878 F.3d 271, 303 (D.C. Cir. 2017) (quoting the 
agency's rule).
    \2858\ Nat'l Ass'n of Home Builders v. E.P.A., 682 F.3d 1032, 
1037-38 (D.C. Cir. 2012).
---------------------------------------------------------------------------

    To be sure, providing ``a more detailed justification'' is 
appropriate in some cases.\2859\ But when ``a more detailed 
justification'' is needed, all that is required is for the agency to 
explain how ``new information arising after'' the previous 
determination ``informed its conclusion'' that a change was 
appropriate: ``Explanations relying on new data are sufficient to 
satisfy the more detailed explanatory obligation.'' \2860\ As one of 
the critical comments itself noted, ``[a]gencies must use `the best 
information available' in reaching their conclusions, and cannot 
lawfully rely on outdated information as circumstances change.'' \2861\ 
Accordingly, when new information became available, the agencies relied 
on it expressly, resulting in a fully-explained change in their 
analysis and ultimately their conclusions.
---------------------------------------------------------------------------

    \2859\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 
(2009) (``Sometimes [the agency] must [provide a more detailed 
justification than what would suffice for a new policy created on a 
blank slate]--when, for example, its new policy rests upon factual 
findings that contradict those which underlay its prior policy; or 
when its prior policy has engendered serious reliance interests that 
must be taken into account.'').
    \2860\ Mingo Logan Coal Co. v. Envtl. Prot. Agency, 829 F.3d 
710, 727 (D.C. Cir. 2016).
    \2861\ CBD et. al, Appendix A, Docket No. NHTSA-2018-0067-12000, 
at 11 (quoting Flyers Rights Education Fund v. FAA, 864 F. 3d 738, 
745 (D.C. Cir. 2017)).
---------------------------------------------------------------------------

    While ``[i]t would be arbitrary or capricious to ignore such 
matters,''\2862\ the agencies have not ignored them. NHTSA has 
satisfied these standards. The NPRM expressly and repeatedly 
acknowledged that it represented a change from the 2012 final rule, the 
Draft TAR, and EPA's Original Determination, appropriately justifying 
the change by citing shifts in policy priorities or new facts and 
changed circumstances that became apparent since the Original 
Determination.\2863\ The agencies are fully cognizant of the facts and 
circumstances that have changed since the Original Determination, 
expressly acknowledged them in the Revised Determination and SAFE Rule 
NPRM, and adapted to accept them now in the final rule.
---------------------------------------------------------------------------

    \2862\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 
(2009).
    \2863\ See, e.g., 83 FR at 43213 (Aug. 24, 2018).
---------------------------------------------------------------------------

    Several commenters invoked requests to the agencies under the 
Freedom of Information Act (``FOIA'') regarding material sought in 
connection with the rulemaking.\2864\ These comments ranged from simple 
references to existing FOIA requests to the agencies, to the actual 
submission of the FOIA requests as a comment posted to the rulemaking 
docket.\2865\ These commenters sought a variety of information, which 
included calendars and internal correspondence of specific agency 
personnel, communications with non-governmental stakeholders, and 
technical materials and clarifications relating to aspects of the 
agencies' analysis.\2866\
---------------------------------------------------------------------------

    \2864\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
12371.
    \2865\ Compare, e.g., Joint Submission from the States of 
California et al. and the Cities of Oakland et al., NHTSA NHTSA-
2018-0067-11735, with, e.g., Office of the Attorney General of the 
State of New York, NHTSA-2018-0067-3613.
    \2866\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
12397; Office of the Attorney General of the State of New York, 
NHTSA-2018-0067-3613; California Air Resources Board, NHTSA-2018-
0067-4166.
---------------------------------------------------------------------------

    To the extent these requests sought substantive material, those 
matters are addressed in other sections herein that pertain to the 
respective underlying issues implicated. Although the submission of 
FOIA requests through an online rulemaking docket is a very unusual 
form of submitting a FOIA request to an agency, the agencies 
nevertheless processed the comments that requested materials by 
invoking FOIA as formal FOIA requests. As such, once identified, those 
comments were forwarded to the agencies' respective FOIA offices, which 
commenced the intake process of the letters as FOIA requests. In turn, 
the agencies' FOIA offices transmitted receipt acknowledgement letters 
to the requestors and conducted searches for the applicable material. 
The agencies responded to the requestors by producing the responsive 
non-exempt records identified, applying the appropriate FOIA standards 
applicable to the records and requests. Like all other typical FOIA 
requests, the requestors were provided with an opportunity to 
administratively appeal the FOIA decision and, if desired, subsequently 
seek judicial review of the agencies' decisions. Several commenters 
availed themselves of this procedure.\2867\
---------------------------------------------------------------------------

    \2867\ See generally, e.g., New York v. U.S. Envtl. Prot. Agency 
and Nat'l Highway Traffic Safety Admin., Case No. 1:19-cv-00712 
(S.D.N.Y.) (FOIA litigation concerning a FOIA request submitted as a 
comment from the Office of the Attorney General of the State of New 
York, NHTSA-2018-0067-3613).

---------------------------------------------------------------------------

[[Page 25159]]

    Thus, the agencies fully satisfied their obligations under the 
governing FOIA provisions. In fact, other commenters noted the 
agencies' responses to these FOIA requests and incorporated information 
disclosed in the responses into their comments.\2868\ Moreover, several 
of the FOIA requests submitted as comments requested information that 
had already been published on the agencies' websites for the rulemaking 
or in the rulemaking dockets.
---------------------------------------------------------------------------

    \2868\ See James H. Stock, Kenneth Gillingham & Wade Davis, EPA-
HQ-OAR-2018-0283-6220, at p. 6.
---------------------------------------------------------------------------

    Although the agencies fulfilled their obligations under all 
applicable FOIA law, the agencies also stress that FOIA compliance is 
wholly irrelevant to conformity to governing APA standards in the 
rulemaking process. FOIA arises from an independent statutory 
framework, which contains unique provisions for judicial review.\2869\ 
These provisions for judicial review provide ``an adequate form of 
relief'' such that the APA is not typically even an appropriate 
mechanism to seek the disclosure of further information requested under 
FOIA.\2870\ Likewise, the APA's principles governing rulemaking 
procedures, including disclosures of information for such rulemakings, 
exist as autonomous statutory and jurisprudential concepts totally 
untethered from the principles of disclosure under FOIA.
---------------------------------------------------------------------------

    \2869\ 5 U.S.C. 552(a)(4)(B).
    \2870\ See, e.g., Feinman v. FBI, 713 F. Supp. 2d 70, 76 (D.D.C. 
2010) (``This court and others have uniformly declined jurisdiction 
over APA claims that sought remedies made available by FOIA.'').
---------------------------------------------------------------------------

    Similarly, as an independent statutory framework from the APA, the 
susceptibility of materials and records for production under FOIA has 
no bearing on whether such materials should have been made public under 
the APA as part of a rulemaking. The scope of materials for production 
under FOIA arises from the Agency's reasonable interpretation of the 
language of the FOIA request, as well as the exemptions potentially 
applicable to the records under the applicable FOIA statutes and 
implementing regulations.\2871\ In contrast, in an APA review of 
rulemaking procedures, separate standards exist to govern the scope of 
materials an agency must make available during the rulemaking 
process.\2872\ Thus, records may be responsive to a FOIA request, but 
not appropriate for publication under the APA--even if the FOIA request 
concerns the proposed rule in question. The FOIA requests at issue here 
are illustrative of this distinction. For example, one of the specific 
FOIA requests identified by commenters describes the requests as 
pertaining to the NPRM, but seeks Outlook calendars of DOT and NHTSA 
personnel.\2873\ While such materials may be responsive to the 
underlying FOIA requests, which expressly mention the calendars, an 
employee's entire list of calendar appointments--including appointments 
unrelated to the rulemaking--is clearly not contemplated by the APA as 
material necessary for publication along with a proposed rule. Thus, 
while the agencies sought to comply with their independent statutory 
obligations under FOIA, to the extent commenters invoke purported FOIA 
noncompliance, the agencies consider such arguments irrelevant to the 
rulemaking analysis. Likewise, any production of records in connection 
with any FOIA request that invokes the proposed rule is not a 
recognition by the agencies that the material should have also been 
made available during the rulemaking under the APA.
---------------------------------------------------------------------------

    \2871\ See 5 U.S.C. 552. See also, e.g., Weisberg v. U.S. Dep't 
of Justice, 745 F.2d 1476, 1485 (DC Cir. 1984) (discussing standards 
applicable to the scope of an Agency's search for records under 
FOIA).
    \2872\ See Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7 (DC 
Cir. 1999) (discussing the scope of materials for an agency to make 
available during a notice and comment period).
    \2873\ See Environmental Defense Fund, NHTSA-2018-0067-12397.
---------------------------------------------------------------------------

    Several commenters also criticized the agencies, and specifically 
the EPA, for not publishing an updated version of the Optimization 
Model for Reducing Emissions of Greenhouse Gases from Automobiles 
(``OMEGA'') along with the proposed rule.\2874\ As described in further 
detail in Section IV herein, OMEGA is a fleet compliance model 
developed by the EPA and used in previous rulemakings. While many 
commenters raised technical arguments comparing the OMEGA model to the 
CAFE Model utilized in this rulemaking, such technical analysis and 
comments are addressed elsewhere in this final rule analysis. See 
Section IV. Likewise, while several comments refer to FOIA requests for 
OMEGA model materials, the Agencies' discussion of FOIA comments are 
addressed above.
---------------------------------------------------------------------------

    \2874\ See, e.g., International Council on Clean Transportation, 
NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    Most other commenters who raised more procedural arguments 
concerning the unavailability of an updated version of the OMEGA model 
argued that an updated version of the model should have been released 
because the EPA utilized the model during an interagency review of the 
proposed rule.\2875\ In considering these comments, the agencies 
emphasize that neither NHTSA, the EPA, nor any other interagency 
reviewer relied upon the OMEGA model for the preparation of either the 
proposed or the final versions of the SAFE Vehicles Rule. Instead, as 
clearly expressed in rulemaking descriptions and documents accompanying 
both this final rule and the proposed rule, the agencies relied on a 
separate model to perform the analysis that helped to inform the 
agencies regarding potential effects of various fuel economy standards. 
This independent model, the CAFE Model, was developed by the Department 
of Transportation's Volpe National Transportation Systems Center.
---------------------------------------------------------------------------

    \2875\ See, e.g., Sallie E. Davis, NHTSA-2018-0067-12430.
---------------------------------------------------------------------------

    In fact, most commenters discussing the OMEGA model understood and 
expressly acknowledged that the agencies relied upon the CAFE Model 
rather than the OMEGA model for this rulemaking.\2876\ Several 
commenters even paradoxically argued both that the agencies 
unreasonably failed to utilize the OMEGA model and that the agencies 
denied meaningful opportunity for comment by utilizing but failing to 
publish an updated OMEGA model.\2877\ Nevertheless, the analysis and 
universe of documents published for the proposed rule made abundantly 
clear that the CAFE Model--not the OMEGA model--performed the 
applicable analysis for this rulemaking. Likewise, the agencies' 
proposed rule published voluminous analyses and supporting documents to 
describe the CAFE Model and explain the underlying methodologies 
incorporated into the model's operation for this rulemaking. The 
agencies also released the full version of the CAFE Model employed in 
this rulemaking, as well as its respective inputs and outputs, in order 
to provide commenters with ample opportunities to understand the 
model's function and operation.
---------------------------------------------------------------------------

    \2876\ See, e.g., Union of Concerned Scientists, NHTSA-2018-
0067-12303-016; Center for Biological Diversity, NHTSA-2018-0067-
12000.
    \2877\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
12108.
---------------------------------------------------------------------------

    The extensive comments on the modeling conducted for this 
rulemaking confirm that the agencies provided the public with 
sufficient information to comment on the modeling process for the 
rulemaking. Comments regarding the OMEGA and CAFE models were 
expansive, spanning hundreds of pages of technical analysis and 
submissions from a variety of commenters. Many of these comments even 
consisted of detailed and technical comparisons of

[[Page 25160]]

the CAFE model used in this rulemaking with past versions of OMEGA 
models used for prior rulemakings.\2878\ Even if certain of these 
commenters disagreed with the Agencies' ultimate approach to the 
modeling, they evidently understood the applicable methodologies and 
performance of the CAFE Model for this rulemaking sufficiently to 
substantively engage with the Agencies on these topics through their 
comments. Therefore, the agencies consider the detailed comments on the 
OMEGA and CAFE models as clear indicia that the extensive information, 
materials, and explanations provided by the agencies in the proposed 
rule enabled significant opportunity for the public to comment on the 
modeling for the rule.
---------------------------------------------------------------------------

    \2878\ See, e.g., California Air Resources Board, NHTSA-2018-
0067-11873; Union of Concerned Scientists, NHTSA-2018-0067-12039; 
Alliance of Automobile Manufacturers, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    To the extent that commenters allege an insufficient opportunity to 
comment by claiming that the EPA actually utilized the OMEGA model in 
the rulemaking process, the agencies consider such comments 
unfounded.\2879\ The agencies did not rely on the OMEGA model during 
the rulemaking process, including during the analysis for the proposed 
and final rules. In past rulemakings, the EPA developed a complete 
final version of the OMEGA model to perform the rulemaking analysis. 
Here, the EPA did not even finalize a completed updated version of the 
OMEGA model, much less rely on such a model in the course of the 
rulemaking. Therefore, no completed version of an updated OMEGA model 
even existed for the agencies to publish as part of the notice of 
proposed rulemaking.
---------------------------------------------------------------------------

    \2879\ See, e.g., Center for Biological Diversity, NHTSA-2018-
0067-12000.
---------------------------------------------------------------------------

    To the extent commenters argue that the EPA should have updated the 
model for this rulemaking, the APA's facilitation of a meaningful 
opportunity to comment neither requires nor contemplates a mandate that 
the agencies develop computational modeling alternatives for the 
public, which were not even incorporated into the agencies' own 
rulemaking analysis.\2880\ In fact, doing so would actually detract 
from the notice and comment process because it would convolute the 
rulemaking docket and inhibit the public's ability to identify the 
modeling materials actually used in the rulemaking process. Thus, such 
extraneous materials would only dilute the rulemaking docket with 
voluminous and complex materials, such as modeling files, input files, 
and statistical figures, that had no influence on the rulemaking in 
question. Indeed, several commenters already claimed that the 
voluminous and complex supporting materials in the rulemaking docket 
required significant time for review, so the introduction of extensive 
totally extraneous material would have been only counterproductive to 
the process.\2881\
---------------------------------------------------------------------------

    \2880\ See, e.g., Center for Biological Diversity et al., NHTSA-
2018-0067-12000.
    \2881\ See, e.g., Institute for Policy Integrity, NHTSA-2018-
0067-5641; Northeast States for Coordinated Air Use Management, 
NHTSA-2018-0067-2158.
---------------------------------------------------------------------------

    Moreover, requiring the EPA to perform the work necessary to fully 
update the OMEGA model solely for a public release--when it did not 
otherwise intend to consider the model in the rulemaking--would divert 
valuable and finite agency resources away from actual rulemaking 
analyses in favor of efforts that further no progress in the 
rulemaking.\2882\ Such an approach would detract from the agencies' 
opportunities to devote time to other considerations that actually 
influenced the rulemaking, such as the substantive analysis 
incorporated into the proposed rule and the drafting of extensive 
language to explain to the public the methodologies applied by the 
agencies for the proposal. Such an inefficient allocation of resources 
undermines both the rulemaking process envisioned by the APA and the 
very notice and comment procedures utilized by these commenters.
---------------------------------------------------------------------------

    \2882\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
12108.
---------------------------------------------------------------------------

    Several commenters also argued that even if the agencies did not 
rely on the model for this rulemaking, the OMEGA model still informed 
the EPA's analysis and interagency review by providing general 
background experience in regulating greenhouse gas emissions--either 
through the agency's work with prior versions of the model or ongoing 
efforts to update the OMEGA model for purposes unrelated to this 
rulemaking. However, even assuming the model provided background 
experience to the EPA in regulating in this arena, federal 
jurisprudence makes clear that ``[t]he Administrative Procedure Act 
does not require that every bit of background information used by an 
administrative agency be published for public comment.'' See B. F. 
Goodrich Co. v. Dep't of Transp., 541 F.2d 1178, 1184 (6th Cir. 1976). 
This is particularly the case when, as here, ``[t]he basic data upon 
which the agency relied in formulating the regulation was available . . 
. for comment.'' Id.; see also Am. Min. Cong. v. Marshall, 671 F.2d 
1251, 1261 (10th Cir. 1982) (``These documents consist of background 
information and data as well as several internal memoranda. There is 
nothing to indicate that the Secretary actually relied on any of these 
documents in promulgating the rule or that the data they contain was 
critical to the formulation of the rule.''). In fact, publishing such 
background information not only exceeds the requirements of the APA, 
but would actually affirmatively undermine the APA's notice and comment 
procedure. If every piece of information ever referenced by the 
agencies or upon which the Agencies drew regulatory experience were 
required to be published, rulemaking dockets would expand to an absurd 
scope of nearly infinite materials, spanning arguably back to even the 
school textbooks the rulemaking personnel used to learn the underlying 
disciplines employed in the rulemaking analysis. Clearly such a scope 
would frustrate rather than further the provision of proper notice to 
the public about a proposed rule.\2883\
---------------------------------------------------------------------------

    \2883\ To the extent commenters seek to understand the manner in 
which the OMEGA model informed prior rulemaking efforts, the EPA has 
released the full versions of prior OMEGA models and applicable 
materials along with the prior rulemakings. In fact, several 
commenters referenced such materials in submitting detailed comments 
comparing the CAFE Model with the OMEGA model. Manufacturers of 
Emission Controls Association, NHTSA-2018-0067-11994. Thus, any 
commenters that were interested in such extraneous background 
information had ample opportunity to access the material.
---------------------------------------------------------------------------

    Moreover, even assuming the premise of several commenters' 
challenges--that the EPA consulted updates to the OMEGA model during 
the interagency review--such a predicate still would not require the 
publication of the model during the rulemaking process.\2884\ As the 
agencies have made clear, the OMEGA model did not affect any part of 
the rule, including the methodologies and analysis underlying the 
formulation of the rule. Therefore, even if consulted, the OMEGA model 
would exist as, at most, supplementary material which had no influence 
on the rulemaking methodologies, all of which were fully disclosed. 
See, e.g., Chamber of Commerce of U.S. v. SEC., 443 F.3d 890, 900 (DC 
Cir. 2006) (``When the agency relies on supplementary evidence without 
a showing of prejudice by an interested party, the procedural 
requirements of the APA are satisfied without further opportunity for 
comment, provided that the agency's response constitutes a logical 
outgrowth

[[Page 25161]]

of the rule initially proposed'') (internal citations omitted).
---------------------------------------------------------------------------

    \2884\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
12406.
---------------------------------------------------------------------------

3. National Environmental Policy Act
    As discussed above, EPCA requires NHTSA to determine the level at 
which to set CAFE standards for each model year by considering the four 
factors of technological feasibility, economic practicability, the 
effect of other motor vehicle standards of the Government on fuel 
economy, and the need of the United States to conserve energy. The 
National Environmental Policy Act (NEPA) directs that environmental 
considerations be integrated into that process.\2885\ To explore the 
potential environmental consequences of this rulemaking action, NHTSA 
prepared a Draft Environmental Impact Statement (``DEIS'') for the NPRM 
and a Final Environmental Impact Statement (``FEIS'') for the final 
rule. The purpose of an EIS is to ``provide full and fair discussion of 
significant environmental impacts and [to] inform decisionmakers and 
the public of the reasonable alternatives which would avoid or minimize 
adverse impacts or enhance the quality of the human environment.'' 
\2886\
---------------------------------------------------------------------------

    \2885\ NEPA is codified at 42 U.S.C. 4321-47. The Council on 
Environmental Quality (CEQ) NEPA implementing regulations are 
codified at 40 CFR parts 1500-08.
    \2886\ 40 CFR 1502.1.
---------------------------------------------------------------------------

    As explained in the NPRM, NEPA is ``a procedural statute that 
mandates a process rather than a particular result.'' \2887\ The 
agency's overall EIS-related obligation is to ``take a `hard look' at 
the environmental consequences before taking a major action.'' \2888\ 
Significantly, ``[i]f the adverse environmental effects of the proposed 
action are adequately identified and evaluated, the agency is not 
constrained by NEPA from deciding that other values outweigh the 
environmental costs.'' \2889\ The agency must identify the 
``environmentally preferable'' alternative but need not adopt it.\2890\ 
``Congress in enacting NEPA . . . did not require agencies to elevate 
environmental concerns over other appropriate considerations.'' \2891\ 
Instead, NEPA requires an agency to develop and consider alternatives 
to the proposed action in preparing an EIS.\2892\ The statute and 
implementing regulations do not command the agency to favor an 
environmentally preferable course of action, only that it make its 
decision to proceed with the action after taking a hard look at the 
potential environmental consequences and consider the relevant factors 
in making a decision among alternatives.\2893\
---------------------------------------------------------------------------

    \2887\ Stewart Park & Reserve Coal., Inc. v. Slater, 352 F.3d 
545, 557 (2d Cir. 2003).
    \2888\ Baltimore Gas & Elec. Co. v. Natural Resources Defense 
Council, Inc., 462 U.S. 87, 97 (1983).
    \2889\ Robertson v. Methow Valley Citizens Council, 490 U.S. 
332, 350 (1989).
    \2890\ 40 CFR 1505.2(b).
    \2891\ Baltimore Gas, 462 U.S. at 97.
    \2892\ 42 U.S.C. 4332(2)(C)(iii).
    \2893\ 40 CFR 1505.2(b).
---------------------------------------------------------------------------

    NHTSA received many comments on the DEIS. Among the comments 
received, many commenters stated that the baseline/no-action standards 
were the environmentally preferable alternative and argued that the 
environmental benefits of the proposal were (1) insufficient and/or (2) 
incorrectly assessed in a variety of ways. Comments regarding the 
environmental analyses presented in this preamble are addressed in 
Section VI above, while those regarding the DEIS are addressed in 
Chapter 10 of the FEIS.
    When preparing an EIS, NEPA requires an agency to compare the 
potential environmental impacts of its proposed action and a reasonable 
range of alternatives. In the DEIS, NHTSA analyzed a No Action 
Alternative and eight action alternatives. In the FEIS, NHTSA analyzed 
the same No Action Alternative and seven action alternatives, including 
a new alternative (the Preferred Alternative) within the range of the 
alternatives considered in the DEIS and FEIS.\2894\ The alternatives 
represent a range of potential actions the agency could take, and they 
are described more fully in Section V above, below in this section, and 
Chapter 2 of the FEIS. The environmental impacts of these alternatives, 
in turn, represent a range of potential environmental impacts that 
could result from NHTSA's setting maximum feasible fuel economy 
standards for passenger cars and light trucks.
---------------------------------------------------------------------------

    \2894\ In its scoping notice, NHTSA indicated that the action 
alternatives analyzed would bracket a range of reasonable annual 
fuel economy standards, allowing the agency to select an action 
alternative in its final rule from any stringency level within that 
range. 82 FR 34740, 34743 (July 26, 2017).
---------------------------------------------------------------------------

    To derive the direct and indirect impacts of the action 
alternatives, NHTSA compared each action alternative to the No Action 
Alternative, which reflects baseline trends that would be expected in 
the absence of any further regulatory action other than finalizing the 
augural standards. More specifically, the No Action Alternative in the 
DEIS and FEIS assumed that NHTSA would not amend the CAFE standards for 
MY 2021 passenger cars and light trucks. In addition, the No Action 
Alternative assumed that NHTSA would finalize the MY 2022-2025 augural 
CAFE standards that were described in the 2012 final rule. Finally, for 
purposes of its analysis, NHTSA assumed that the MY 2025 augural 
standards would continue indefinitely. The augural standards also serve 
as a proxy for EPA's CO2 standards for MYs 2022-2025, which 
were also finalized in the 2012 final rule. The No Action Alternative 
provides an analytical baseline against which to compare the 
environmental impacts of other alternatives presented in the EIS.\2895\
---------------------------------------------------------------------------

    \2895\ See 40 CFR 1502.2(e), 1502.14(d). CEQ has explained that 
``[T]he regulations require the analysis of the no action 
alternative even if the agency is under a court order or legislative 
command to act. This analysis provides a benchmark, enabling 
decision makers to compare the magnitude of environmental effects of 
the action alternatives [See 40 CFR 1502.14(c).] . . . Inclusion of 
such an analysis in the EIS is necessary to inform Congress, the 
public, and the President as intended by NEPA. [See 40 CFR 
1500.1(a).]'' Forty Most Asked Questions Concerning CEQ's National 
Environmental Policy Act Regulations, 46 FR 18026 (Mar. 23, 1981).
---------------------------------------------------------------------------

    For the DEIS, NHTSA analyzed eight action alternatives, 
Alternatives 1 through 8, which ranged from amending the MY 2021 
standards to match the MY 2020 standards and holding those standards 
flat for passenger cars and light trucks through MY 2026 (Alternative 
1) to maintaining the existing MY 2021 standards and subsequently 
requiring average annual increases in fuel economy by 2.0 percent 
(passenger cars) and 3.0 percent (light trucks) (Alternative 8). The 
action alternatives analyzed in the DEIS also reflected different 
options regarding air conditioning efficiency and off-cycle technology 
adjustment procedures, with some alternatives phasing out these 
adjustments in MYs 2022-2026. For the FEIS, NHTSA analyzed seven action 
alternatives, Alternatives 1 through 7, which range from amending the 
MY 2021 standards to match the MY 2020 standards and holding those 
standards flat for passenger cars and light trucks through MY 2026 
(Alternative 1) to maintaining the existing MY 2021 standards and 
subsequently requiring average annual increases in fuel economy by 2.0 
percent (passenger cars) and 3.0 percent (light trucks) (Alternative 7) 
from year to year. The primary differences between the action 
alternatives for the DEIS and FEIS is that the FEIS did not analyze 
alternatives that phased out the air conditioning efficiency and off-
cycle technology adjustments (see Section V above for further 
discussion), and the FEIS added an alternative under which fuel economy 
increased at 1.5 percent per year for both cars and light trucks 
(Alternative 3). Both of the ranges of action alternatives, as well as 
the No

[[Page 25162]]

Action Alternative, in the DEIS and FEIS encompassed a spectrum of 
possible standards NHTSA could determine was maximum feasible based on 
the different ways the agency could weigh EPCA's four statutory 
factors. Throughout the FEIS, estimated impacts were shown for all of 
these action alternatives, as well as for the No Action Alternative. 
For a more detailed discussion of the environmental impacts associated 
with the alternatives, see Chapters 3-8 of the FEIS, as well as Section 
VII above.
    NHTSA's FEIS describes potential environmental impacts to a variety 
of resources, including fuel and energy use, air quality, climate, land 
use and development, hazardous materials and regulated wastes, 
historical and cultural resources, noise, and environmental justice. 
The FEIS also describes how climate change resulting from global carbon 
emissions (including CO2 emissions attributable to the U.S. 
light duty transportation sector under the alternatives considered) 
could affect certain key natural and human resources. Resource areas 
are assessed qualitatively and quantitatively, as appropriate, in the 
FEIS, and the findings of that analysis are summarized here.\2896\
---------------------------------------------------------------------------

    \2896\ The impacts described in this section come from NHTSA's 
FEIS, which is being publicly issued simultaneously with this final 
rule. As described in Section VII.A.4.c.1 above, the FEIS is based 
on ``unconstrained'' modeling rather than ``standard setting'' 
modeling; NHTSA conducts modeling both ways in order to reflect the 
various statutory requirements of EPCA and NEPA. The preamble 
employs the ``standard setting'' modeling in order to ensure that 
the decision-maker does not consider things that EPCA/EISA prohibit, 
but as a result, the impacts reported here may differ from those 
reported elsewhere in this preamble. However, NHTSA considers the 
impacts reported in the FEIS, in addition to the other information 
presented in this preamble, as part of its decision-making process.
---------------------------------------------------------------------------

    As the stringency of the alternatives increases, total U.S. 
passenger car and light truck fuel consumption for the period of 2020 
to 2050 decreases. Total light-duty vehicle fuel consumption from 2020 
to 2050 under the No Action Alternative is projected to be 3,371 
billion gasoline gallon equivalents (GGE). Light-duty vehicle fuel 
consumption from 2020 to 2050 under the action alternatives is 
projected to range from 3,598 billion GGE under Alternative 1 to 3,456 
billion gallons GGE under Alternative 7. Under the Alternative 3, 
light-duty vehicle fuel consumption from 2020 to 2050 is projected to 
be 3,571 GGE. All of the action alternatives would increase fuel 
consumption compared to the No Action Alternative, with fuel 
consumption increases that range from 226 billion GGE under Alternative 
1 to 85 billion GGE under Alternative 7.
    The relationship between stringency and air pollutant emissions is 
less straightforward, reflecting the complex interactions among the 
tailpipe emissions rates of the various vehicle types, the technologies 
assumed to be incorporated by manufacturers in response to the CAFE 
standards, upstream emissions rates, the relative proportions of 
gasoline and diesel in total fuel consumption, and changes in VMT from 
the rebound effect. In general, emissions of criteria and toxic air 
pollutants increase across all action alternatives, with some 
exceptions. Further, the action alternatives would result in increased 
incidence of PM2.5-related adverse health impacts (including 
increased incidences of premature mortality, acute bronchitis, 
respiratory emergency room visits, and work-loss days) due to the 
emissions increases.\2897\
---------------------------------------------------------------------------

    \2897\ As discussed in Section X.E.1, NHTSA also performed a 
national-scale photochemical air quality modeling and health benefit 
assessment for the FEIS, which is included as Appendix E. This 
analysis affirms the estimates that appeared in the DEIS and 
explains conclusions that may be drawn from the FEIS air quality 
discussion.
---------------------------------------------------------------------------

    For CO (in 2025), NOX (in 2025), and SO2, 
emissions generally decrease under the action alternatives compared to 
the No Action Alternative. For CO in 2025, the largest decrease occurs 
under Alternative 1 and the emissions decreases get smaller from 
Alternative 1 through Alternative 7. For NOX in 2025, the 
largest decrease occurs under Alternative 6. For SO2 in 
2025, the largest decrease occurs under Alternative 6; however, 
SO2 emissions under Alternative 7 are greater than under the 
No Action Alternative. For SO2 in 2035, the largest decrease 
occurs under Alternative 2. For SO2 in 2050, the largest 
decrease occurs under Alternative 1 and the emissions decreases get 
smaller from Alternative 1 through Alternative 7. Across all criteria 
pollutants, action alternatives, and analysis years, the smallest 
decrease in emissions is less than 0.1 percent and occurs for 
NOX under Alternative 7 in 2025; the largest decrease is 12 
percent and occurs for SO2 under Alternative 2 in 2050.
    For CO (in 2035 and 2050), NOX (in 2035 and 2050), 
PM2.5, and VOCs, emissions show increases across action 
alternatives compared to the No Action Alternative, with the largest 
increases occurring under Alternative 1 (except CO in 2035, for which 
the largest increase occurs under Alternative 4). The emissions 
increases get smaller from Alternative 1 through Alternative 7. 
Exceptions to this trend are for PM2.5 and VOCs in 2025, 
which show the smallest emissions increase under Alternative 6. Across 
all criteria pollutants, action alternatives, and analysis years, the 
smallest increase in emissions is 0.1 percent and occurs for 
SO2 under Alternative 7 in 2025; the largest increase is 12 
percent and occurs for VOCs under Alternative 1 in 2050.
    Under each action alternative in 2025 compared to the No Action 
Alternative, decreases in emissions would occur for all toxic air 
pollutants except for DPM, for which emissions would increase by as 
much as 2 percent. For 2025, the largest relative decreases in 
emissions would occur for 1,3,-butadiene, for which emissions would 
decrease by as much as 0.5 percent. Percentage reductions in emissions 
of acetaldehyde, acrolein, benzene, and formaldehyde would be less. 
Under each action alternative in 2035 and 2050 compared to the No 
Action Alternative, increases in emissions would occur for all toxic 
air pollutants. The largest relative increases in emissions would occur 
for DPM, for which emissions would increase by as much as 9 percent. 
Percentage increases in emissions of acetaldehyde, acrolein, benzene, 
1,3,-butadiene, and formaldehyde would be less.
    In addition, the action alternatives would result in increased 
incidence of PM2.5-related adverse health impacts due to the 
emissions increases. Increases in adverse health outcomes include 
increased incidences of premature mortality, acute bronchitis, 
respiratory emergency room visits, and work-loss days. In 2025 and 
2035, all action alternatives except for Alternative 6 would result in 
increased adverse health impacts nationwide compared to the No Action 
Alternative as a result of increases in emissions of NOX, 
PM2.5, and DPM. The increases in adverse health impacts are 
largest for the least stringent alternative (Alternative 1). The 
increases get smaller from Alternative 1 to Alternative 4, get larger 
from Alternative 4 to Alternative 5, then smaller from Alternative 5 to 
Alternative 6, and larger again from Alternative 6 to Alternative 7. In 
2050, all action alternatives would result in decreased adverse health 
impacts nationwide compared to the No Action Alternative as a result of 
decreases in emissions of SOX. The decreases in adverse 
health impacts get smaller from Alternative 1 to Alternative 7.
    The action alternatives would increase U.S. passenger car and light 
truck fuel consumption and CO2 emissions compared with the 
No Action

[[Page 25163]]

Alternative, resulting in minor increases to the anticipated increases 
in global CO2 concentrations, temperature, precipitation, 
and sea level, and minor decreases in ocean pH that would otherwise 
occur, as described below. They could also, to a small degree, increase 
the impacts and risks of climate change. Uncertainty exists regarding 
the magnitude of impact on these climate variables, as well as to the 
impacts and risks of climate change. Still, the impacts of the action 
alternatives on global mean surface temperature, precipitation, sea 
level, and ocean pH would be extremely small in relation to global 
emissions trajectories. This is because of the global and multi-
sectoral nature of climate change. These effects would be small, would 
occur on a global scale, and would not disproportionately affect the 
United States.
    According to the FEIS, passenger cars and light trucks are 
projected to emit 85,900 million metric tons of carbon dioxide 
(MMTCO2) from 2021 through 2100 under the No Action 
Alternative. Alternative 1 would increase these emissions by 10 percent 
through 2100 (approximately 8,800 MMTCO2). Alternative 7 
would increase these emissions by 4 percent through 2100 (approximately 
3,100 MMTCO2). Emissions increases would be highest under 
Alternative 1 and would decrease across the action alternatives, with 
emissions being the lowest under the No Action Alternative.
    In the FEIS, NHTSA presented two different analyses based on these 
emissions changes to illustrate potential impacts to certain climate 
variables. In the first analysis, to represent the direct and indirect 
impacts of this action, NHTSA used the Global Change Assessment Model 
(GCAM) Reference scenario (i.e., future global emissions assuming no 
additional climate policy [``business-as-usual'']) to represent the 
reference case emissions scenario. Under that analysis, total global 
CO2 emissions from all sources are projected to be 4,950,865 
MMTCO2 under the No Action Alternative from 2021 through 
2100, which means that the action alternatives are expected to increase 
global CO2 emissions between 0.06 (Alternative 7) and 0.17 
(Alternative 1) percent by 2100. The estimated CO2 
concentrations in the atmosphere for 2100 would range from 789.89 parts 
per million (ppm) under Alternative 1 to approximately 789.11 ppm under 
the No Action Alternative, indicating a maximum atmospheric 
CO2 increase of approximately 0.78 ppm compared to the No 
Action Alternative.
    Changes in CO2 emissions translate to changes in global 
mean surface temperature, sea levels, global mean precipitation, and 
ocean pH, among other things. Under the first analysis, global mean 
surface temperature is projected to increase by approximately 
3.48[deg]C (6.27 [deg]F) under the No Action Alternative by 2100. 
Implementing the lowest-emissions action alternative (Alternative 7) 
would increase this projected temperature rise by 0.001[deg]C (0.002 
[deg]F), while implementing the highest-emissions alternative 
(Alternative 1) would increase projected temperature rise by 
0.003[deg]C (0.005 [deg]F). Projected sea-level rise in 2100 ranges 
from a low of 76.28 centimeters (30.03 inches) under the No Action 
Alternative to a high of 76.35 centimeters (30.06 inches) under 
Alternative 1. Alternative 1 would result in an increase in sea level 
equal to 0.07 centimeter (0.03 inch) by 2100 compared with the level 
projected under the No Action Alternative, compared to an increase 
under Alternative 7 of 0.02 centimeter (0.001 inch) compared with the 
No Action Alternative. Global mean precipitation is anticipated to 
increase by 5.85 percent by 2100 under the No Action Alternative. Under 
the action alternatives, this increase in precipitation would be 
increased further by 0.01 percent. Finally, ocean pH in 2100 is 
anticipated to be 8.2715 under Alternative 7, about 0.0001 less than 
the No Action Alternative. Under Alternative 1, ocean pH in 2100 would 
be 8.2712, or 0.0004 less than the No Action Alternative.
    In the second analysis, NHTSA used the GCAM6.0 scenario instead of 
the default scenario to represent the reference case emissions 
scenario. The GCAM6.0 scenario assumes a moderate level of global GHG 
reductions and corresponds to stabilization, by 2100, of total 
radiative forcing and associated CO2 concentrations at 
roughly 678 ppm. By assuming a moderate level of global GHG reduction, 
NHTSA attempts to capture the cumulative impacts of this action (i.e., 
the impact on the environment which results from the incremental impact 
of the action when added to other past, present, and reasonably 
foreseeable future actions). In the FEIS, NHTSA documented a number of 
domestic and global actions that indicate that a moderate reduction in 
the growth rate of global GHG emissions is reasonably foreseeable in 
the future.
    Under the second analysis, compared with projected total global 
CO2 emissions of 4,044,005 MMTCO2 from all 
sources from 2021 to 2100, the incremental impact of this rulemaking is 
expected to increase global CO2 emissions between 0.08 
(Alternative 7) and 0.22 (Alternative 1) percent by 2100. Estimated 
atmospheric CO2 concentrations in 2100 range from a low of 
687.3 ppm under the No Action Alternative to a high of 688.04 ppm under 
Alternative 1. Alternative 7, the lowest CO2 emissions 
alternative, would result in CO2 concentrations of 687.55 
ppm, an increase of 0.26 ppm compared with the No Action Alternative. 
Global mean surface temperature increases for the action alternatives 
compared with the No Action Alternative in 2100 range from a low of 
0.001[deg]C (0.002 [deg]F) under Alternative 7 to a high of 0.004[deg]C 
(0.007 [deg]F) under Alternative 1. Global mean precipitation is 
anticipated to increase by 4.77 percent by 2100 under the No Action 
Alternative. Under the action alternatives, this increase in 
precipitation would be increased further by 0.01 percent. Projected 
sea-level rise in 2100 ranges from a low of 70.22 centimeters (27.65 
inches) under the No Action Alternative to a high of 70.30 centimeters 
(27.68 inches) under Alternative 1, indicating a maximum increase of 
sea-level rise of 0.07 centimeter (0.03 inch) by 2100. Sea-level rise 
under Alternative 7 would be 70.25 centimeters (27.66 inches), a 0.03 
centimeter (0.01-inch) increase compared to the No Action Alternative. 
Ocean pH in 2100 is anticipated to be 8.2721 under Alternative 7, about 
0.0001 less than the No Action Alternative. Under Alternative 1, ocean 
pH in 2100 would be 8.2719, or 0.0004 less than the No Action 
Alternative.
    For several other resources, NHTSA is unable to provide a 
quantitative measurement of potential impacts. Instead, the FEIS 
presents a qualitative discussion on potential impacts. In most cases, 
NHTSA presents the findings of a literature review of scientific 
studies, such as in Chapter 6, where NHTSA provides a literature 
synthesis focusing on existing credible scientific information to 
evaluate the most significant lifecycle environmental impacts from some 
of the fuels, materials, and technologies that may be used to comply 
with the alternatives. In Chapter 7, NHTSA discusses land use and 
development, hazardous materials and regulated waste, historical and 
cultural resources, noise, and environmental justice. Finally, in 
Chapter 8, NHTSA discusses cumulative impacts related to energy, air 
quality, and climate change, and provides a literature synthesis of the 
impacts on key natural and human resources of changes in climate change 
variables. In these chapters, NHTSA concludes that impacts would be 
proportional to changes in emissions that would result

[[Page 25164]]

under the alternatives. As a result, among the action alternatives, 
Alternative 1 would have the highest impact on these resources while 
Alternative 7 would have the lowest.
    Based on the foregoing, NHTSA concludes from the FEIS that the No 
Action Alternative is the overall environmentally preferable 
alternative because, assuming full compliance were achieved regardless 
of the agency's assessment of the costs to industry and society, it 
would result in the largest reductions in fuel use and CO2 
emissions among the alternatives considered. In addition, the No Action 
Alternative would result in the lowest overall emissions levels of 
criteria air pollutants (with the exception of sulfur dioxide) and of 
the toxic air pollutants studied by NHTSA. Impacts on other resources 
(especially those described qualitatively in the FEIS) would be 
proportional to the impacts on fuel use and emissions, as further 
described in the FEIS, with the No Action Alternative expected to have 
the fewest negative impacts.\2898\ Although the CEQ regulations require 
NHTSA to identify the environmentally preferable alternative,\2899\ the 
agency need not adopt it, as described above. The following section 
(Section VIII.B.4) explains how NHTSA balanced the relevant factors to 
determine which alternative represented the maximum feasible standards, 
including why NHTSA does not believe that the environmentally 
preferable alternative is maximum feasible.
---------------------------------------------------------------------------

    \2898\ Among the action alternatives considered, Alternative 7 
would be the environmentally preferable alternative, as it is 
closest in stringency to the No Action Alternative.
    \2899\ 40 CFR 1505.2(b).
---------------------------------------------------------------------------

4. Evaluating the EPCA Factors and Other Considerations To Arrive at 
the Proposed Standards
    As discussed in this section, NHTSA is required to consider four 
enumerated factors when establishing maximum feasible CAFE standards 
under 49 U.S.C. chapter 329: ``technological feasibility, economic 
practicability, the effect of other motor vehicle standards of the 
Government on fuel economy, and the need of the United States to 
conserve energy.'' \2900\ For this final rule, NHTSA has considered a 
wide range of potential CAFE standards (Baseline/No Action Alternative 
and Alternatives 1 through 7), ranging from the augural standards set 
forth in 2012 (Baseline/No Action Alternative), through a number of 
less stringent alternatives, including the proposed preferred 
alternative (Alternative 1, 0 percent per year stringency improvement) 
and what has been chosen as the final standards (Alternative 3, 1.5 
percent per year stringency improvement). NHTSA has determined that 
Alternative 3, which would increase the stringency of the MY 2020 
standards by 1.5 percent per year for both passenger cars and light 
trucks from MY 2021 through 2026, represents the maximum feasible CAFE 
standards under 49 U.S.C. 39202. In addition to technological 
feasibility, economic practicability, the effects of other motor 
vehicle standards of the Government on fuel economy, and the need of 
the United States to conserve energy, NHTSA has also considered the 
impact of the standards on safety and the environment.
---------------------------------------------------------------------------

    \2900\ 49 U.S.C. 32902(f).
---------------------------------------------------------------------------

How did the Agency balance the factors for the NPRM?
    In the NPRM, NHTSA began its discussion of the tentative balancing 
of factors by explaining that ``NHTSA well recognizes that the decision 
it proposes to make in today's NPRM is different from the one made in 
the 2012 final rule that established standards for MY 2021 and 
identified ``augural'' standard levels for MYs 2022-2025. Not only do 
we believe that the facts before us have changed, but we believe that 
those facts have changed sufficiently that the balancing of the EPCA 
factors and other considerations must also change. The standards we are 
proposing today reflect that balancing.'' \2901\ NHTSA highlights this 
discussion at the outset in response to the number of commenters who 
claimed that NHTSA had not acknowledged or explained in the NPRM how or 
why the proposal was different from past work or policy decisions.
---------------------------------------------------------------------------

    \2901\ 83 FR at 43213.
---------------------------------------------------------------------------

    The NPRM balancing discussion went on to explore the definition of 
``to conserve'' in the context of what ``energy conservation'' and 
``the need of the U.S. to conserve energy'' should be interpreted to 
mean, in recognition of the major structural changes in global oil 
markets since EPCA was originally passed, and even since the 2012 final 
rule that set forth the augural standards. NHTSA examined these changes 
from both a demand perspective and a supply perspective. On the demand 
side, U.S. demand and global demand have both changed over time. The 
NPRM discussed the fact that the U.S. consumes a much smaller share of 
global oil output than it did at the CAFE program's outset, both 
because U.S. fleet fuel economy has improved, and because other 
countries that were not major petroleum consumers in the 1970s have 
rapidly increased their share of consumption, and continue to do so. A 
more globalized market means that risk of price spikes is spread 
around--making the U.S. in particular less likely to bear a 
disproportionate burden of price spikes. The NPRM also discussed the 
decreasing energy intensity of the U.S. economy over time and the 
improving balance of payments in petroleum, including the likelihood 
that the U.S. is poised to become a net petroleum exporter in the near 
future. Related to the decreasing energy intensity of the U.S. economy, 
on the demand side, the NPRM discussed the proliferation of fuel-
efficient vehicle options in the market in response to CAFE increases 
over time, and the fact that consumers who wish to purchase more fuel 
efficient vehicles have largely done so, and may continue to do so over 
time if they wish.
    On the supply side, the NPRM explained, vast increases in U.S. 
petroleum production, largely from shale formations, have introduced a 
major new stable supply into the global market. Shale oil production 
costs may be higher than the cost (for example, to OPEC members) to 
produce traditional oil, but that itself acts as a lever on global 
prices. Prices of goods like oil are affected by demand and supply--
given that global demand trends increase relatively steadily, if OPEC 
States want to increase revenues by selling more of the total oil 
consumed globally, they have to try to control global supply volume by 
controlling production volumes (to avoid shale production increasing in 
response to higher prices). In short, the higher global prices trend, 
the more U.S. shale production increases in response, and as supply 
increases, prices fall. The NPRM discussed the responsiveness of U.S. 
shale production and suggested it could be higher than traditional 
producers in some instances. Traditional oil producers seeking to 
maintain market share have a new incentive to keep prices below a 
certain threshold, and U.S supply helps to buffer the impact of 
geopolitical events. The NPRM looked at then-current EIA oil price 
forecasts, under which U.S. gasoline prices were not forecast to exceed 
$4/gallon through 2050, and acknowledged that while price shocks could 
still occur, NHTSA tentatively concluded that from the supply side, it 
is possible that the oil market conditions that created the price 
shocks in the 1970s may no longer exist.
    In light of these changes in global oil markets, the NPRM 
tentatively concluded that many aspects of the need of the U.S. to 
conserve energy had

[[Page 25165]]

improved enough over time to merit further consideration of what the 
need of the United States is to conserve oil today and going forward. 
With regard to environmental considerations, the NPRM returned to the 
definition of ``to conserve'' and suggested that differences of 
thousandths of a degree Celsius in 2100 resulting from higher levels of 
carbon dioxide emissions under the proposal as compared to the augural 
standards might not rise to the level of ``wasteful,'' given the other 
considerations discussed. With regard to consumer costs, the NPRM 
discussed the interplay of oil market conditions with prior arguments 
about consumer ``myopia'' with regard to the benefits of fuel savings, 
and tentatively concluded that U.S. consumers may be valuing fuel 
savings appropriately and purchasing the vehicles they want to 
purchase--i.e., that using CAFE standards as a tool to compel consumers 
to save money may not be necessary.
    Given the discussion above, NHTSA tentatively concluded that the 
need of the U.S. to conserve energy may no longer function as assumed 
in previous considerations of what CAFE standards would be maximum 
feasible. In that discussion, NHTSA stated that the overall risks 
associated with the need of the U.S. to conserve oil have entered a new 
paradigm with the risks substantially lower today and projected into 
the future than when CAFE standards were first issued and in the recent 
past. NHTSA explained that the effectiveness of CAFE standards in 
reducing the demand for fuel combined with the increase in domestic oil 
production have contributed significantly to the current situation and 
outlook for the near- and mid-term future. NHTSA tentatively concluded 
that the world has changed, and the need of the U.S. to conserve 
energy, at least in the context of the CAFE program, has also changed.
    Of two other factors under 32902(g), the NPRM explained that the 
changes were perhaps less significant. NHTSA suggested that all of the 
alternatives appear as though they could narrowly be considered 
technologically feasible, in that they could be achieved based on the 
existence or the projected future existence of technologies that could 
be incorporated on future vehicles. With regard to the effect of other 
motor vehicle standards of the Government on fuel economy, the NPRM 
explained that it was similarly not heavily limiting during this 
rulemaking time frame. The NPRM analysis projected that neither safety 
standards nor Tier 3 compliance obligations appeared likely to make it 
significantly harder for industry to comply with more stringent CAFE 
standards, and that EPA's CO2 standards should have no 
greater effect on difficulty in meeting CAFE standards than already 
existed.
    For economic practicability, the NPRM considered the traditional 
definition used by the agency, and expressed concern that all of the 
alternatives considered in the proposal could raise economic 
practicability concerns. NHTSA stated that it believed there could be 
potential for unreasonable elimination of consumer choice, loss of U.S. 
jobs, and a number of adverse economic consequences under nearly all if 
not all of the regulatory alternatives considered in the NPRM. NHTSA 
explored consumer choice issues given a foreseeable future of 
relatively low fuel prices and the likelihood that more stringent CAFE 
standards could cause automakers to add technology to new vehicles that 
consumers do not want, or prevent the addition of technology to new 
vehicles that consumers do want, and suggested that there could be risk 
that such elimination of consumer choice could be unreasonable. NHTSA 
explained its assumption, based on repeated manufacturer input, that 
fuel-saving technologies that paid for themselves within 2.5 years 
would be added regardless of CAFE stringency, meaning that the power of 
CAFE standards (by themselves) to compel fuel savings was reduced. 
NHTSA suggested that requiring more technology to be added than 
consumers were willing to pay for could have dampening effects on 
vehicle sales, particularly given forecasted relatively low gas prices, 
increasing the likelihood of automaker non-compliance with more 
stringent standards due to difficulty in selling higher-fuel-economy 
models. NHTSA examined the levels of electrification necessary to meet 
the various regulatory alternatives evaluated in the NPRM and compared 
them with information about consumers' willingness to purchase vehicles 
with these technologies and even to spend money on fuel economy 
improvements generally. NHTSA suggested that if the market for higher 
fuel-economy vehicles exists and is already possibly saturated, 
increasing fuel economy requirements could create economic 
practicability concerns by affecting sales and consumer choice.
    NHTSA recognized that automakers cross-subsidize regulation-driven 
cost increases and expressed concern about their ability to do that 
under sustained, ongoing increases over many years, and the 
corresponding concern that continued cross-subsidizing could create 
affordability problems for lower-income consumers if manufacturers pass 
costs forward to consumers more broadly rather than concentrating them 
in high-volume, higher-profit vehicles. NHTSA suggested that higher 
vehicle prices and monthly vehicle payments could outweigh, for at 
least some new vehicle purchasers, the benefit of fuel savings, because 
vehicle payments are fixed costs and fuel costs may be less fixed. 
NHTSA expressed concern that as vehicles get more expensive in response 
to higher CAFE standards, it will become more and more difficult for 
finance companies and dealers to continue creating loan terms that keep 
monthly payments low and do not result in consumers' still owing 
significant amounts of money on the vehicle by the time they can be 
expected to be ready for a new vehicle. This situation may imply a 
bubble in new vehicle sales, the effects of which could fall 
disproportionately on new and low-income buyers. NHTSA suggested that 
these effects could impact both fleet-wide safety (by slowing fleet 
turnover) and consumer choice. The NPRM also expressed concern that the 
sales and employment analyses were unable to capture (1) the risk that 
manufacturers and dealers may not be able to continue keeping monthly 
new vehicle payments low, or (2) the risk that manufacturing could 
shift overseas as manufacturing costs rise.
    NHTSA also examined the net benefits of the various regulatory 
alternatives, and noted that the analysis showed that consumers recoup 
only a portion of the costs associated with increasing stringency under 
all of the alternatives, because the fuel savings resulting from each 
of the alternatives was substantially less than the costs associated 
with the alternative, meaning that net savings for consumers improved 
as stringency decreased. NHTSA explained that it recognized that this 
was a significantly different analytical result from the 2012 rule, 
which showed the opposite trend, and explained that the result was 
different because the facts and analysis underlying the result were 
also different, and enumerated the noteworthy differences, such as 
payback assumptions; fleet composition; what levels of technologies had 
already been applied; the costs and effectiveness values for some of 
those technologies; fuel price forecasts; the value of the rebound 
effect; the value of the social cost of carbon; accounting for price 
impacts on fleet turnover; not limiting mass reduction to only the 
largest vehicles; and the value of a statistical life having increased. 
NHTSA explained that all of these changes, together, meant

[[Page 25166]]

that the standards under any of the regulatory alternatives (compared 
to the preferred alternative) were more expensive and had lower 
benefits than if they had been calculated using the inputs and 
assumptions of the 2012 analysis. This assessment, in turn, contributed 
to the agency's decision to reevaluate what standards might be maximum 
feasible in the model years covered by the rulemaking. NHTSA explained 
that it had thus both relied on new facts and circumstances in 
developing the proposal and reasonably rejected prior analyses relied 
on in the 2012 final rule.\2902\
---------------------------------------------------------------------------

    \2902\ See FCC v. Fox Television Stations, 556 U.S. at 514-515; 
see also NAHB v. EPA, 682 F.3d 1032 (D.C. Cir. 2012).
---------------------------------------------------------------------------

    NHTSA then considered that ``maximum feasible'' may change over 
time as the agency assessed the relative importance of each factor that 
Congress requires it to consider, and tentatively concluded that 
proposing CAFE standards that hold the MY 2020 curves for passenger 
cars and light trucks constant through MY 2026 would be the maximum 
feasible standards for those fleets and would fulfill EPCA's 
overarching purpose of energy conservation in light of the facts before 
the agency and as the agency expected them to be in the rulemaking time 
frame. NHTSA recognized that this was a different interpretation from 
the 2012 final rule and explained that the context of that rulemaking 
was meaningfully different from the current context, because the facts 
had changed the importance of the need of the U.S. to conserve energy, 
and NHTSA recognized that under that circumstance, while more stringent 
standards may be possible, insofar as production-ready technology 
exists that the industry could physically employ to reach higher 
standards, it was not clear that higher standards would be economically 
practicable in light of current U.S. consumer needs to conserve energy. 
Therefore, NHTSA stated, it viewed the determination of maximum 
feasible standards as a question of the appropriateness of standards 
given that their need--either from the societal-benefits perspective in 
terms of risk associated with fuel price shocks or other related 
catastrophes, or from the private-benefits perspective in terms of 
consumer willingness to purchase new vehicles with expensive 
technologies that may allow them to save money on future fuel 
purchases--seems likely to remain low for the foreseeable future. NHTSA 
also considered the effects of the standards on highway safety and 
expressed concern that because more stringent standards could depress 
sales and slow fleet turnover, and because higher fuel economy leads to 
more driving and more exposure to crash risk, all regulatory 
alternatives would improve safety as compared to the augural standards.
(b) What comments did NHTSA receive regarding how it balanced the 
factors in the NPRM?
    In addition to comments on each of the factors NHTSA considered 
discussed above, comments also were received on how NHTSA should 
balance these factors in determining the maximum feasible final 
standards. Hundreds of thousands of comments addressed stringency and, 
thus, the agency's evaluation of what standards were maximum feasible. 
Most of those focused on the augural standards: Many individual 
commenters supported reducing the stringency of the standards from 
augural levels--some citing estimates of cost, and some citing concerns 
about consumer choice. Many comments by other individual commenters 
supported retaining stringency at augural levels or increasing 
stringency beyond that level--generally citing concerns about climate 
change and increased fuel costs under less stringent standards. A few 
commenters, like CEI, expressly supported the proposal, and even 
suggested that stringency should be decreased further. Many other 
commenters, including environmental and consumer groups, health 
advocacy organizations, and a number of State organizations, argued 
that the proposal was flawed and/or that the augural standards should 
be finalized because more stringent standards help to reduce climate 
change and address other air quality issues.\2903\ The Congressional 
Tri-Caucus commenters supported maintaining the augural standards, 
stating that they contribute to employment and protect low income 
communities and communities of color.\2904\
---------------------------------------------------------------------------

    \2903\ See, e.g., Harvard Environmental Law Clinic, EPA-HQ-OAR-
2018-0283-5486, at 1; University of San Francisco graduate students, 
EPA-HQ-OAR-2018-0283-2676, at 1-2; Vanderbilt student organizations, 
EPA-HQ-OAR-2018-0283-4189, at 1-2; Blue Planet Foundation, EPA-HQ-
OAR-2018-0283-4207, at 1; Green Energy Institute (Lewis and Clark 
Law School), et al., EPA-HQ-OAR-2018-0283-4193, at 1-3; CBD et al., 
NHTSA-2018-0067-12057, at 2; NESCAUM, NHTSA-2018-0067-11691, at 3-4.
    \2904\ Congressional Tri-Caucus, NHTSA-2018-0067-1424, at 1.
---------------------------------------------------------------------------

    The Alliance and Global Automakers both supported final standards 
that increased in stringency year over year. The Alliance stated that 
it could support stringency increases between 0 percent per year and 2-
3 percent per year ``along with the inclusion of appropriate 
flexibilities.'' \2905\ Global stated that increases should be 
``meaningful'' \2906\ and suggested that ``[i]n order for the U.S. auto 
industry to remain competitive and continue to export vehicles to the 
rest of the world, industry is best served by a reasonable, steady ramp 
rate that accounts for investments made and the global nature of the 
market. Steady increases allow for long-term planning and create an 
environment of security that fosters ongoing investment in vehicle 
technology and consumer confidence in purchasing new vehicles. It also 
provides a level playing field upon which automakers can compete.'' 
\2907\ Toyota made similar points, and argued that while the standards 
set in 2012 are beyond maximum feasible today, the ``statutes support 
an adjustment to those standards that reflect the realities of the 
market, consumer choice, and the pace of technological advancement 
acceptable to consumers.'' \2908\ Mazda stated that it supported 
``increasing requirements for fuel efficiency. . ., if they are 
sensible and achievable under changing market conditions.'' \2909\
---------------------------------------------------------------------------

    \2905\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 8.
    \2906\ Global, NHTSA-2018-0067-12032, at 3.
    \2907\ Global, NHTSA-2018-0067-12032, Attachment A, at A-11.
    \2908\ Toyota, NHTSA-2018-0067-12150, at 31.
    \2909\ Mazda, NHTSA-2018-0067-11727, at 2.
---------------------------------------------------------------------------

    NADA commented that it was willing to support standards that 
increased in stringency (i.e., more stringent than the proposal) if 
they were economically practicable and technologically feasible, based 
on the evidence before the agencies; if they ensured consumer choice 
and ``the strongest possible rate of fleet turnover;'' and if passenger 
car and light truck standards increased at the same rate.\2910\ The 
Alliance for Vehicle Efficiency (AVE) argued that compliance shortfalls 
are evidence that the current rate of stringency increase is beyond 
maximum feasible, and that the assumptions that enabled those rates to 
be chosen ``are no longer feasible based on consumer adoption.'' \2911\ 
AVE suggested that a rate of increase of 2.5 percent per year for both 
cars and trucks, retroactively imposed beginning in MY 2018, would be 
feasible given sufficient flexibilities.\2912\
---------------------------------------------------------------------------

    \2910\ NADA, NHTSA-2018-0067-12064, at 12.
    \2911\ AVE, NHTSA-2018-0067-11696, at 6-8.
    \2912\ Id., at 10.
---------------------------------------------------------------------------

    NADA also stressed the importance of flexibilities as a compliance 
tool for meeting standards that increase faster

[[Page 25167]]

than the proposal.\2913\ The Minnesota agencies supported maintaining 
standards at the augural levels, commenting that automakers has simply 
``requested additional flexibility . . ., not a wholesale rollback of 
the standards,'' and suggesting that additional flexibilities would 
enable augural levels.\2914\ IPI disagreed with the suggestion in the 
NPRM that heavy automaker reliance on credits for compliance might 
indicate that standards were beyond maximum feasible, arguing that 
automakers must be either using credits about to expire, or counting on 
future standards being cheaper to meet due to rising consumer demand 
for fuel economy, technology costs decreasing over time, and the cost-
effectiveness of EPA's EV multiplier incentive.\2915\
---------------------------------------------------------------------------

    \2913\ NADA, NHTSA-2018-0067-12064, at 12.
    \2914\ 1 Minnesota agencies, NHTSA-2018-0067-11706, at 6-7.
    \2915\ IPI, NHTSA-2018-0067-12213, Appendix, at 25-26.
---------------------------------------------------------------------------

    With regard to analysis of costs and benefits, IPI argued that the 
final rule needed, like the 2012 rule, to cite costs and benefit 
expressly in discussing balancing of statutory factors, but with a 
``proper'' accounting of costs and benefits. IPI claimed that in the 
NPRM the factors were balanced ``in a way that conflicts with the . . 
.controlling statute[ ] and weighed . . .without regard for the 
accuracy of the accompanying cost-benefit analysis.'' \2916\ IPI stated 
that ``. . . the agencies' analysis produced biased and irrational 
results at each of the steps in that causal chain, leading to a 
Proposed Rule that vastly overstates the benefits of the rollback and 
understates the benefits society foregoes with the rollback,'' and that 
``[a] full and balanced analysis of all the costs and benefits that the 
agencies are charged with considering would reveal--as the midterm 
review recently confirmed--that the baseline standards will deliver 
massive net social benefits, and the proposed rollback is 
unjustified.'' \2917\
---------------------------------------------------------------------------

    \2916\ Id.
    \2917\ IPI, NHTSA-2018-0067-12213, Appendix, at 1-2.
---------------------------------------------------------------------------

    With regard to net benefits, the States and Cities commenters 
stated that prior analyses had concluded that the net benefits of the 
augural standards were extremely high,\2918\ while the Alliance stated 
that ``[t]he NERA-Trinity Assessment confirms the Agencies' findings 
that Alternatives 1, 5, and 8 result in increased net benefits relative 
to the no-action alternative augural CAFE standards.''\2919\ Michalek 
and Whitefoot commented that ``maximizing net benefits is among the 
most important factors to consider in policy selection because it is an 
effort to weigh a variety of policy implications on a common basis and 
seek decisions that are beneficial to society overall,'' but also 
cautioned that estimates are inherently uncertain and should be 
transparent and clearly justified; that sensitivity analysis is 
necessary; that a net benefits analysis will not be able to capture 
distributional effects or changes in behavior caused by the policy; and 
that ``it is not clear that there is necessarily any relationship 
between MNB and setting the `maximum feasible' criteria while 
considering `economic practicability.' ''\2920\ IPI disagreed with the 
NPRM's suggestion that feasibility concerns could lead NHTSA not to 
maximize net benefits, stating that ``if a standard were truly not 
feasible, then its costs would be prohibitively high, and a full and 
fair cost-benefit analysis would reflect that.'' \2921\
---------------------------------------------------------------------------

    \2918\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 6.
    \2919\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 13.
    \2920\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 14-15.
    \2921\ IPI, NHTSA-2018-0067-12213, Appendix, at 11.
---------------------------------------------------------------------------

    CARB argued that ``[a]lthough EPCA provides NHTSA with some 
discretion with respect to balancing the four factors, that discretion 
is nevertheless constrained by EPCA's overriding mandate of conserving 
energy.'' \2922\ CARB further stated that EPCA ``envision[s] the 
promulgation of increasingly stringent requirements to ensure the 
continued reductions of both emissions and fuel consumption from motor 
vehicles.'' \2923\ Michalek and Whitefoot similarly commented that the 
requirement that standards be maximum feasible necessarily means that 
stringency must increase over time, because technology capabilities and 
cost are constantly improving; international regulations are constantly 
increasing in stringency; and if standards are held constant, 
automakers will always exceed them.\2924\ The States and Cities 
commenters cited the CAS language from the D.C. Circuit that ``[i]t is 
axiomatic that Congress intended energy conservation to be a long term 
effort that would continue through temporary improvements in energy 
availability,'' and argued that ``[w]hile NHTSA purports to acknowledge 
this purpose and the importance of improving fuel economy over time, 
NHTSA proposes to do the opposite: roll back fuel economy standards for 
a period of at least six years.'' \2925\ The States and Cities 
commenters further argued that NHTSA had ``departed sharply from its 
past interpretations and practice without an adequate explanation, 
often without even an acknowledgement,'' citing Fox Television, insofar 
as the 2012 final rule justification had noted that less stringent 
regulatory alternatives would have conserved less energy than the then-
finalized standards, as compared to ``[w]ith the Proposed Rollback, 
NHTSA has radically changed positions--assuming energy conservation 
provides little, if any, benefits, for example--without explaining or 
even acknowledging this complete reversal of course.'' \2926\ The 
States and Cities commenters concluded that it was ``impermissible'' 
for NHTSA to balance ``the factors in a manner that contravenes EPCA's 
central purpose of energy conservation.'' \2927\
---------------------------------------------------------------------------

    \2922\ CARB, NHTSA-2018-0067-11783, Detailed Comments, at 78.
    \2923\ Id., at 80.
    \2924\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 3-4.
    \2925\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 64-65.
    \2926\ Id., at 65.
    \2927\ Id.
---------------------------------------------------------------------------

    ACEEE commented that NHTSA did not have discretion to assess 
whether the need of the U.S. to conserve energy was as great as when 
EPCA was first passed, arguing that ``[t]he statute does not ask for a 
determination on whether the nation needs to save energy. It assumes 
the need and directs that the need be taken into account along with 
other considerations.'' \2928\ Securing America's Energy Future 
commented that the need of the U.S. to conserve energy continued, and 
that ``[a]lthough the nation is undoubtedly more energy secure than it 
was before the start of the U.S. shale oil revolution ten years ago,'' 
\2929\ ``[u]ntil the U.S. transportation sector is no longer beholden 
to oil, the country will be vulnerable to oil price volatility. 
Improving the fuel efficiency of the U.S. vehicle fleet is a valuable 
insurance policy against this volatility.'' \2930\ IPI also commented 
that fuel efficiency standards act as insurance, but against 
unpredictable future fuel prices.\2931\ IPI stated that anticipating 
relatively low future fuel prices was not an appropriate basis for 
finalizing the proposal, both because fuel costs may rise in the 
future, and also because

[[Page 25168]]

EPA's Final Determination ``found that that even with the lowest prices 
projected in AEO 2016 of close to $2, the `lifetime fuel savings 
significantly outweigh the increased lifetime costs' of the GHG 
standards.'' \2932\ IPI further argued that ``[i]n ignoring the [FD] 
analysis, the Proposed Rule has failed to provide a `reasoned 
explanation' for dismissing the `facts and circumstances that underlay' 
the original rule, rendering its analysis arbitrary and capricious.'' 
\2933\ IPI also argued that NHTSA had not adequately explained its 
``shift since 2012 in its interpretation and application of the need to 
conserve energy factor,'' stating that ``[a]ctual fuel savings, and the 
associated benefits to consumers, the environment, and society, were at 
the heart of NHTSA's analysis of the need to conserve energy factor 
back in 2012. Now the agency ignores those conclusions from 2012 and 
relies on mistaken and inconsistent interpretations of petroleum import 
projections and the urgency of climate change to justify ignoring this 
statutory factor and giving primacy instead to economic practicability 
and safety effects. The failure to explain this shift in approach is 
arbitrary.'' \2934\
---------------------------------------------------------------------------

    \2928\ ACEEE, NHTSA-2018-0067-12122, main comments, at 1.
    \2929\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 17.
    \2930\ Id., at 7, 8.
    \2931\ IPI, NHTSA-2018-0067-12213, Appendix, at 31.
    \2932\ Id., at 32.
    \2933\ Id.
    \2934\ Id., at 6.
---------------------------------------------------------------------------

    UCS argued that the need of the United States to conserve energy is 
``the most important of the four required factors'' according to CBD v. 
NHTSA, and claimed that ``NHTSA has manipulated the evaluation of the 
factors to produce a result that supports the preferred option in the 
NPRM.'' \2935\ The States and Cities commenters argued that it was 
``[c]ynical. . .'' for NHTSA to justify the proposal on the basis that 
``the oil intensity of U.S. GDP has continued to decline'' in part as a 
result of increasingly stringent CAFE standards, and on the basis that 
``[m]anufacturers have responded to fuel economy standards and to 
consumer demand over the last decade to offer a wide array of fuel-
efficient vehicles in different segments and with a wide array of 
features.'' \2936\
---------------------------------------------------------------------------

    \2935\ UCS, NHTSA-2018-0067-12039, at 3, 7.
    \2936\ States and Cities, NHTSA-2018-0067-11735, Detailed 
Comments, at 64-65.
---------------------------------------------------------------------------

    CARB and CBD et al. argued that if NHTSA's analysis indicates that 
automakers will voluntarily exceed the standards, then the standards 
cannot be maximum feasible.\2937\ Robertson commented relatedly that 
standards should not be set below augural levels because ``Much higher 
fuel economy and reduced emissions have been achieved by several lower 
priced makes and models using hybrid technology.'' \2938\ Blue Planet 
Foundation stated that the augural standards are feasible because 
automakers have already invested in technologies, and electrification 
is projected to continue to grow cheaper over time, so that ``even the 
up-front cost of an EV will begin to reach parity with gas-powered cars 
by 2024.'' \2939\ ACEEE also cited the voluntary overcompliance in the 
NPRM analysis as evidence that there could not be diminishing returns 
from higher fuel efficiency standards, because ``the list of [cost-
effective] technology [must] continually regenerate itself'' if 
manufacturers would continue applying it in the absence of future 
standards. Moreover, ACEEE argued, past analyses had always found 
plenty of available cost-effective technologies, and automakers would 
find a way to apply them.\2940\
---------------------------------------------------------------------------

    \2937\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84; 
CBD et al., NHTSA-2018-0067-12057, at 2.
    \2938\ Robertson, EPA-HQ-OAR-2018-0283-0787, at 3.
    \2939\ Blue Planet Foundation, EPA-HQ-OAR-2018-0283-4207, at 1-
2.
    \2940\ ACEEE, NHTSA-2018-0067-12122, main comments, at 9.
---------------------------------------------------------------------------

c) How is NHTSA Balancing the Factors to Determine the Maximum Feasible 
Final CAFE Standards?
    EPCA/EISA grants the Secretary (by delegation, NHTSA) discretion in 
how to balance the relevant statutory factors, while bearing in mind 
EPCA's overarching purpose of energy conservation. Many commenters 
cited the Ninth Circuit's language in CBD v. NHTSA that ``the 
overarching purpose of EPCA is energy conservation,'' \2941\ and the 
D.C. Circuit's language in CAS v. NHTSA that ``[i]t is axiomatic that 
Congress intended energy conservation to be a long term effort that 
would continue through temporary improvements in energy availability.'' 
\2942\ NHTSA has considered those comments and those court decisions 
carefully as it made the decision set forth in the final rule. Based on 
the information before the agencies and considering carefully the 
comments received, NHTSA has determined that the preferred alternative 
identified in the proposal--amending the MY 2021 standards to match MY 
2020, and holding those standards flat through MY 2026--does not 
represent the maximum feasible standards, and that the maximum feasible 
standards for MYs 2021-2026 passenger cars and light trucks increase in 
stringency by 1.5 percent per year from the MY 2020 standards. The 
following discussion walks through NHTSA's evaluation and balancing of 
the relevant factors in light of the information before it.
---------------------------------------------------------------------------

    \2941\ CBD, 508 F.3d 508, 537 (9th Cir. 2007), opinion vacated 
and superseded on denial of reh'g, 538 F.3d 1172 (9th Cir. 2008).
    \2942\ CAS, 793 F.2d 1322, 1340 (D.C. Cir. 1986).
---------------------------------------------------------------------------

(1) Need of the U.S. to Conserve Energy
    NHTSA agrees with commenters that energy conservation remains 
important, and that changed conditions, even significantly changed 
conditions, do not obviate NHTSA's obligation to set maximum feasible 
CAFE standards as directed by Congress. Many commenters disagreed 
strongly with NHTSA's suggestion in the NPRM that increased U.S. 
petroleum production, and the U.S.'s likely imminent status as a net 
petroleum exporter, decreased the need of the U.S. to conserve energy. 
NHTSA agrees that there is still a need to conserve energy, and oil in 
particular. Like an insurance policy or a savings account, continuing 
to move the needle forward on CAFE helps position Americans better to 
weather certain types of possible future uncertainty. NHTSA believes 
that it is reasonable to be somewhat conservative about this risk, and 
thus to set CAFE standards that increase in stringency year over year 
through MY 2026.
    That said, NHTSA believes that there are limits to how much 
uncertainty the CAFE program can mitigate--continuing to make progress 
is important, but it is also important to be transparent and realistic 
about what is being accomplished, even if NHTSA were able to set 
standards beyond levels that NHTSA considers maximum feasible. NHTSA 
also continues to believe that structural changes in global oil markets 
over the last 10 years, driven in part by changes in demand both in the 
U.S. and abroad, and in part by the significant growth in U.S. 
petroleum production, have led to a fundamental shift in the dynamics 
of global oil prices, which has in turn improved U.S. (and possibly, 
global) energy security. NHTSA believes that this shift is important to 
consider as NHTSA weighs the need of the Nation to conserve energy.
    NHTSA acknowledges that price shocks can still happen. The large 
scale attack on Saudi Arabia's Abqaiq processing facility--the world's 
largest crude oil processing and stabilization plant--on September 14, 
2019 caused ``the largest single-day [crude oil] price increase in the 
past decade,'' of between $7 and $8, according to EIA.\2943\ The Abqaiq 
facility has a capacity to process

[[Page 25169]]

7 million barrels per day, or about 7 percent of global crude oil 
production capacity. By September 17, however, also according to EIA,
---------------------------------------------------------------------------

    \2943\ https://www.eia.gov/todayinenergy/detail.php?id=41413.

    Saudi Aramco reported that Abqaiq was producing 2 million 
barrels per day, and they expected its entire output capacity to be 
fully restored by the end of September. In addition, Saudi Aramco 
stated that crude oil exports to customers will continue by drawing 
on existing inventories and offering additional crude oil production 
from other fields. Tanker loading estimates from third-party data 
sources indicate that loadings at two Saudi Arabian export 
facilities were restored to the pre-attack levels. Likely driven by 
news of the expected return of the lost production capacity, both 
Brent and WTI crude oil prices fell on Tuesday, September 17.\2944\
---------------------------------------------------------------------------

    \2944\ Id.

    Thus, the largest single-day oil price increase in the past decade 
was largely resolved within a week, and assuming very roughly that 
average crude oil prices were $70/barrel in September 2019 (slightly 
higher than actual), an increase of $7/barrel would represent a 10 
percent increase as a result of the Abqaiq attack. Contrast this with 
the 1973 Arab oil embargo, which lasted for months and raised prices 
350 percent.\2945\ Saudi Arabia could have benefited, revenue-wise, 
from higher prices following the Abqaiq attack, but instead moved 
rapidly to restore production and tap reserves to control the risk of 
resulting price increases, likely recognizing that long-term sustained 
price increases would reduce their ability to control global supply 
(and thus prices, and thus their own revenues) by relying on their 
lower cost of production.\2946\ Even if the NPRM discussion was perhaps 
overconfident about the ability of U.S. shale producers to act as 
``swing'' supply, as some commenters suggested, it seems clear from 
events that the existence of U.S. production has a stabilizing effect 
on global oil prices. This has played out in important ways in the 
first quarter of 2020, with the dissolution of the ``OPEC+'' coalition 
as Russia and Saudi Arabia compete for market share in response to U.S. 
shale production and also in the wake of global demand downturn.\2947\
---------------------------------------------------------------------------

    \2945\ See Jeanne Whalen, ``Saudi Arabia's oil troubles don't 
rattle the U.S. as they used to,'' Washington Post, September 19, 
2019, available at https://www.washingtonpost.com/business/2019/09/19/saudi-arabias-oil-troubles-dont-rattle-us-like-they-used/.
    \2946\ See, e.g., ``Dynamic Delivery: America's Evolving Oil and 
Natural Gas Transportation Infrastructure,'' National Petroleum 
Council (2019) at 18, available at: https://dynamicdelivery.npc.org/downloads.php. See also ``Oil prices plunge as Trump speech eases 
Iran fears,'' CNN, available at https://www.cnn.com/2020/01/07/business/oil-prices-iran-attack-iraq/index.html.
    \2947\ See, e.g., EIA, ``This Week in Petroleum--OPEC shift to 
maintain market share will result in global inventory increases and 
lower prices,'' March 11, 2020, https://www.eia.gov/petroleum/weekly/; DOE, ``DOE Responds to Recent Oil Market Activity,'' March 
9, 2020, https://www.energy.gov/articles/doe-responds-recent-oil-market-activity; Reid Standish, Keith Johnson, ``No End in Sight to 
the Oil Price War Between Russia and Saudi Arabia,'' March 14, 2020, 
https://foreignpolicy.com/2020/03/14/oil-price-war-russia-saudi-arabia-no-end-production/; Alex Ward, ``The Saudi Arabia-Russia oil 
war, explained,'' March 9, 2020, https://www.vox.com/2020/3/9/21171406/coronavirus-saudi-arabia-russia-oil-war-explained.
---------------------------------------------------------------------------

    Even though the effect of significant supply disruptions appears 
much lower than was the case several years ago, the analysis for this 
final rule (like the NPRM analysis) does, in fact, explicitly account 
for the possible occurrence of price shocks. The cost penalty used in 
the analysis to represent the consequences of those shocks attempts to 
quantify the negative impact on U.S. GDP created by abrupt, short-term 
increases in the world oil price. The values used in the NPRM were 
based on arguably outdated work, and commenters cited more recent 
studies of relevance in their comments on the NPRM--one of which formed 
the basis for the estimates in today's analysis. The final rule 
estimate of this cost are based on a recent study which states that 
``[i]n recent years, the United States has become much more self-
reliant in producing oil, and a newer economics literature suggests 
that oil demand may be more elastic and U.S. GDP may be less sensitive 
to world oil price shocks than was previously estimated. These 
developments suggest somewhat lower security costs may be associated 
with U.S. oil consumption.'' \2948\ These more recent studies concede 
that the fact that ``the world has not seen a major oil supply 
disruption since 2003,'' and that therefore ``we have no reliable 
method to quantify the effects of these disruptions,'' \2949\ but even 
the range of uncertainty suggests that the risk has decreased relative 
to prior estimates. The price shock cost estimate employed in the NPRM 
was at least twice as large as the upper bound of the range in Brown's 
new estimates, and consistently close to the upper bound of the range 
of his more conservative estimates. The approach taken today, which 
relies on median estimates in Brown's study, implies that risk is more 
properly estimated here than in the NPRM.
---------------------------------------------------------------------------

    \2948\ Brown, Stephen, ``New estimates of the security costs of 
U.S. oil consumption,'' Energy Policy 113 (2018) 171-192, at 171. 
Cited in Securing America's Energy Future, NHTSA-2018-0067-12172, at 
29.
    \2949\ Brown, at 181.
---------------------------------------------------------------------------

    Commenters (Bordoff, SAFE, CARB, IPI) argued that increased U.S. 
petroleum production, which improves the stability of the global supply 
and reduces the probability of supply interruptions, does not reduce 
U.S. exposure to petroleum price shocks, which are still determined by 
the dynamics of the global market. By reducing the probability of 
supply disruptions in the global market, the U.S. does reduce its 
vulnerability to price shocks. However, to the extent that the 
vulnerability to price shocks is a function of exposure, commenters are 
correct that looming petroleum independence does not entirely insulate 
the U.S. economy from the consequences of global oil price shocks. Some 
commenters further argued that the proposed standard would leave the 
U.S. more exposed to oil price shocks, which would harm consumers. 
Basic mathematics means that a less efficient on-road fleet necessarily 
would spend more on fuel than a more efficient on-road fleet in the 
event of a sudden, unexpected, and dramatic increase in oil price. The 
suggestion in these comments, however, is that finalizing the augural 
standards would sufficiently insulate U.S. consumers from harm during 
such an event, while finalizing any other regulatory alternative would 
not. NHTSA disagrees that finalizing the augural standards, as compared 
to the standards we are finalizing, would make a meaningful difference 
in this case.
    A continuous, but slow, price increase over several years is 
fundamentally different from the kinds of acute price shocks over which 
commenters have expressed understandable concern. Long-term price 
increases signal consumers to make investments in fuel economy, in both 
the new and used vehicle markets, and to diversify the vehicles in 
their household fleets. In a side analysis using outputs from the CAFE 
Model, the agencies examined the consequences of a gasoline price spike 
in 2030--increasing the price from $3.40/gallon to $6/gallon for eight 
months, then reverting back to $3.40/gallon.\2950\ By choosing a year 
so far in the future, the agencies consider a larger gap in fleet fuel 
efficiency than is attributable to this action. If the agencies 
increase stringency again after MY 2026, the efficiency gap between the 
on-road fleet in the final standards and baseline would be smaller than 
simulated here. This side analysis showed that even a nearly doubling 
of the fuel price, sustained for more than half a year, would result in 
less than 1 percent savings in fuel expenditures for that

[[Page 25170]]

year under the final standards (relative to the proposal), compared to 
about 5 percent reduction in expenditures under the augural standards. 
This demonstrates that even though finalizing the augural standards 
would mitigate American drivers' increase in fuel expenditures by more 
than the standards the agencies are finalizing today, it would only do 
so by a few percent. This is important to understanding concerns about 
differences in the amount of fuel saved under today's final standards 
versus if the augural standards were finalized, as will be discussed 
more below. And as also discussed below, NHTSA believes the augural 
standards are beyond maximum feasible at this time.
---------------------------------------------------------------------------

    \2950\ Docketed in NHTSA-2018-0067.
---------------------------------------------------------------------------

    Some commenters raised the possibility that the U.S. might ban 
fracking at some point in the future, and suggested that therefore the 
need of the U.S. to conserve energy could not be assumed away. NHTSA 
acknowledges that the future is uncertain. Without the supply of U.S. 
oil in the global market, NHTSA agrees that it is foreseeable that 
conditions could revert somewhat to how global oil market conditions 
were before the ramp-up in U.S. supply--i.e., that the global market as 
a whole could be somewhat less stable and thus fuel prices could be 
somewhat more prone to change unexpectedly and for longer periods. 
Pulling out of the market on the supply side means that the agencies 
would lose the ability to influence the market on that side. 
Presumably, part of the policy objective of banning fracking would be 
to accelerate a transition to a post-oil transportation system. In that 
scenario, presumably decision-makers would consider higher fuel prices 
to be an acceptable tradeoff for less driving and lower emissions. That 
said, the availability of shale oil resources does exist today, and is 
not realistically in question. And, even if the future availability of 
that capacity was realistically doubtful, any increase in fuel economy 
above current levels, like the final rule will require, will help 
somewhat to mitigate the economic pain to drivers of that event were it 
to occur, as shown above.\2951\ To the extent that current events cause 
pauses or consolidation in the shale industry's development, while that 
may lead to transitory difficulty for the shale industry, the resources 
will continue to exist, and U.S. shale will continue to be able to act 
as a lever to keep global prices from rising very high for very long.
---------------------------------------------------------------------------

    \2951\ See also Letter from Alliance for Automotive Innovation, 
NADA, and MEMA to Congress, Mar. 23, 2020, available at https://www.autosinnovate.org/wp-content/uploads/2020/03/COVID-19-Letter-to-Congress-NADA-MEMA-AAI-March-23.pdf.
---------------------------------------------------------------------------

    As noted above, Securing America's Energy Future commented that 
``[a]lthough the nation is undoubtedly more energy secure than it was 
before the start of the U.S. shale oil revolution ten years ago,'' 
\2952\ ``[u]ntil the U.S. transportation sector is no longer beholden 
to oil, the country will be vulnerable to oil price volatility. 
Improving the fuel efficiency of the U.S. vehicle fleet is a valuable 
insurance policy against this volatility.'' \2953\ (Emphasis added.) 
NHTSA agrees fully with this comment. Energy security concerns were the 
driving force behind the creation of the CAFE program, as discussed in 
the NPRM. U.S. energy security has improved, but the only way to 
resolve petroleum-related energy security concerns entirely would be 
for the U.S. vehicle fleet to stop using oil. And doing so would not 
avoid energy-related concerns entirely, but rather shift them away from 
petroleum (and the Middle East) and toward battery-related security 
(and lithium-, nickel-, cobalt-, and other metals-producing 
countries).\2954\
---------------------------------------------------------------------------

    \2952\ Securing America's Energy Future, NHTSA-2018-0067-12172, 
at 17.
    \2953\ Id., at 7, 8.
    \2954\ While progress is being made on developing and improving 
domestic sources for many of the minerals necessary for battery 
development, the U.S. is still heavily dependent on imports of both 
raw materials and batteries. Regarding minerals production and 
import dependence, see Schulz, K.J., DeYoung, J.H., Jr., Seal, R.R., 
II, and Bradley, DC, eds., Critical mineral resources of the United 
States--Economic and environmental geology and prospects for future 
supply: U.S. Geological Survey Professional Paper 1802 (see 
particularly Chapter K, p. K1-K21 on lithium), available at https://www.commerce.gov/sites/default/files/2020-01/Critical_Minerals_Strategy_Final.pdf and https://pubs.usgs.gov/pp/1802/k/pp1802k.pdf. Regarding vehicle battery supply chains, see 
Coffin, D., and J. Horowitz, ``The Supply Chain for Electric Vehicle 
Batteries,'' Journal of International Commerce and Economics, 
December 2018, available at https://www.usitc.gov/publications/332/journals/the_supply_chain_for_electric_vehicle_batteries.pdf.
---------------------------------------------------------------------------

    Our relationship to the global energy market has changed 
significantly since the CAFE program was created, with most of this 
change occurring over the last decade. The United States has become 
energy independent, and is currently a net exporter of petroleum 
products. Rising world oil prices no longer only mean a financial 
burden on U.S. drivers and a wealth transfer to foreign nations. While 
rising prices continue to affect U.S. motorists, we have taken steps to 
insulate our transportation system from exogenous price shocks. CAFE 
standards (and, recently, CO2 standards) have increased the 
efficiency of new vehicles for more than a decade, and these 
increasingly efficient vehicles are still working their way into the 
on-road fleet as older models are retired. Accompanying any increase in 
the global oil price is an increase in revenue to the U.S. oil 
industry. To the extent that motorists are spending more on oil 
everywhere, the dollars spent on domestically produced petroleum 
products stay within the U.S. and additional revenue from foreign 
buyers flows into our domestic energy industry. To the extent that the 
U.S. transportation system is able to further reduce its dependence on 
petroleum in a cost-effective manner, it is sensible to do so. But in 
the current environment, in which motorized transportation is 
increasingly energy efficient and U.S. energy producers are not only 
supplying our demand but exporting petroleum products to other nations, 
the nationwide benefits of reducing petroleum consumption are 
substantially diminished.
    There is also the opposite concern to bear in mind--that energy 
security is not just about oil becoming more expensive, but also about 
other changes in oil prices. Major fluctuations in either direction, as 
well as oil price collapse, can potentially have seriously 
destabilizing geopolitical effects. Many major oil producing countries 
(some of whom are allies) rely heavily on oil revenues for public 
revenue, and sustained losses in public revenue in certain countries 
and regions can foreseeably create new energy-related security risks, 
not only for the U.S. As the world works toward transitioning away from 
oil for transportation, keeping prices reasonably stable may best help 
that transition remain peaceful and steady. In short, energy security 
can cut both ways, and the current estimates of price shock that we 
model inherently do not account for the longer-term stabilizing effect 
of steady global oil consumption (of which the U.S. is a part) on 
global security. Steady trends in consumption can facilitate steady 
changes in production, which can facilitate a steady security 
situation.
    NHTSA does not interpret EPCA/EISA to mean that Congress expected 
the CAFE program to take the U.S. auto fleet off of oil entirely--
indeed, EISA renders doing so impossible because it amended EPCA to 
prohibit NHTSA from considering the fuel economy of dedicated 
alternative fuel vehicles, including electric vehicles, when setting 
maximum feasible standards. This means that standards cannot be set 
that assume increased usage of full electrification for compliance. 
Reading that prohibition together with the obligation to set maximum 
feasible standards by considering (which is hard

[[Page 25171]]

to do without balancing) factors like economic practicability with the 
need of the U.S. to conserve energy, NHTSA believes that Congress 
intended CAFE to try to mitigate the risk of gas lines, but not to 
shift the fleet entirely off of oil. Moreover, the EISA-added 
requirement that standards ``increase ratably'' for MYs through 2020 
ceases to apply beginning in MY 2021. While NHTSA unquestionably has 
discretion to determine that standards should continue to increase 
post-MY 2020, NHTSA does not interpret EPCA/EISA as requiring that they 
do, as long as they are maximum feasible. Several commenters suggested 
that standards that do not continue to increase, by definition, cannot 
be maximum feasible, but NHTSA believes that this interpretation does 
not account for the clear requirement that maximum feasible standards 
be determined with reference to the four statutory factors. The statute 
does not preclude an interpretation that non-increasing standards could 
be maximum feasible, depending on the facts before the agency. Neither 
does the statute preclude an interpretation that amending standards 
downward can be maximum feasible, as has occurred in the past in 
response to changes in consumer demand.\2955\
---------------------------------------------------------------------------

    \2955\ See, Center for Auto Safety v. NHTSA (CAS), 793 F.2d 1322 
(D.C. Cir. 1986).
---------------------------------------------------------------------------

    Nevertheless, for purposes of this final rule, NHTSA does believe 
that standards that increase in stringency are maximum feasible; the 
question remains by how much those standards should increase. While 
NHTSA agrees that CAFE standards must conserve energy, the improvement 
in energy security discussed above is entirely relevant to how much 
energy should be conserved. If the marginal improvement in energy 
security of increasing CAFE stringency from one regulatory alternative 
to another is very small, as it appears to be based on the above 
discussion, then other aspects of the need of the U.S. to conserve 
energy must be considered next to see what effect they have.
    Consumer costs, as discussed above, is another aspect of the need 
of the U.S. to conserve energy. The final rule analysis estimates that 
all alternatives besides the baseline/augural standards would result in 
higher fuel costs for consumers than the baseline/augural standards 
would result in, as follows:
[GRAPHIC] [TIFF OMITTED] TR30AP20.735

    A number of commenters stated that the 2012 rulemaking had relied 
on fuel savings as part of its justification, and argued that the NPRM 
had not adequately grappled with the fact that the proposal would have 
cost consumers more in fuel expenditures than if NHTSA finalized the 
augural standards. In fact, NHTSA explained in the NPRM that while fuel 
costs would be higher, NHTSA believed that the higher upfront (and 
ongoing, if financed) costs of new vehicles and associated taxes and 
registration fees--as well as the opportunity cost associated with 
those upfront costs--would outweigh, for many consumers, the additional 
fuel costs that would be incurred if standards were less stringent than 
augural. That continues to be the case under the final rule analysis, 
as discussed below. In addition, Section VI.D. discusses how past 
rulemaking analyses assumed that consumers were `myopic' and/or did not 
have adequate information about the benefits of fuel savings, which led 
them to choose to purchase less efficient vehicles than they otherwise 
would if they better understood the costs or savings they would accrue. 
As Section VI.D. explains, the agencies are less certain today that 
consumers improperly value fuel savings. Vehicle buyers today have more 
information about fuel costs than ever before, including right on the 
window sticker when considering a new vehicle purchase, and it is 
ultimately a private choice whether consumers prefer improvements in 
other vehicle attributes over additional fuel economy. When fuel costs 
are expected to rise manageably over time, it may be that consumers are 
comfortable choosing to absorb an additional $1,375 over the vehicle's 
lifetime, the estimated difference in lifetime expenditures between the 
proposal and if NHTSA was choosing to finalize the augural standards, 
and are even more comfortable choosing to absorb an additional $1,125, 
the estimated difference in lifetime expenditures between the final 
standards and what

[[Page 25172]]

the augural standards would have required. If fuel prices rise less 
than anticipated, as they have done since the 2012 final rule, or even 
decrease over time, buyers face an even smaller tradeoff between 
foregone fuel savings and the value of improvements in other aspects of 
new cars.
    Consumer expenditures on fuel are important to understanding the 
benefits (and net benefits) of CAFE and CO2 standards. Every 
analysis of CAFE/CO2 standards relies on hundreds of 
assumptions, and estimates of costs and benefits developed as part of 
those analyses, by their very nature, depend on those assumptions. 
Specifically, the net benefits associated with each alternative result 
from the assumptions used and the relationships between vehicle 
production, ownership, and usage in which the assumptions interact. Put 
more simply, inputs affect outputs. As discussed in the section above 
on economic practicability, net benefits may be a consideration in the 
determination of maximum feasible standards, among the many other 
things the agency considers. While some commenters have asserted that 
the analysis for this rulemaking has ``put a thumb on the scale by 
undervaluing the benefits and overvaluing the costs of more stringent 
standards,'' \2956\ this final rule has identified a number of critical 
assumptions in the 2012 final rule that were problematic in the other 
direction (i.e., undervaluing the costs and overvaluing the benefits), 
for a variety of reasons. For example, the projected fuel prices in the 
2012 analysis inflated the value of fuel savings relative to what has 
actually occurred. That assumption about how fuel prices were projected 
to rise over time was solidly grounded at the time, but is no longer 
so, and continuing to use it would not be reasonable, even if that 
means that the benefits of all of the regulatory alternatives decrease 
as compared to what the 2012 analysis showed. Lower oil prices mean 
that fuel savings benefits for consumers are lower under any CAFE 
standards, whether the augural standards or the standard being 
finalized today--consumers may yet spend less on fuel under more 
stringent standards, but how much less matters.
---------------------------------------------------------------------------

    \2956\ See CBD v. NHTSA, 538 F.3d 1172, 1189 (9th Cir. 2008).
---------------------------------------------------------------------------

    Additionally, the assumption in 2012 that no market exists for fuel 
economy improvements at any fuel price or technology cost artificially 
inflated the value of fuel savings attributable to the standards in 
each regulatory alternative. The combination of assumptions and 
relationships (the examples above, and others) in the 2012 final rule 
produced estimates of net benefits that continued to increase with 
stringency from 1 percent per year through 6 percent per year.\2957\ 
Under some alternatives, benefits actually would have appeared to be 
infinite, growing faster than the discount rate, if the analysis had 
been extended far enough into the future. No market works this way, and 
there is no reasonable set of assumptions under which costs could never 
exceed benefits no matter how much technology was deployed or how much 
stringency was required. Rather than demonstrating meaningfully that 
more stringent standards are always more beneficial to society, the 
result from the 2012 analysis suggests that that analysis was 
critically flawed. That said, while the 2012 analysis appeared to show 
that more technology, at a faster pace, is always preferable from the 
perspective of net benefits, the agencies ultimately relied on other 
features of the analysis and considerations of impacts in choosing a 
preferred alternative. While today's analysis produces an inflection 
point at a 3 percent discount rate--a level of stringency where further 
increases reduce net benefits as the tradeoff between regulatory costs 
and resulting net benefits tips the other way \2958\--the agencies 
similarly rely on considerations beyond net benefits in choosing the 
preferred alternative.\2959\
---------------------------------------------------------------------------

    \2957\ The 7 percent per year alternative happened to be 
indistinguishable from the 6 percent alternative in that analysis.
    \2958\ See Table VII-95.
    \2959\ See CBD v. NHTSA, 538 F.3d 1172, 1188 (9th Cir. 2008).
---------------------------------------------------------------------------

    NHTSA also agrees with many commenters that environmental (both 
climate and air quality) concerns are relevant to the need of the U.S. 
to conserve oil, as explained above. As the Supreme Court stated in 
Massachusetts v. EPA, ``[a] reduction in domestic emissions would slow 
the pace of global emissions increases,'' \2960\ and there is no 
question that CAFE standards directly affect CO2 emissions. 
Besides providing information on differences between the regulatory 
alternatives in terms of million metric tons of CO2 emitted, 
the NPRM also provided a chart illustrating the difference between the 
estimated atmospheric CO2 concentration (789.76 ppm) in 2100 
under the proposal as compared to the estimated level under the augural 
standards (789.11 ppm) in a scenario where no CO2 emissions 
reduction measures are implemented throughout the planet.\2961\ The 
NPRM noted that this translated to 3/1000ths of a degree Celsius 
increase in global average temperatures by 2100, relative to the 
augural standards. Many commenters strongly objected to the framing of 
these findings, as discussed above in the section on the environmental 
implications of the need of the U.S. to conserve energy. Changing the 
framing does not change the agency's findings.\2962\ For this final 
rule, the Preferred Alternative would result in 922.5 million metric 
tons of CO2 more than the estimated emissions if the augural 
standards were to be finalized (for MY 2017-MY 2029 vehicles between 
calendar years 2017 and 2070), which is 160.2 million fewer tons than 
if the proposed Preferred Alternative were to be finalized. It is 
reasonable to consider these raw million-metric-ton estimates in terms 
of their effects, namely, on estimated temperature change and sea level 
rise, which are the primary climate effects referred to and estimated. 
The FEIS accompanying today's rule estimates that, by 2100, global mean 
surface temperature will increase by 3.487 degrees (Celsius) under 
either the proposed or final standards, versus 3.484 degrees under the 
augural standards. The FEIS shows corresponding sea level rise in 2011 
reaching 76.34 cm under the final standards, 76.35 cm under the 
proposed standards, and 76.28 cm under the augural standards. This is 
accounted for in economic terms (i.e., translated from fractions of a 
degree temperature rise and from millimeters of sea level rise, among 
other things, into dollar-based effects) in the measure of the social 
cost of carbon, described in Section VI.D.1.b)(13).
---------------------------------------------------------------------------

    \2960\ Mass. v. EPA, 549 U.S. at 526.
    \2961\ 83 FR at 42996-97 (Aug. 24, 2018).
    \2962\ In fact, NHTSA's analysis in Section 8.6.4.2 of the FEIS 
illustrates that the differences between alternatives are similar in 
reference to other GCAM scenarios. Regardless of whether there will 
be widespread global efforts to mitigate climate change, the impacts 
of this action are roughly the same.
---------------------------------------------------------------------------

    NHTSA is mindful of the language in Massachusetts v. EPA that 
``[a]gencies . . . do not generally resolve massive problems in one 
fell regulatory swoop,'' \2963\ and acknowledges the concerns of many 
commenters that standards less stringent than augural may result in 
higher CO2 emissions. In response, it is important to 
remember that even under the proposal, sales of new vehicles would, 
over time, have continued to improve the fuel economy and reduce the 
CO2 emissions of the on-road fleet through fleet turnover 
effects, as discussed in Section IV. Under the final rule, those rates 
of improvement will likely be faster than they would have been if NHTSA 
were finalizing the

[[Page 25173]]

proposal. Emissions are still being reduced under the final rule, and 
the on-road fleet will be less energy and carbon intensive than it is 
today. NHTSA is taking the impacts of CO2 emissions into 
account, while also considering the other statutory factors in its 
balancing.
---------------------------------------------------------------------------

    \2963\ Mass. v. EPA, 549 U.S. at 524.
---------------------------------------------------------------------------

    It is also important to note that the science of climate change and 
the models used to assess effects on climate variables (and other 
effects discussed in Section VII.A.4.b, and in the DEIS/FEIS) are 
subject to various types and degrees of uncertainty. In light of this, 
NHTSA also conducted climate sensitivity analyses in the FEIS.\2964\ In 
these analyses, NHTSA considered a range of climate sensitivities (1.5 
[deg]C, 2.0 [deg]C, 2.5 [deg]C, 3.0 [deg]C, 4.5 [deg]C, and 6.0 [deg]C) 
for a doubling of CO2 compared to preindustrial atmospheric 
concentrations (278 ppm CO2). Even under the least stringent 
alternative considered (the proposal) and assuming the highest level of 
climate sensitivity (6.0 [deg]C), the global mean surface temperature 
increase in 2100 was 0.006 [deg]C higher than under the augural 
standards. Thus, accounting for some of this uncertainty, impacts on 
global mean surface temperature resulting from this action remain very 
small.
---------------------------------------------------------------------------

    \2964\ See Sections 5.4.2.3 and 8.6.4.2 of the FEIS.
---------------------------------------------------------------------------

    NHTSA received many comments about the costs of delaying 
CO2 emissions reductions and the potential of crossing 
climate tipping points and triggering abrupt climate change. Many of 
these costs and risks are factored in to the social cost of carbon, and 
are therefore considered as part of the agency's cost-benefit analysis. 
And many of these costs and risks cannot be quantified at all: The 
current state of science does not allow for quantifying how increased 
emissions from a specific policy or action might affect the probability 
and timing of abrupt climate change. However, NHTSA does recognize that 
while these costs cannot be quantified, they do exist and must also be 
taken into account. Ultimately, the costs of delaying CO2 
emissions reductions (both the ones that can be accounted for 
quantitatively and those that can only be considered qualitatively) 
must also be balanced against the costs associated with more stringent 
alternatives. Some of the costs associated with more stringent 
alternatives are direct, such as the additional costs passed on to 
consumers for technology that improves fuel economy. Other costs are 
indirect, such as environmental costs associated with more stringent 
fuel economy standards. For example, the increased electrification of 
motor vehicles can result in localized impacts associated with the 
production and recycling of lithium-ion batteries. Similarly, the 
increased reliance on material substitution for vehicle mass reduction 
could result in various environmental impacts associated with 
manufacture and recycling. Certainly, the benefits of these 
technologies in reducing carbon emissions outweighs the other life-
cycle environmental impacts, but that does not mean NHTSA can just 
ignore those impacts, either.
    Many commenters claimed that NHTSA ignored the effects of climate 
change or determined they were inevitable, not urgent enough to act 
upon, or not worth the effort to address at all. NHTSA makes none of 
those determinations here. On the contrary, NHTSA has considered the 
material on this subject in the administrative record and the plethora 
of public comments we received on the topic. The agency recognizes what 
is at stake, but we also recognize that NHTSA is not charged by 
Congress to single-mindedly address carbon emissions at the expense of 
all other considerations. The question before NHTSA is not whether to 
conserve energy (and thereby reduce carbon emissions, which drive 
climate change) but by how much each year. Taking climate change into 
account elevates the importance of the ``need of the United States to 
conserve energy'' criterion in NHTSA's balancing. However, in light of 
the limits in what the agency can achieve, the potential offsetting 
impacts to the environment, and the statutory requirement to consider 
other factors, the impacts of carbon emissions alone cannot drive the 
outcome of NHTSA's decision-making.
    NHTSA also recognizes the potential impacts of this rulemaking on 
air quality. To be clear, this final rule does not directly involve the 
regulation of pollutants such as carbon monoxide, smog-forming 
pollutants (nitrogen oxides and unburned hydrocarbons), or ``air 
toxics'' (e.g., formaldehyde, acetaldehyde, benzene). Nevertheless, 
NHTSA recognizes that this rule is expected to impact such emissions 
indirectly (by reducing travel demand and accelerating fleet turnover 
to newer and cleaner vehicles on one hand while, on the other, 
increasing activity at refineries and in the fuel distribution system). 
Based on a review of Section VII.A.4.c. above and the FEIS, NHTSA 
believes these impacts are much smaller than impacts on fuel use and 
CO2 emissions, and therefore factor in less to the need of 
the U.S. to conserve energy.\2965\
---------------------------------------------------------------------------

    \2965\ For an explanation of how NHTSA considers environmental 
impacts and the differences between the preamble and FEIS analyses, 
see Section VII.A.4.c.1 above.
---------------------------------------------------------------------------

    For criteria pollutants, NHTSA estimates that emissions over the 
lifetimes of vehicles through MY 2029 under the alternatives will not 
change significantly. Tailpipe emissions of most pollutants will 
generally decrease, while upstream emissions will generally increase. 
Overall emissions under the action alternatives for most pollutants 
will increase over time. Changes are not uniform year-to-year, however, 
reflecting the complex interaction of the amount of highway travel, the 
distribution of that travel among different vehicles, upstream 
processes, etc. Generally, tailpipe air toxic emissions decrease while 
upstream air toxic emissions increase. Over the long term, however, the 
upstream emissions increase further while the decreases in tailpipe 
emissions become less pronounced. Overall, NHTSA anticipates that air 
toxic emission will increase over time under the action alternatives. 
Most alternatives result in cumulative increases in adverse health 
impacts associated with total upstream and tailpipe pollutant 
emissions. Although some alternatives would have resulted in decreases, 
the differences among alternatives across the lifetime of vehicles 
through MY 2029 are not large.
    NHTSA also considered the various impacts reported qualitatively in 
the FEIS and described briefly above in Section VIII.B.3. Although the 
agency cannot compare the impacts of the alternatives quantitatively 
(except to the degree that they are otherwise covered by the agency's 
monetary cost-benefit analysis, such as through the social cost of 
carbon), NHTSA recognizes that such impacts would generally increase 
under all the action alternatives compared to the augural standards. In 
Chapter 8 of the FEIS, for example, NHTSA provides a qualitative 
discussion of the long-term impacts of climate change on key natural 
and human resources. While these impacts would be expected to increase 
under the action alternatives, the change is expected to be very small. 
In contrast, the FEIS also discusses some environmental impacts that 
would decrease with the lower stringencies considered in this 
rulemaking. For example, in Chapter 6 of the FEIS, NHTSA provides a 
literature review of potential lifecycle impacts as a result of 
manufacturer use of various materials and technologies to meet the 
standards. NHTSA can account for the benefits to

[[Page 25174]]

tailpipe emissions of these technologies as part of its evaluation of 
technology effectiveness. However, as discussed in the FEIS, accounting 
for the upstream emissions associated with the processes used in the 
manufacture of these technologies can be complicated. Because the 
adoption of these materials and technologies would vary across 
alternatives, and each has varying upstream impacts, the agency cannot 
provide meaningful comparisons across alternatives. Still, any benefit 
to tailpipe CO2, criteria pollutant, or air toxic emissions 
of more stringent alternatives would be offset by the increased 
upstream impacts reported in that section.\2966\
---------------------------------------------------------------------------

    \2966\ In most cases, tailpipe emissions benefits offset 
upstream environmental impacts associated with materials and 
technologies NHTSA considered in its analysis. However, in some 
cases, results may not align with conventional wisdom. For example, 
while EVs can offer significant life-cycle GHG emissions savings 
over conventional vehicles, this is highly dependent on the time and 
location of charging. In some regions, life-cycle impacts are 
similar for EVs and conventional vehicles.
---------------------------------------------------------------------------

    In total, environmental impacts factor into the need of the U.S. to 
conserve energy and potentially elevate that criterion, but those 
impacts cannot be considered in isolation. While some impacts are more 
significant than others, NHTSA must consider how much weight to place 
on this factor as well as the relative weight of other factors.
    Thus, even if the agency no longer interprets the need of the U.S. 
to conserve energy as necessarily boundless as it once did, as it 
explained in the NPRM and again in the discussion above, NHTSA 
continues to believe that the factor functions in the overall balancing 
to push toward increases in stringency, and notes that any increase in 
stringency over the last binding standards--not in question at this 
point, the standards for MY 2020--does conserve energy and reduce 
negative environmental impacts. In fact, fleet turnover over time means 
that less energy is being consumed by the fleet over time even if 
standards did not increase year over year. Even if new vehicles are not 
all as efficient as would have been required under more stringent 
standards, they are still more efficient on average than the older 
vehicles they are replacing, particularly after a decade of successive 
increases in CAFE standard stringency, as Section IV above discusses. 
The on-road fleet has well over 250 million vehicles, dwarfing the 
roughly 16 million new vehicles sold each year. Comprehensive energy 
savings come from turning over legacy vehicles in the fleet so that 
overall fleet fuel economy increases. If the NPRM's preferred 
alternative were finalized, the fuel consumption of the passenger car 
and light truck fleet would have fallen from roughly 8.5 million 
barrels per day (currently) to roughly 7 million barrels per day by 
2050 as the fleet turned over. Finalizing the 1.5 percent alternative 
reduces that number to 6.3 million barrels per day. That breaks the 
trend of increasing oil consumption over time, and conserves energy.
(2)Technological Feasibility and the Effect of Other Motor Vehicle 
Standards of the Government on Fuel Economy
    As in the 2012 final rule, technological feasibility and the effect 
of other motor vehicle standards of the Government on fuel economy do 
weigh in NHTSA's balancing of the relevant factors, but they play a 
less significant role because they vary less across regulatory 
alternatives than the other factors vary. Technological feasibility, as 
explained above and as similarly explained in 2012, relates to whether 
technologies exist and can be commercially applied during the 
rulemaking timeframe. None of the regulatory alternatives under 
consideration today would require brand new technologies to be 
invented--they can all be met with technology that exists currently. 
However, as recognized in the 2012 final rule, ``some technologies that 
currently have limited commercial use cannot be deployed on every 
vehicle model in MY [2021], but require a realistic schedule for 
widespread commercialization to be feasible. . . . Any of the 
alternatives could thus be achieved on a technical basis alone if the 
level of resources that might be required to implement the technologies 
is not considered.'' As explained above in the discussion of economic 
practicability, however, resources must be, and are, considered. The 
2012 final rule further explained that ``If all alternatives are at 
least theoretically technologically feasible in the [rulemaking] 
timeframe, and the need of the nation is best served by pushing 
standards as stringent as possible, then the agency might be inclined 
to select the alternative that results in the very most stringent 
standards considered.'' The 2012 final rule stated, however, that such 
a selection would be inappropriate because ``the agency must also 
consider what is required to practically implement technologies, which 
is part of economic practicability, and to which the most stringent 
alternatives give little weight.''
    NHTSA considers technological feasibility similarly to how it has 
long considered that factor--for the most part, the question of what 
standards are maximum feasible is less about technological feasibility 
than about economic practicability. All of the regulatory alternatives 
considered in this final rule are likely technologically feasible, but 
that does not mean that any of them could be maximum feasible, just as 
we concluded in evaluating alternatives in 2012. NHTSA must now account 
for how the need of the U.S. to conserve oil has changed, and this 
consideration tips our balancing away from the most stringent 
standards.
    For the effect of other motor vehicle standards of the Government 
on fuel economy, there is relatively little variation across regulatory 
alternatives, as discussed in the FRIA. As in the 2012 final rule, in 
developing this final rule NHTSA considered the effects of compliance 
with known and possible NHTSA safety standards and known EPA emission 
standards in developing this final rule, and has accounted for those 
effects in the analysis. The effect of other motor vehicle standards of 
the Government does not, therefore, have a noticeable effect on NHTSA's 
balancing of factors to determine maximum feasible standards.
(3) Economic Practicability
    Economic practicability remains a complex factor to consider and 
balance, as discussed above, encompassing a variety of different issues 
that are each captured to various degrees through the analysis. As 
NHTSA stated in the 2012 final rule, ``The agency does not necessarily 
believe that there is a bright-line test for whether a regulatory 
alternative is economically practicable, but there are several metrics 
. . . that we find useful for making the assessment.'' \2967\ In 2012, 
as today, NHTSA looks to factors like:
---------------------------------------------------------------------------

    \2967\ 77 FR at 63038 (Oct. 15, 2012).
---------------------------------------------------------------------------

     Per-vehicle cost, in terms of ``even if the technology 
exists and it appears that manufacturers can apply it consistent with 
their product cadence, if meeting the standards will raise per-vehicle 
cost more than we believe consumers are likely to accept, which could 
negatively impact sales and employment in this sector, the standards 
may not be economically practicable;'' \2968\
---------------------------------------------------------------------------

    \2968\ Id.
---------------------------------------------------------------------------

     Application rate of technologies, because ``even if 
shortfalls are not extensive, whether it appears that a regulatory 
alternative would impose undue burden on manufacturers in either or 
both the near and long term in terms of how much and which technologies 
might be required'' can be

[[Page 25175]]

relevant to manufacturers' difficulty with meeting standards; \2969\
---------------------------------------------------------------------------

    \2969\ Id.
---------------------------------------------------------------------------

     Consumer demand, which NHTSA described in 2012 as ``other 
. . . considerations related to the application rate of technologies, 
whether it appears that the burden on several or more manufacturers 
might cause them to respond to the standards in ways that compromise . 
. . other aspects of performance that are important to consumer 
acceptance of new products'' \2970\
---------------------------------------------------------------------------

    \2970\ Id.
---------------------------------------------------------------------------

     Manufacturer compliance shortfalls, because ``If it 
appears, in our modeling analysis, that a significant portion of the 
industry cannot meet the standards defined by a regulatory alternative 
in a model year, given that our modeling analysis accounts for 
manufacturers' expected ability to design, produce, and sell vehicles 
(through redesign cycle cadence, technology costs and benefits, etc.), 
then that suggests that the standards may not be economically 
practicable;'' \2971\
---------------------------------------------------------------------------

    \2971\ Id.
---------------------------------------------------------------------------

     Uncertainty and consumer acceptance of technologies, which 
the 2012 final rule said was ``not accounted for expressly in our 
modeling analysis, but [was] important to an assessment of economic 
practicability given the time frame of this rulemaking.'' \2972\
---------------------------------------------------------------------------

    \2972\ Id.
---------------------------------------------------------------------------

    Thus, estimated impacts on per-vehicle cost are one issue; 
estimated sales and employment impacts are issues; uncertainty 
surrounding future market conditions and consumer demand for fuel 
economy (versus consumer demand for other vehicle attributes) are other 
issues. Consumers may respond to per-vehicle cost increases by choosing 
to keep their current vehicle or buy used vehicles instead of new 
vehicles, with consequent effects on new vehicle sales and the overall 
fleet makeup; consumers may respond to new fuel-economy-improving 
technologies on certain models by choosing to buy other models, 
especially when fuel costs are not expected to increase significantly 
in the ownership timeframe and consumers value other vehicle attributes 
more than they value fuel economy. Either of these responses may cause 
manufacturers both to lose money and to face further difficulties in 
meeting the CAFE standards. While there are significant benefits for 
both manufacturers and consumers under attribute-based standards, 
manufacturers must still sell enough ``target-beaters'' to balance out 
sales of less-fuel-efficient vehicles and meet their overall fleet-
average compliance obligations. If consumer demand shifts strongly away 
from target-beaters, CAFE compliance will be a struggle, even if the 
target-beaters are widely available. Section IV above discusses this 
phenomenon in more detail. And if consumers buy fewer new vehicles in 
response to per-vehicle cost increases, which the agencies are 
beginning to see already, \2973\ the fleet as a whole will turn over 
more slowly, and fuel conservation gains may also be slowed. NHTSA does 
not believe that that is EPCA's goal. Manufacturers struggling to sell 
new vehicles will have less capital to devote to further technological 
improvements; may choose to move manufacturing jobs outside the U.S. to 
places with lower labor costs; and so forth. A net benefits analysis 
may be informative to attempting to quantify some of the issues 
described above, but not all of these issues lend themselves to clear 
quantification. The following discussion will evaluate what the 
agencies believe has been reasonably accounted for.
---------------------------------------------------------------------------

    \2973\ See, e.g., Jackie Charniga, ``Prime buyers flood used-
vehicle market in Q4,'' Automotive News, March 4, 2020, https://www.autonews.com/finance-insurance/prime-buyers-flood-used-vehicle-market-q4.
---------------------------------------------------------------------------

(a) Per-Vehicle Costs, Sales, and Employment as Part of Economic 
Practicability
    Per-vehicle cost estimates are relevant to NHTSA's consideration of 
economic practicability because, when cost increases associated with 
more stringent standards are passed through to consumers as price 
increases, they affect consumers' willingness and ability to purchase 
new vehicles, and thus influence vehicle sales and fleet turnover. A 
similar effect occurs in reverse when stringency is decreased. Table 
VIII-7 below shows the estimated effects on per-vehicle costs by 
regulatory alternative in MY 2029:
[GRAPHIC] [TIFF OMITTED] TR30AP20.736

    Generally speaking, per-vehicle costs increase as stringency 
increases. The agencies estimate that, by MY 2029, costs for additional 
fuel-saving technology (beyond that present on vehicles in MY 2017) 
would average about $2,800 under the augural CAFE standards, as 
compared to about $1,400 under the proposed CAFE standards,

[[Page 25176]]

and about $1,650 under the final CAFE standards for MYs 2021-2026. The 
next most stringent alternative beyond the 1.5 percent alternative is 
the ``2%/3%'' alternative. Under 2%/3%, the agencies estimate that 
costs would increase by $2,000 per vehicle on average. NHTSA 
understands that many readers may not find an extra $350 per vehicle to 
be a compelling reason to reject the 2%/3% alternative, or even find an 
additional $1,125 per vehicle a reason to reject the baseline/augural 
standards. As the NPRM discussed, ``. . . the corresponding up-front 
and monthly costs may pose a challenge to low-income or credit-
challenged purchasers. . . . such increased costs will price many 
consumers out of the market--leaving them to continue driving an older, 
less safe, less efficient, and more polluting vehicle, or purchasing 
another used vehicle that would likewise be less safe, less efficient, 
and more polluting than an equivalent new vehicle.'' \2974\ This 
continues to be a concern: For example, the average MY 2025 prices 
estimated here under the baseline, final, and 2%/3% CAFE standards are 
about $38,100, $36,850, and $37,150, respectively. The buyer of a new 
MY 2025 vehicle might thus avoid the following purchase and first-year 
ownership costs under the final standards as compared to the baseline 
standards or 2%/3% standards:
---------------------------------------------------------------------------

    \2974\ 83 FR at 43222 (Aug. 24, 2018).
    [GRAPHIC] [TIFF OMITTED] TR30AP20.737
    
    While the buyer of the average vehicle would also purchase somewhat 
more fuel under the final standards than the baseline standards, this 
difference might average less than four gallons per month during the 
first year of ownership. Some purchasers may consider it more important 
to avoid these very certain (e.g., being reflected in signed contracts) 
cost savings than the comparatively uncertain (because, e.g., some 
owners drive considerably less than others, and may purchase fuel in 
small increments as needed) fuel savings. For some low-income 
purchasers or credit-challenged purchasers, the cost savings may make 
the difference between being able or not to purchase the desired 
vehicle. As vehicles get more expensive in response to higher CAFE 
standards, it will get more and more difficult for manufacturers and 
dealers to continue creating loan terms that both keep monthly payments 
low and do not result in consumers still owing significant amounts of 
money on the vehicle by the time they can be expected to be ready for a 
new vehicle. These considerations were discussed in the NPRM and they 
remain true for this final rule.
---------------------------------------------------------------------------

    \2975\ Edmunds estimates that the average down payment for a new 
vehicle in 2019 was 11.7% of the vehicle's price, see https://www.edmunds.com/car-buying/how-much-should-a-car-down-payment-be.html.
---------------------------------------------------------------------------

    Per-vehicle cost and fuel economy both affect sales estimates in 
the final rule analysis. Table VIII-9 below shows the estimated effects 
on fleet-wide sales by regulatory alternative from 2017-2030, where the 
augural standards represent absolute sales and all other alternatives 
represent increases relative to the augural sales:
BILLING CODE 4910-59-P

[[Page 25177]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.738

BILLING CODE 4910-59-C
    The final rule analysis indicates that industry sales decrease as 
stringency increases, and increase as stringency decreases. While sales 
under both the

[[Page 25178]]

proposal and the final rule are comparable, each represents about a 1.5 
percent reduction in total sales over the period from 2017--2030. In 
the context of 16-17 million new vehicle sales annually, NHTSA does not 
believe that the sales volume effects here, while significant, are 
necessarily determinative for economic practicability, even after 
accounting for fuel economy effects in the sales analysis as some 
commenters recommended. That said, NHTSA recognizes that the final rule 
sales analysis does not account for a number of factors that could 
cause differences between alternatives to result in changes in new 
vehicle sales (perhaps greater). For example, as explained above, NHTSA 
remains concerned that significant increases in fixed upfront prices 
(which for many people translate to monthly financing costs) are harder 
for certain segments of new vehicle buyers to manage than fuel costs, 
which can be managed to some extent through vehicle switching or travel 
decisions. The sales analysis for this final rule indicates that more 
stringent standards tend to result in higher light truck sales and 
lower passenger car sales. While NHTSA does not have specific 
information (or a vehicle choice model) to inform the agency about 
which consumers (by income) buy which vehicles, and while NHTSA 
acknowledges that it does not account for price cross-subsidization by 
manufacturers to keep ``entry-level'' new vehicle (often, passenger 
car) prices low, NHTSA continues to be concerned about the possibility 
of a bubble in the market for new vehicles. As the Wall Street Journal 
reported in November 2019, ``Some 33% of people who traded in cars to 
buy new ones in the first nine months of 2019 had negative equity, 
compared with 28% five years ago and 19% a decade ago, according to 
car-shopping site Edmunds . . . . Rising car prices have exacerbated an 
affordability gap that is increasingly getting filled with auto debt.'' 
\2976\ The sales analysis for this final rule does not directly account 
for these effects, but NHTSA is concerned that they may be 
considerable. NHTSA notes that this analysis does not take into account 
potential economic turmoil or recession, which may have a significant 
impact on vehicle sales and industry viability.\2977\
---------------------------------------------------------------------------

    \2976\ AnnaMaria Andriotis and Ben Eisen, ``A $45,000 Loan for a 
$27,000 Ride: More Borrowers are Going Underwater on Car Loans,'' 
Wall Street Journal, November 9, 2019.
    \2977\ Letter from Alliance for Automotive Innovation, NADA, and 
MEMA to Congress, Mar. 23, 2020, available at https://www.autosinnovate.org/wp-content/uploads/2020/03/COVID-19-Letter-to-Congress-NADA-MEMA-AAI-March-23.pdf.
---------------------------------------------------------------------------

    The final rule analysis also looked at employment effects under the 
different regulatory alternatives. A number of commenters argued that 
more stringent standards improved employment opportunities, as shown in 
the NPRM analysis and in other analyses, due to the need for workers to 
manufacture the additional technology needed to meet those more 
stringent standards. Similar to the NPRM analysis, the agencies' 
updated analysis shows labor utilization, on balance, increasing 
slightly with stringency, as this effect outweighs the opposing effect 
of changes in vehicle sales. Table VIII-11 below shows the estimated 
effects on U.S. auto industry employment by regulatory alternative in 
MY 2029:
[GRAPHIC] [TIFF OMITTED] TR30AP20.739

[GRAPHIC] [TIFF OMITTED] TR30AP20.740

    It is important to note, however, that the reduction in person-
years described in this table merely reflects the fact that, when 
compared to the standards set in 2012, fewer jobs will be specifically 
created to meet infeasible regulatory requirements. It is also 
important to note that the $15 billion in avoided required technology 
costs (in MY 2029) can be invested by manufacturers into other areas, 
or passed on to consumers. Moreover, consumers can either take those 
cost savings in the form of a reduced vehicle price, or used toward the 
purchase of specific automotive features that they desire (potentially 
including a more-efficient vehicle or optional safety features that can 
reduce risk of injury or death for all vehicle occupants on the road), 
which would increase employment among suppliers and manufacturers.
    Generally speaking, the agencies' analysis shows net labor 
utilization increasing with stringency, because the additional labor 
utilization involved with producing additional fuel-saving

[[Page 25179]]

technology outweighs the foregone labor utilization involved with the 
foregone sales. As indicated above, for the scope of labor utilization 
accounted for in today's analysis, the agencies show about 1.20 million 
person-years under the augural CAFE standards and about 1.19 million 
person-years under either the proposed or final standards. As for 
sales, it is arguably instructive to consider these estimates in the 
broader context of U.S. employment. BLS data indicates that roughly 129 
million people in the U.S. are employed full-time at the time of 
writing,\2978\ and that roughly 1.4 million people were employed in 
motor vehicle and motor vehicle equipment manufacturing in 2018.\2979\ 
The agencies estimate that, compared to the augural standards, the 
final standards will reduce automotive labor utilization associated 
with production of the MY 2029 fleet by about 1.1%, a slightly smaller 
reduction than the 1.4% estimated to occur under the proposed 
standards. For comparison, the Synapse Report cited often by commenters 
concluded that vehicle standards result in ``nationwide employment 
increases of more than 100,000 in 2025 and more than 250,000 in 2035. . 
. these increases represent less than 0.2 percent of current U.S. 
employment levels.'' \2980\ Even at these levels, which NHTSA does not 
necessarily agree are accurate, the employment effects of standards are 
in the range of the average of more than 216,000 jobs added to the U.S. 
economy during each month of 2018.\2981\ That said, as for sales, NHTSA 
recognizes that the final rule labor utilization analysis does not 
account for a number of factors that could cause differences between 
alternatives to be different (perhaps greater), as discussed further 
below.
---------------------------------------------------------------------------

    \2978\ https://www.bls.gov/cps/cpsaat08.htm.
    \2979\ https://www.bls.gov/cps/cpsaat18b.htm.
    \2980\ https://www.synapse-energy.com/sites/default/files/Cleaner-Cars-and%20Job-Creation-17-072.pdf, at ES-2.
    \2981\ Payroll employment increased by 2.6 million jobs in 2018, 
an average of 216,667 per month. ``The Employment Situation--
December 2018,'' Bureau of Labor Statistics, available at: https://www.bls.gov/news.release/archives/empsit_01042019.pdf.
---------------------------------------------------------------------------

(b) Application Rates for New Technologies as Part of Economic 
Practicability
    The sales analysis for this final rule also does not account for 
the potential consumer acceptance issue of more stringent standards 
effectively requiring the application of technologies not yet ready for 
widespread deployment. As widely noted, the 2012 rule assumed extremely 
high penetration of dual-clutch transmissions in response to standards. 
While the agencies stated throughout that final rule that the analysis 
was not meant to represent the expected response to the standards, Ford 
did apply DCTs to a number of vehicle models in its fleet, that 
resulted in major customer satisfaction issues and ultimately caused 
extensive buyback campaigns, customer service programs, and class-
action litigation.\2982\ Sales can be impacted as a result of standards 
if technologies applied in response to those standards have 
operational, maintenance, or customer acceptance problems, or if 
consumers are unwilling to pay for it. Manufacturer capital to develop 
and add new technologies and manage these rollout issues is finite, as 
discussed. Insufficient capital can also cause quality problems. The 
cost effects modeled in this final rule analysis, that drive the sales 
and scrappage analyses, only include technology costs and RPE--they do 
not include the cost of stranded capital or lost consumer surplus, 
which are things that could drive up costs, drive down benefits, and 
therefore impact sales and scrappage beyond what today's analysis 
shows.
---------------------------------------------------------------------------

    \2982\ See https://www.autonews.com/technology/dual-clutch-gearbox-complaints-haunt-ford.
---------------------------------------------------------------------------

    As Section IV above notes, a great deal of fuel economy-improving 
technology has already been added to the fleet since 2012, which means 
that the amount of fuel economy-improving technology left to be added 
in response to higher standards is less than it was assumed to be in 
2012. Looking at the technology penetration rates modeled in today's 
analysis, it appears that the augural standards are projected to 
require nearly 20 percent total electrification in MY 2029, while the 
proposal would have required nearly 7 percent and the final standards 
would require nearly 8 percent. Table VIII-11 below shows projected 
electrification rates by 2029 for the regulatory alternatives--
electrification refers to all models with strong hybrids, PHEVs, or 
full EVs:
[GRAPHIC] [TIFF OMITTED] TR30AP20.741


[[Page 25180]]


    As the table shows, the analysis projects that meeting the augural 
standards could require over twice as much electrification as the final 
rule standards could require.\2983\ The current market penetration for 
all such vehicles is only approximately 4 percent even though the 
technology is well-established, with hybrids having been first 
introduced with the Honda Insight in 1999 and Toyota Prius in 2000, 
plug in hybrids with the Chevrolet Volt in late-2010 and electric 
vehicles with the Tesla Roadster in 2008 and Nissan Leaf in late 2010. 
As Mr. Kreucher commented, and as Figure VIII-2 shows, consumers appear 
to be driven by fuel price. Given anticipated fuel prices during this 
timeframe and evidence in the market today of cannibalization within 
these vehicle segments (not to mention the continued phasing out of 
government incentives for these vehicles),\2984\ NHTSA is concerned 
that there could be consumer acceptance problems associated with 
further electrification under more stringent alternatives, which could 
have sales impacts.
---------------------------------------------------------------------------

    \2983\ While NHTSA is prohibited by statute from considering 
battery electric vehicles as a compliance mechanism, we are aware 
that many OEMs will likely opt to produce a smaller number of fully 
electric vehicles rather than a large number of strong hybrid 
models.
    \2984\ 26 U.S.C. Section 30D provides for tax credits ranging 
from $2,500 to $7,500 for purchasers of qualifying plug-in hybrid 
(PHEV) and battery electric (BEV) vehicles, with a phaseout applying 
to vehicle manufactured by an automaker once they sell 200,000 
qualifying vehicles. Both Tesla and General Motors have reached this 
threshold and the tax credit applicable to purchasers of new PHEV 
and BEV vehicles from those manufacturers has been reduced gradually 
and will phase out completely on January 1, 2020 for Tesla, and 
April 1, 2020 for General Motors.
    The California Clean Vehicle Rebate Project was launched in 2010 
to provide incentives of up to $5,000 for purchasers or lessees of 
qualifying PHEV, BEV, and certain other alternative fuel vehicles. 
Since then, the program has undergone significant changes, including 
the addition of income eligibility criteria for certain incentives, 
and excluding eligibility toward the purchase or lease of a vehicle 
with an MSRP exceeding $60,000.
    Separately, in 2005, California passed a law allowing hybrid 
electric vehicle (HEV), plug in hybrid electric vehicle (PHEV), and 
battery electric vehicle (BEV), and other qualifying alternative 
fuel vehicle owners to apply for a sticker allowing single-occupant 
access to High Occupancy Vehicle (HOV) lanes. HEV access was phased 
out in 2011, with eligibility being limited to PHEV, BEV and other 
qualifying alternative fuel vehicle owners. Access is now limited to 
a four-year period, and only to individuals who do not receive a 
rebate under the California Clean Vehicle Rebate Project (unless 
meeting income eligibility requirements).
---------------------------------------------------------------------------

    We underscore that the table above simply shows the analytical 
results of the modeling for today's final rule based upon the most 
cost-effective means of achieving a given standard--it does not show 
how manufacturers would, or could, comply with the CAFE standards 
represented by the different regulatory alternatives. The discussion 
below covers the topic of manufacturer compliance shortfalls, and this 
discussion and that one are connected: The final rule analysis does not 
show significant compliance shortfalls under any regulatory 
alternative, but NHTSA believes that this is in large part because the 
CAFE model is not programmed with assumptions about consumer acceptance 
of strong hybrid technologies. In effect, the model lets manufacturers 
lean on hybridization to achieve compliance at a lower cost than if 
manufacturers instead pursued, for example, more advanced engine 
technologies. If cost-effectiveness is the only concern, that may be a 
valid compliance choice. If consumer acceptance of hybrid vehicles is 
accounted for, especially in a time of foreseeably low fuel prices, it 
may not be a valid compliance choice.
[GRAPHIC] [TIFF OMITTED] TR30AP20.742


[[Page 25181]]


    As Figure VIII-2 illustrates, the market share of strong hybrids in 
the new vehicle market has mostly tracked fuel prices. The bars 
represent the market share (left axis) and the line tracks the price of 
fuel (on the right axis). The light numbers inside of each bar 
represent the number of unique strong hybrid models offered for sale in 
that year. Initially, we see rapid growth that continues during the 
fuel price increases of the mid-2000s and peaking at around 3.5 percent 
market share. The figure shows that neither the passage of time, where 
consumers become more familiar with the technology over successive 
vehicle purchases, nor the number of models offered for sale have much 
of an impact on the market share for strong hybrids. Despite a doubling 
of the number of models offered for sale in subsequent years, market 
share continued to track fuel price closely, and fell dramatically as 
prices fell in 2015 and 2016. At fuel prices at or above $3.50/gallon, 
strong hybrids were able to capture additional market share. However, 
the current projection does not show prices returning to those levels 
for quite some time--leaving manufacturers uncertain about their 
ability to sell strong hybrids in the numbers estimated to be needed to 
comply with CAFE and CO2 standards before MY 2026.
    The agencies conducted a sensitivity analysis to evaluate the 
impact of compliance pathways that did not rely on strong hybrids (see 
Chapter 7 of the Final RIA). As we discuss in the sensitivity analysis, 
in the absence of strong hybrids, compliance pathways tend toward a 
greater reliance on advanced engines and transmissions, and more 
aggressive exploitation of opportunities to reduce vehicles' mass. 
These alternative technology pathways carry with them additional 
technology costs that increase compliance costs in the baseline and 
increase the savings associated with the preferred alternative.
    Under the CAFE program, where battery electric vehicles are not a 
compliance option (due to statutory restrictions on their consideration 
for rulemaking), the additional cost of advanced engine technology in 
the baseline increases baseline technology cost by about $800 per 
vehicle, and increases the cost savings under the preferred 
alternative, which has a much smaller reliance on strong hybrids to 
achieve compliance, by about $600 per vehicle. This difference is 
sufficient to change the sign on net social benefits for the preferred 
alternative to being slightly negative, to being very positive (nearly 
$80 billion at a 3 percent discount rate). The magnitude of this impact 
is comparable to the impact of varying fuel price projections.
    As shown in, Figure VIII-2 even the preferred alternative requires 
levels of strong hybridization (and PHEV share) that would be about 
twice what has been observed at the market, even at its peak. Both the 
baseline and the 2%/3% alternative have even greater reliance on 
hybridization--more than twice as much in the baseline. The compliance 
costs associated with each alternative in today's rule depend upon the 
estimated levels of hybridization in the compliance scenarios being 
possible to achieve in the new vehicle market. The sensitivity analysis 
shows that manufacturers can still reach comparable levels of fuel 
economy without additional reliance on hybridization, but at 
significantly higher per-vehicle costs. Those higher costs have 
implications for the sales response, vehicle retirement rates in the 
existing vehicle population, and the penetration rate of emerging 
safety features.
(c) Consumer Demand as Part of Economic Practicability
    As discussed above, NHTSA's consideration of consumer demand as 
relevant to economic practicability has been upheld by the D.C. Circuit 
in Center for Auto Safety v. NHTSA. A number of commenters argued that 
consumers do, in fact, demand more fuel economy than the NPRM analysis 
assumed; that consumers will appreciate more widespread application of 
fuel economy-improving technologies that NHTSA appears to believe they 
will tolerate; that NHTSA was wrong to assume that fuel prices will 
remain relatively low in the future and continue to dampen consumer 
demand for fuel economy; and that vehicle manufacturers will not make 
tradeoffs between investments in fuel economy improvements and 
investments in other vehicle characteristics which consumers also 
demand, such that requiring manufacturers to meet more stringent 
standards will not impair consumer demand for new vehicles because less 
of those other characteristics will be available. Those commenters also 
often highlighted the CAS language stating that consideration of 
consumer demand may not undermine EPCA's goal of energy conservation.
    NHTSA agrees with commenters that some consumers seek out vehicle 
models with higher fuel efficiency, and notes that those consumers have 
increasing numbers of relatively high-efficiency vehicle models to 
choose from in the current new-vehicle market, as shown in the previous 
section. CAFE does not affect fuel economy improvements that are 
supported by consumer demand--market forces will take care of that. 
Instead, it specifically addresses fuel economy improvements that are 
not preferred by consumers, and the agency sets standards that require 
manufacturers to make fuel economy improvements that consumers are not 
otherwise seeking. Section IV.B.3 discusses at some length the fact 
that alternative powertrains and higher fuel-efficiency vehicle models 
have proliferated widely since 2011--consumers no longer lack for 
choice if fuel economy is what they want. NHTSA's concern regarding 
consumer demand is that in an era of relatively low gasoline prices--as 
EIA currently projects and NHTSA has no basis to second-guess, and 
which may be even lower than currently projected--it does not appear 
likely that the market for higher fuel-economy vehicles and alternative 
powertrains in particular will increase significantly in the rulemaking 
timeframe, beyond the 30-month payback period that the agencies 
currently use as a proxy for market demand for fuel economy. It is 
worth citing the CAS case at greater length here in light of its 
parallels: As the D.C. Circuit stated in that case,

    [T]he petitioners do not challenge the consideration of consumer 
demand per se, but rather the weight the agency has given the factor 
in downgrading standards when, they argue, the principal 
impracticability is paying a civil penalty [note: today, using or 
purchasing credits]. Until the model years at issue here, there has 
been little tension between consumer demand and the fuel 
conservation goals of EPCA. The agency now relies on market 
projections in a setting in which falling gas prices have relaxed 
consumer demand for fuel efficiency. Earlier consideration of 
consumer demand in setting standards could not have alerted Congress 
to the agency's current application of this factor. Because Congress 
has not spoken clearly on the issue before us, it must be determined 
whether the agency's interpretation represents a reasonable 
accommodation of the policies embodied in the statute.
    . . .
    The agency concluded that if manufacturers had to restrict the 
availability of larger trucks and engines in order to adhere to CAFE 
standards, the effects ``would go beyond the realm of `economic 
practicability' as contemplated in the Act.'' [Citation omitted.] 
The original projections of technological feasibility for the 1985 
model year standards were based on the assumption that gasoline 
prices would remain high and consumer demand for fuel-efficient 
vehicles would remain strong. No one disputes that actual 
circumstances have deviated from these assumptions. NHTSA acted 
within the reasonable range of interpretations of the statute in 
correcting the 1985 standards to

[[Page 25182]]

account for these changed conditions. Consideration of product mix 
effects was also reasonable in setting the standards for 1986, as 
there is no evidence that the same trends in consumer demand will 
not continue.
    . . .
    In short, while it may be disheartening to witness the erosion 
of fuel conservation measures in the face of changes in consumer 
priorities, this court is nonetheless compelled to uphold the 
agency's standards. They are the result of a balancing process 
specifically committed to the agency by Congress, and, in this case, 
the weight given to consumer demand was not outside the range 
permitted by EPCA.

    CAS, 793 F.2d 1322, 1340-41 (D.C. Cir. 1986). As in the situation 
presented in the CAS case, the agencies believed in 2012 based on the 
evidence then before them that fuel prices would be significantly 
higher than the fuel prices currently projected today. Using the fuel 
prices currently projected, which are lower because of the structural 
changes to the global oil market described at length above, Figure 
VIII-3 shows the difference in annual fuel consumption for a typical 
driver under the augural standards, proposed standards, and final 
standards. As the figure shows, the difference in annual consumption 
(for a user that drives 14K miles per year) \2985\ is fewer than 40 
gallons by MY 2030--the largest difference between the alternatives. 
Rising fuel prices over time increase the value of those forty gallons, 
but the diminishing returns to successive increases in fuel economy are 
nonetheless evident.\2986\
---------------------------------------------------------------------------

    \2985\ Parts of the central analysis assume a typical new 
vehicle is driven 14,000 miles per year, for each of the first three 
years it is owned. In practice, the average is slightly higher, 
through affected by a smaller number of users that drive much more 
than average. There is no single value that is representative of all 
households, and the National Household Travel Survey has shown lower 
annual usage estimates than 14,000 miles per year for a typical new 
vehicle.
    \2986\ In general, because fuel savings are subject to 
diminishing returns as CAFE standards become more stringent, and 
per-vehicle costs increase as CAFE standards become more stringent, 
the relationship between per-vehicle costs and the value of fuel 
savings is more of a curve than a line, although the slope of the 
curve is reduced by the fact that we rely on EIA's forecast of 
rising fuel prices over time.
[GRAPHIC] [TIFF OMITTED] TR30AP20.743

    Thus, on the supply side, greater and more stable global oil 
supply, which reduces projected fuel prices, means that the benefits of 
more stringent CAFE standards are lower than they appeared to be in 
2012 when the agencies believed oil supply would be scarcer and less 
stable, and projected fuel prices were consequently higher.
    On the demand side, as already explained, while NHTSA agrees that 
some consumers do seek out higher fuel economy, those consumers have 
vastly more higher fuel-economy-vehicle options than they did when the 
agencies wrote the 2012 final rule, as shown in Section IV above. For 
the other consumers who are driven more by the economics of their 
vehicle-purchasing decisions, NHTSA believes that they are likely 
making reasonably informed decisions about the new vehicle attributes 
they want in light of expectations about future fuel costs. This can be 
illustrated by examining estimated payback periods under the different 
regulatory alternatives, because payback period directly compares 
estimated future fuel savings with estimated vehicle purchase and 
ownership costs. A number of commenters suggested that per-vehicle cost 
was not a meaningful metric in isolation, because consumers would also 
be saving money on fuel under more stringent standards. The agencies 
discuss affordability issues further below, but the rulemaking presents 
Table VIII-12 here as a comparison of per-vehicle costs to lifetime 
fuel savings to illustrate the point raised by commenters:

[[Page 25183]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.744

    Table VIII-12 shows the differences in regulatory costs, other 
registration costs (taxes and financing, though the cost of insurance 
also increases to cover more expensive vehicles), lifetime fuel 
savings, and the payback relative to a MY 2017 vehicle. It is important 
to compare apples to apples, so in this case, because the agencies are 
considering fuel costs over a vehicle's full lifetime, this rulemaking 
needs to compare that against a broader lifetime cost of ownership, 
instead of comparing it simply to the estimated increase in initial 
purchase price. Under the augural standards, the analysis projects that 
it would take a full five years for the undiscounted value of fuel 
savings to offset the estimated upfront increase in purchase cost 
(relative to a MY 2017 vehicle). For reference, the average new car 
buyer holds on to that car for about six or seven years.\2987\ 
Naturally, this payback period, and the fuel savings on which it is 
based, depend upon fuel prices. Higher fuel prices shorten payback 
periods, while declining fuel prices lengthen them. For this analysis, 
the agencies have employed fuel prices estimated using the version of 
NEMS used to produce AEO 2019, as discussed in Section VI.
---------------------------------------------------------------------------

    \2987\ IHS Markit estimates the average length of new vehicle 
ownership at about 79 months, see https://www.forbes.com/sites/jimgorzelany/2018/01/12/the-long-haul-15-vehicles-owners-keep-for-at-least-15-years/#4e971b576237.
---------------------------------------------------------------------------

    Thus, all of the regulatory alternatives considered in today's 
analysis result in significantly longer payback periods than the 2.5 
years assumed by the agencies, the industry, and the NAS--i.e., while 
fuel economy would foreseeably improve in the rulemaking timeframe in 
the absence of regulation, it would do so at a rate slower even than 
the proposal would have required.\2988\ NHTSA thus does not expect that 
consumer demand for fuel-efficient vehicles will grow significantly in 
the rulemaking timeframe without regulation to prop up manufacturer 
sales of significantly larger volumes of more fuel-efficient models. 
This increases the economic practicability of regulatory alternatives 
that represent less stringent standards, as compared to those that 
represent more stringent standards.
---------------------------------------------------------------------------

    \2988\ While presented at the industry level, technology 
application and compliance simulation occur at the level of each 
individual manufacturer's respective fleets. Some OEMs and fleets 
are able to increase CAFE more easily than others--starting from 
more favorable positions and adding less expensive technology, or 
taking advantage of credit provisions, to improve the fuel economy 
of their fleets. However, for several OEMs, even the proposed 
standards are binding, and the costs associated with bringing their 
fleets into compliance are significant. At the level of the industry 
average, the cost of compliance with the proposal--and as a 
corollary, with the other alternatives--exceeds the 2.5 year payback 
for fuel economy technology, even while a small amount of 
overcompliance occurs at the industry level.
---------------------------------------------------------------------------

(d) Manufacturer Compliance Shortfalls as Part of Economic 
Practicability
    Manufacturer compliance shortfalls given the pace of increase in 
standard stringency over time are also relevant to economic 
practicability, and were considered as part of the 2012 final rule. 
Some commenters argued that it was not reasonable for NHTSA to 
interpret automakers' fuel economy improvements over time as evidence 
that less stringent standards might be maximum feasible, suggesting 
that evidence of improvements means that improvements are possible, and 
that automakers' stated difficulties with meeting more stringent 
standards may be overstated. Fleet fuel economy improvements over time 
have been possible, NHTSA agrees. NHTSA does not agree, however, that 
improvements thus far constitute de facto evidence of automakers' 
ability to meet rapidly increasing standards indefinitely into the 
future. Section IV above illustrates this clearly--many more very fuel-
efficient models are available now than in 2012, while fuel prices have 
been trending downward on an absolute basis over the same time period. 
Simultaneously and relatedly, the rate at which various manufacturer 
fleets have been falling short of their standards has been increasing 
steadily. As Section IV explains, at the time of the 2012 analysis, 
most manufacturers were in reasonable shape in terms of compliance. The 
total fleet outperformed CAFE standards by a full mile per gallon--
reflecting the historical trend that the full fleet always exceeds

[[Page 25184]]

the average fuel economy target.\2989\ Of the then 45 import passenger 
car, domestic passenger car, and light truck compliance fleets in the 
2012 model year, 26 of the fleets exceeded their fuel economy targets, 
while 19 failed to meet their standard.\2990\ Of those 19 fleets that 
failed to meet their standard, the total shortfall was 41,033,802 
credits--the equivalent of $225,685,911 in penalties.\2991\ That is no 
longer the case. 2016 marked the first model year in CAFE history that 
the entire light duty fleet failed to meet its target.\2992\ This 
continued in the 2017 model year (the most recent full model year of 
compliance data).\2993\ In the 2017 model year, of the now 42 
compliance fleets, only 14 fleets exceeded their targets.\2994\ 25 
failed to meet their target, with a total shortfall of 166,715,863 
credits--the equivalent of $1,133,430,584 in penalties.\2995\ Required 
manufacturer reporting data shows the situation continuing to get worse 
in the 2018 and 2019 model years,\2996\ despite manufacturers' 
increasing ability to utilize generous credit provisions related to 
alternative fueled vehicles and A/C efficiency and off-cycle 
adjustments.
---------------------------------------------------------------------------

    \2989\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed Dec. 27, 
2019.
    \2990\ NHTSA MY 2011-2019 Industry CAFE Compliance, https://one.nhtsa.gov/cafe_pic/MY%202011-MY_2019_Credit_Shortfall_Report_v08.pdf.
    \2991\ Id. While we denominate shortfalls in terms of credits, 
that is simply for convenience, and any given manufacturer's 
shortfall is measured in tenths of a mile per gallon for compliance 
purposes.
    \2992\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed Dec. 27, 
2019.
    \2993\ Id.
    \2994\ NHTSA MY 2011-2019 Industry CAFE Compliance, https://one.nhtsa.gov/cafe_pic/MY%202011-MY_2019_Credit_Shortfall_Report_v08.pdf.
    \2995\ Id.
    \2996\ Id.
---------------------------------------------------------------------------

    Although each year has continued to see improvements in fuel 
economy performance, each successive increase in stringency requires 
many fleets not only to achieve the new level from the resulting 
increase, but to resolve deficits from the prior year as well. The 
problem is particularly marked in the light truck fleet, where sales of 
lower fuel-economy vehicles have proliferated over this time period, 
despite availability of higher fuel-economy models. But the passenger 
car fleet is facing compliance challenges as well, as more consumers 
have shifted away from sedans and into crossover utility vehicles that 
are considered passenger cars for compliance purposes. While the 
agencies' move toward footprint based standards account for vehicle 
length and track width--which certainly affect fuel economy as 
described above--they do not account for mass-intensive increases in 
vehicle ride height that crossover purchasers value, the additional 
frontal area and higher drag at highway speeds, or the additional power 
required to achieve similar performance as the equivalent sedan. These 
issues are further exacerbated by the fact that consumers are demanding 
more powerful engines than the baseline efficient four cylinder 
versions the agencies assumed consumers would find acceptable, instead 
opting to upgrade to more powerful powertrains.\2997\ If the augural 
standards were finalized and energy prices remain as currently 
projected, the shortfall situation could well erase large portions of 
assumed fuel savings/emissions reduction benefits from higher 
standards.
---------------------------------------------------------------------------

    \2997\ Mr. Rykowski's comments for EDF, for example, stated that 
EPA's recent Fuel Economy and CO2 Trends Reports show 
clearly that manufacturers have been improving vehicle performance 
at the expense of fuel economy. See NHTSA-2018-0067-12018, at 31.
---------------------------------------------------------------------------

    In the current analysis, gasoline prices are projected to rise 
steadily from about $2.50/gallon in 2017 to $3.5/gallon by 2035. While 
CAFE can provide some insurance against unexpected and sudden price 
increases, in the case of sustained, consistent increases in gasoline 
prices, market demand for fuel economy would outpace the standards over 
time. In an earlier analysis, the agencies considered the impact of a 
sudden gasoline price shock in a single year, where the price of 
gasoline jumped from $3.50/gallon to $6/gallon for most of a year. If 
instead of that one-year spike, the price of gasoline rose steadily 
from current levels to $6/gallon by 2040, the response of both 
consumers and manufacturers in the marketplace would cause the industry 
to consistently over-comply with even the augural standards.\2998\ The 
payback assumption in this analysis, where consumers are willing to pay 
for any fuel economy improvement that pays for itself in the first 2.5 
years of vehicle usage, would likely be too short in a world with $6/
gallon gasoline, where the cost of operating a vehicle consumed a 
larger share of a household's budget and even longer payback periods 
could be seen as sound investments. Thus, if it turns out that fuel 
prices rise steadily over the next decade, at a significantly faster 
rate than currently projected, the market will end up demanding more 
efficient vehicles and the gap between the baseline and the preferred 
alternative will shrink further. However, the agencies do not currently 
have information that projects $6/gallon fuel in 2040 is likely, for 
the reasons discussed at length above.
---------------------------------------------------------------------------

    \2998\ We simulated this response in the CAFE Model, where all 
other inputs were identical to the central analysis.
---------------------------------------------------------------------------

    As also discussed above, while the analysis for this final rule 
does not show significant shortfalls under any regulatory alternative, 
that appearance of compliance is predicated on the assumption that 
automakers will be able to sell the hybrids that we simulate them 
producing in response to the standards. Again, given foreseeably low 
fuel prices going forward, it is also foreseeable that selling greater 
volumes of hybrid vehicles will be even more difficult than at present. 
It is very possible that manufacturer compliance shortfalls could end 
up being worse than the agency's analysis currently forecasts for the 
more stringent alternatives.
    Given the ongoing shortfall problem illustrated above, and given 
the payback period estimates, the proposal might appear to be the 
correct answer in the absence of other considerations. NHTSA believes 
that the bubble concerns may be significant, and the diminishing 
returns of higher standards identified in Section IV above calls into 
question the value of pushing that bubble. Compliance shortfalls 
represent a growing problem with the current standards and will 
continue to be a problem if stringency does not converge at least 
somewhat more closely with what the market appears willing to bear. If 
industry is unable to comply with standards, that non-compliance means 
that the standards are not achieving what they set out to achieve in 
terms of fuel savings or emissions reductions, or at least they are not 
achieving what NHTSA estimated they would achieve. The NPRM disagreed 
with the idea that ``if you build it, they will come''--that 
manufacturers would find a way to market higher fuel-economy vehicles, 
and consumers would eventually buy them. Comments on that topic were 
mixed: some commenters agreed with the NPRM's sentiment, while other 
commenters argued that manufacturers' past ability to exceed standards 
combined with consumers' growing interest in fuel economy/lower 
emissions meant that concerns about the market's ability to bear 
further increases were misplaced. The shortfall discussion above and in 
Section IV suggests that the NPRM's sentiment may be accurate, but this 
difference in perspective highlights the core philosophical question of 
the CAFE program--whether consumers should choose for themselves how 
much fuel

[[Page 25185]]

economy they want, or whether the government should choose for them.
(4) Considering Safety Along With the Other Factors in Determining 
Maximum Feasible Standards
    In addition to the above, as explained in the NPRM and as discussed 
extensively by commenters, NHTSA considers safety effects in 
determining maximum feasible CAFE standards. A number of commenters 
objected to aspects of the safety analysis, as discussed in Section VI 
above, and some made suggestions for improvement. In response to those 
comments, NHTSA took a very conservative approach in making a number of 
changes to the safety analysis for this final rule:
     Commenters disagreed with certain aspects of the sales and 
scrappage effects on the safety analysis; in response to those 
comments, changes have been made and the scrappage effect on fatalities 
is lower now than it was in the NPRM;
     Commenters disagreed with certain aspects of mass 
reduction; in response to those comments, changes have been made there;
     Commenters argued that additional technologies should be 
accounted for; in response to those comments, many of those 
technologies have been added;
     Commenters argued that the NPRM did not account for crash 
avoidance technologies; in response to those comments, the final rule 
accounts for the effects of crash avoidance technologies;
     Commenters argued that the NPRM did not account for the 
mortality/morbidity effects of criteria pollution differences between 
the alternatives; in response, the final rule accounts for these 
effects explicitly in these values.
    Overall, the final rule analysis suggests that fatalities may be 
lower than the NPRM analysis showed; injuries may be greater; and the 
safety effects overall are less than the NPRM suggested, but they are 
still significant. Less-stringent standards remain better for safety 
and are projected to save thousands of lives and prevent tens of 
thousands of hospitalizations, even if the amount by which they are 
better is lower than previously estimated.
    EPCA/EISA directs NHTSA to conserve energy and consider the need of 
the U.S. to conserve energy, while simultaneously directing NHTSA to 
set attribute-based standards whose outcome varies depending on what 
consumers choose to buy, and directing NHTSA to consider economic 
practicability. The greater the need of the U.S. to conserve energy, 
the more the government should decide for consumers how much fuel 
economy will be in their new vehicles. Based on the information before 
NHTSA in this final rule, NHTSA agrees with the commenters who 
suggested that increasing CAFE stringency can function as ``insurance'' 
against future oil price volatility, although as illustrated above, the 
short-term effects of that insurance may be relatively minor and the 
longer-term effects may be too uncertain to consider meaningfully. 
NHTSA also agrees that environmental considerations necessitate energy 
conservation, though the long-term benefits of emissions reductions 
(even accounting for the increased costs of delayed action) require 
consideration of the immediate costs to consumers, the industry, and 
the environment.
    Balancing all of the factors and issues identified above, NHTSA 
concludes that standards that increase at 1.5% per year are the maximum 
feasible for passenger cars and light trucks for MYs 2021-2026, based 
on the information currently before the agency. We recognize that more 
stringent standards, including the baseline/augural standards, could 
conserve more energy and might be technologically feasible (in the 
narrowest sense), but the additional incremental fuel savings, 
emissions reductions, and environmental benefits of higher standards is 
not significant enough to outweigh the immediate economic costs. There 
is still risk to the U.S. from circumstances outside our control that 
the CAFE program may be able to mitigate, but there must also be 
recognition of the limited extent to which this program can address 
that risk, certainly without exacerbating considerable challenges 
currently being faced by automakers, dealers, and consumers. Economic 
practicability would be best served by slower increases, as discussed 
above. And while these two factors weigh in different directions, NHTSA 
has discretion to accommodate conflicting statutory priorities in a 
reasonable manner. Beginning with MY 2021, the first MY addressed by 
this rule, Congress eliminated the obligation to increase FE standards 
ratably.\2999\ Thus, the appropriateness of an increase, if any, is 
within NHTSA's discretion based on its balancing of statutory 
factors.\3000\
---------------------------------------------------------------------------

    \2999\ Previously applied for MYs 2011-2020.
    \3000\ NHTSA also notes that it was expressly anticipated in the 
2012 final rule that the current rulemaking could determine that the 
augural standards were not maximum feasible. NHTSA stated that 
``Whether different alternatives may be maximum feasible can also be 
influenced by differences and uncertainties in the way in which key 
economic factors (e.g., the price of fuel and the social cost of 
carbon) and technological inputs could be assessed and valued. While 
NHTSA believes that our analysis for this final rule uses the best 
and most transparent technology-related inputs and economic 
assumption inputs that the agencies could derive for MYs 2017-2025, 
we recognize that there is uncertainty in these inputs, and the 
balancing could be different if the inputs were different. When the 
agency undertakes the future rulemaking to develop final standards 
for MYs 2022-2025, for example, we expect that much new information 
will inform that future analysis, which may potentially lead us to 
choose different standards than the augural ones presented today.'' 
(emphasis added) 77 FR at 63037 (Oct. 15, 2012).
---------------------------------------------------------------------------

    In past rulemakings, as discussed above, NHTSA has expressly 
considered the point at which net benefits appear to be maximized as 
potentially relevant to determining maximum feasible CAFE 
standards.\3001\ Whether the standards maximize net benefits has thus 
been a significant, but not dispositive, factor in the past for NHTSA's 
consideration of economic practicability. Executive Order 12866, as 
amended by Executive Order 13563, states that agencies should ``select, 
in choosing among alternative regulatory approaches, those approaches 
that maximize net benefits . . .'' In practice, however, NHTSA must 
consider that the modeling of net benefits does not capture all 
considerations relevant to the EPCA statutory factors. Additionally, 
nothing in EPCA or EISA mandates that NHTSA set standards at the point 
at which net benefits are maximized, and case law confirms that whether 
to maximize net benefits in determining maximum feasible standards is 
within NHTSA's discretion.\3002\ As explained extensively in prior 
rulemakings, even if the agency believed it could quantify enough 
relevant factors to determine the CAFE

[[Page 25186]]

levels at which net benefits were maximized with reasonable accuracy, 
there may be other considerations which lead the agency to conclude 
that maximum feasible CAFE standards are not the ones that maximize net 
benefits. For example, in 2012, NHTSA rejected the regulatory 
alternative that appeared to maximize net benefits (and all 
alternatives more stringent than that one) based on the conclusion that 
even though net benefits were maximized, the ``resultant technology 
application and cost'' were simply too high, and thus made those 
standards economically impracticable, and thus beyond maximum 
feasible.\3003\
---------------------------------------------------------------------------

    \3001\ See, e.g., the 2006 final rule, which concluded that the 
point at which net benefits were maximized was the maximum feasible 
CAFE level (71 FR 17566 (Apr. 6, 2006)); the 2010 final rule, which 
considered among the regulatory alternatives one that maximized net 
benefits, but explained that nothing in EPCA or EISA mandated that 
NHTSA choose CAFE standards that maximize net benefits (75 FR 25324, 
at 25606, 25167 (May 7, 2010)); and the 2012 final rule, which also 
considered among the regulatory alternatives one that maximized net 
benefits, and also explained that nothing in EPCA or EISA mandated 
that NHTSA choose CAFE standards that maximize net benefits, in 
fact, directly rejecting the regulatory alternative that maximized 
net benefits as beyond maximum feasible for the MYs 2017-2025 
timeframe (77 FR 62624 (Oct. 15, 2012)).
    \3002\ The Ninth Circuit has agreed with NHTSA that ``EPCA 
neither requires nor prohibits the setting of standards at the level 
at which net benefits are maximized,'' stating further that ``The 
statute is silent on the precise question of whether a marginal 
cost-benefit analysis may be used. See Chevron, 467 U.S. at 843, 104 
S.Ct. 2778. Public Citizen and Center for Auto Safety persuade us 
that NHTSA has discretion to balance the oft-conflicting factors in 
49 U.S.C. 32902(f) when determining ``maximum feasible'' CAFE 
standards under 49 U.S.C. 32902(a).'' CBD v. NHTSA, 538 F.3d 1172, 
1188 (9th Cir. 2008).
    \3003\ 77 FR at 63050 (Oct. 15, 2012).
---------------------------------------------------------------------------

    Table VII-95 and Table VII-96, above, appear to suggest that net 
benefits would be maximized under a 3 percent discount rate by choosing 
the 2%/3% alternative, and under a 7 percent discount rate by choosing 
the 0% (proposed) alternative. Across all alternatives under either 
discount rate, the variation in net benefits is within $20 billion over 
the lifetimes of vehicles produced during the rulemaking timeframe. 
While $20 billion may seem like a large amount of money, it must be 
understood within context--the auto industry accounted for 
approximately $89 billion of U.S. GDP in 2018 alone,\3004\ and 
Americans spent approximately $370 billion on gasoline in 2019 
alone.\3005\ For a program this large, if the difference between the 
net benefits created by different regulatory alternatives is within $20 
billion (over the full lifetimes of six model years), the net benefits 
are relatively small. Furthermore, given how close together the net 
benefits are across the range of regulatory alternatives considered, 
NHTSA does not believe that the point at which net benefits are 
maximized is meaningful for determining maximum feasible CAFE standards 
in this final rule.
---------------------------------------------------------------------------

    \3004\ See Bureau of Economic Analysis, GDP by Industry, ``Value 
Added by Industry,'' Oct. 29, 2019, https://apps.bea.gov/iTable/iTable.cfm?ReqID=51&step=1 (accessed Mar. 18, 2020)
    \3005\ Using EIA estimates of an average of $2.60/gallon 
gasoline cost in 2019 (https://www.eia.gov/todayinenergy/detail.php?id=42435) and EIA estimates of about 142 billion gallons 
total gasoline consumed (https://www.eia.gov/tools/faqs/faq.php?id=23&t=10).
---------------------------------------------------------------------------

    Important to that conclusion is the fact that the net benefits 
estimates produced by the analysis depend heavily on EIA's future 
forecasts of fuel prices, which were made prior to the recent collapse 
of oil prices. If the former OPEC+ members continue to pursue market 
share, fuel prices will likely continue to drop. If, instead of 
pursuing market share, they try to control prices by restricting 
supply, U.S. shale production can ramp back up and exert downward 
pressure on price. If fuel prices end up even lower than our analysis 
assumes, benefits from saving additional fuel will be worth even less 
to consumers. Our analysis captures none of these effects. Depending 
upon future fuel prices, net benefits estimates described above could 
foreseeably be overstated, possibly by a significant amount. It is 
possible, depending on future fuel prices, that the final rule 1.5 
percent annual increase standards could end up being more stringent 
than standards that would maximize net benefits. Moreover, sustained 
low oil prices can be expected to have real effects on consumer demand 
for additional fuel economy, which will have real effects on sales, 
jobs, and many other things relevant to NHTSA's consideration of what 
standards would be maximum feasible. Choosing a regulatory alternative 
more stringent than the final rule's 1.5 percent annual increases could 
foreseeably either lead to more hybridization than the market is likely 
to bear given foreseeably low fuel prices, or lead to significantly 
more cost than the analysis currently suggests. Neither of those 
outcomes would be beneficial for consumers or for industry, even 
considering the additional fuel savings for consumers.\3006\
---------------------------------------------------------------------------

    \3006\ It is within NHTSA's discretion to adopt an alternative 
based on unquantified/unquantifiable benefits. See, e.g., Inv. Co. 
Inst. v. Commodity Futures Trading Comm'n, 720 F.3d 370, 379 (D.C. 
Cir. 2013) (``The appellants further complain that CFTC failed to 
put a precise number on the benefit of data collection in preventing 
future financial crises. But the law does not require agencies to 
measure the immeasurable. CFTC's discussion of unquantifiable 
benefits fulfills its statutory obligation to consider and evaluate 
potential costs and benefits. See Fox, 556 U.S. at 519, 129 S.Ct. 
1800 (holding that agencies are not required to `adduce empirical 
data that' cannot be obtained). Where Congress has required 
`rigorous, quantitative economic analysis,' it has made that 
requirement clear in the agency's statute, but it imposed no such 
requirement here. American Financial Services Ass'n v. FTC, 767 F.2d 
957, 986 (DCCir.1985); cf., e.g., 2 U.S.C. 1532(a) (requiring the 
agency to `prepare a written statement containing . . . a 
qualitative and quantitative assessment of the anticipated costs and 
benefits' that includes, among other things, `estimates by the 
agency of the [rule's] effect on the national economy').''); 
BellSouth Corp. v. FCC, 162 F.3d 1215, 1221 (D.C. Cir.1999) (`When . 
. . an agency is obliged to make policy judgments where no factual 
certainties exist or where facts alone do not provide the answer, 
our role is more limited; we require only that the agency so state 
and go on to identify the considerations it found persuasive').''
---------------------------------------------------------------------------

    NHTSA concludes that steady increases at 1.5 percent annually, with 
the same rate for cars and trucks as suggested by several commenters, 
are the optimal way to move the needle forward on fuel economy, fuel 
savings, and emissions reductions without imposing excessive cost on 
automakers and consumers and overly reducing vehicle sales. Requiring 
demand changes (through CAFE standards) much faster than what the 
market will bear creates a substantial likelihood of a mis-match 
between what companies produce and what consumers buy. While companies 
can manage that mis-match for short periods through incentivization and 
cross-subsidization, we have seen that over time automakers begin to 
fall short on fuel economy performance relative to the standards. Over 
time, if swaths of the industry continually fall short of fuel economy 
targets, and consumer demand for fuel economy does not significantly 
increase, then continuing to force technology into the fleet does not 
achieve the program's objectives (i.e., energy conservation). This is 
the case regardless of how much manufacturers spend manufacturing 
vehicles that consumers do not purchase (implicating concerns with 
economic practicability) to reduce their compliance liability. This is 
one part of why NHTSA believes that the 1.5 percent alternative is 
maximum feasible during the rulemaking timeframe.
    While the 1.5 percent alternative being finalized is new for the 
final rule, it is responsive to comments requesting steady increases at 
the same rate for both cars and trucks, and it is within the range of 
rates of increase considered in the NPRM. As both the NPRM analysis and 
the final rule analysis show, after MY 2020 the proposed (0%) standards 
are not binding at the industry level (though some manufacturers, and 
fleets, remain below their standard after that model year) as a 
consequence of market demand for fuel economy at projected gasoline 
prices. However, the preferred (1.5% percent) alternative, while 
producing slightly higher achieved CAFE levels, tracks closely to the 
level produced by the combination of existing CAFE standards (through 
MY 2020) and subsequent market demand for fuel economy represented by 
the proposal. It is also likely close to the point at which net 
benefits will be maximized, even if it remains unclear exactly where 
that point will end up.
    As a kind of insurance policy against future fuel price volatility, 
standards that increase at 1.5 percent per year for cars and trucks 
will help to keep fleet fuel economy higher than they would be 
otherwise when fuel prices are low, which is not improbable over the 
next several years.\3007\ These standards will

[[Page 25187]]

also enable industry to choose how to spend the capital that would 
otherwise be spent meeting more stringent standards on more of what 
consumers are demanding, which could also include more fuel economy if 
the market heads unexpectedly in that direction. As explained above, 
even if more stringent standards might be technologically feasible in a 
narrow sense, and even if the effect of other motor vehicle standards 
of the Government does not vary significantly between regulatory 
alternatives, economic practicability concerns still counsel against 
more stringent standards, and the need of the U.S to conserve energy 
does not, at present, appear to counsel toward higher stringency. 
Standards that increase at 1.5 percent per year represent a reasonable 
balance of additional technology and required per-vehicle costs, 
consumer demand for fuel economy, fuel savings and emissions avoided 
given the foreseeable state of the global oil market and the minimal 
effect on climate between finalizing 1.5 percent standards versus more 
stringent standards. The final standards will also result in year-over-
year improvements in fleetwide fuel economy, resulting in energy 
conservation that helps address environmental concerns, including 
criteria pollutant, air toxic pollutant, and carbon emissions. All 
things considered, NHTSA determines that an increase of 1.5 percent per 
year is maximum feasible for both passenger cars and light trucks for 
MYs 2021-2026.
---------------------------------------------------------------------------

    \3007\ For example, EIA currently expects U.S. retail gasoline 
prices to average $2.14/gallon in 2020, compared to $2.69/gallon in 
2019 (see https://www.eia.gov/outlooks/steo/archives/mar20.pdf), and 
$3.68/gallon in 2012 (see https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=EMM_EPM0_PTE_NUS_DPG&f=A). While gasoline 
prices may foreseeably rise over the rulemaking time frame, it is 
also very foreseeable that they will not rise to the $4-5/gallon 
that many American saw over the 2008-2009 time frame, that caused 
the largest shift seen toward smaller and higher-fuel-economy 
vehicles. See, e.g., Figure VIII-2 above.
---------------------------------------------------------------------------

Compliance and Enforcement

A. Introduction

1. Overview
    The CAFE and CO2 emissions standards are both fleet-
average standards, and for both programs, determining compliance begins 
by testing vehicles on dynamometers in a laboratory over pre-defined 
test cycles under controlled conditions.\3008\ A machine is connected 
to the vehicle's tailpipe while it performs the test cycle, which 
collects and analyzes the resulting exhaust gases; a vehicle that has 
no tailpipe emissions has its performance measured differently, as 
discussed below. CO2 quantities, as one of the exhaust 
gases, can be evaluated for vehicles that produce CO2 
emissions directly. Fuel economy is determined from the amount of 
CO2 emissions, because the two are directly mathematically 
related.\3009\ Manufacturers generally perform their own testing, and 
EPA confirms and validates those results by testing a sample of 
vehicles at the National Vehicle and Fuel Emissions Laboratory (NVFEL) 
in Ann Arbor, Michigan. The results of this testing form the basis for 
determining a manufacturer's compliance in a given model year, through 
the following steps:
---------------------------------------------------------------------------

    \3008\ For readers unfamiliar with this process, it is similar 
to running a car on a treadmill following a program--or more 
specifically, two programs. 49 U.S.C. 32904(c) states that, in 
testing for fuel economy, EPA must ``use the same procedures for 
passenger automobiles [that EPA] used for model year 1975 (weighted 
55 percent urban cycle and 45 percent highway cycle), or procedures 
that give comparable results.'' Thus, the ``programs'' are the 
``urban cycle,'' or Federal Test Procedure (abbreviated as ``FTP'') 
and the ``highway cycle,'' or Highway Fuel Economy Test (abbreviated 
as ``HFET''), and they have not changed substantively since 1975. 
Each cycle is a designated speed trace (of vehicle speed versus 
time) that vehicles must follow during testing--the FTP is meant 
roughly to simulate stop and go city driving, and the HFET is meant 
roughly to simulate steady flowing highway driving at about 50 mph. 
The 2-cycle dynamometer test results differ somewhat from what 
consumers will experience in the real world driving environment 
because of the lack of high speeds, rapid accelerations, and hot and 
cold temperatures evaluations with the A/C operation. These added 
conditions are more so reflected in the EPA 5-cycle test results 
listed on each vehicle's fuel economy label and on the 
fueleconomy.gov website.
    \3009\ Technically, for the CAFE program, carbon-based tailpipe 
emissions (including CO2, CH4, and CO) are 
measured, and fuel economy is calculated using a carbon balance 
equation. EPA uses carbon-based emissions (CO2, 
CH4, and CO, the same as for CAFE) to calculate the 
tailpipe CO2 equivalent for the tailpipe portion of its 
standards.
---------------------------------------------------------------------------

     Each vehicle model's performance on the test cycles is 
calculated;
     The number of vehicles of that model that were produced is 
divided by the performance;
     That number, in turn is summed for all the manufacturer's 
model types;
     The manufacturer's total product volume is then divided by 
the summed value of all the model types; and
     That number represents the manufacturer's fleet harmonic 
average performance.
    That performance is then compared to the manufacturer's unique 
compliance obligation (standard). This compliance obligation is 
calculated using the same approach that is used to determine 
performance, except that the fuel economy or CO2 target 
value (based on the footprint of each vehicle model) is used instead of 
the model's measured performance value. The fuel economy or 
CO2 target values for each of the vehicle models in the 
manufacturer's fleet and production volumes are used to derive the 
manufacturer's fleet harmonic average standard. Using fuel economy 
targets to illustrate the concept, the following figure shows two 
vehicle models produced in a model year for which passenger cars are 
subject to a fuel economy target function that extends from about 30 
mpg for the largest cars to about 41 mpg for the smallest cars:

[[Page 25188]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.745

    If these are the only two vehicle models the manufacturer produces, 
the manufacturer's required CAFE obligation is determined by 
calculating the production-weighted harmonic average of the fuel 
economy target values applicable at the hatchback and sedan footprints 
(from the curve, about 41 mpg for the hatchback and about 33 mpg for 
the sedan). The manufacturer's achieved CAFE level is determined by 
calculating the production-weighted harmonic average of the hatchback 
and sedan fuel economy levels (in this example the values shown in the 
boxes in Figure IX-1, 48 mpg for the hatchback and 25 mpg for the 
sedan). Depending on the relative mix of hatchbacks and sedans the 
manufacturer produces, the manufacturer's fleet may meet the standard, 
or perform better than the standard (if required CAFE is less than 
achieved CAFE) and thereby earn credits or perform worse than the 
standard (if required CAFE is greater than achieved CAFE) and thereby 
have a shortfall that may be made up, in whole or in part, using CAFE 
credits, discussed below, or be subject to civil penalties. Although 
the arithmetic is different for CO2 standards (which do not 
involve harmonic averaging), the underlying concept is the same.
    There are thus two parts to the foundation of compliance with CAFE 
and CO2 emissions standards: First, how well any given 
vehicle model performs relative to its target, and second, how many of 
each vehicle model a manufacturer produces. While no given model need 
precisely meet its target (and virtually no model exactly meets its 
target in the real-world), if a manufacturer finds itself producing 
large numbers of vehicles that fall well short of their targets, it 
will have to find a way of offsetting that shortfall, either by 
increasing production of vehicles that exceed their targets, or by 
taking advantage of compliance flexibilities and incentives, or the 
manufacturer will be subject to civil penalties. Given that 
manufacturers typically need to produce for sale vehicles that 
consumers want to buy, and not all consumers value fuel economy, their 
options for pursuing the former approach can often be limited.
    The CAFE and CO2 programs both offer a number of 
compliance flexibilities and incentives, discussed in more detail 
below. For example, starting in model year 2017, manufactures have 
flexibility to account for efficiency improvements in air conditioning 
(A/C) systems and/or for the application fuel economy improving 
technologies that increase fuel economy in the real-world, but that 
are, in whole or in part, not accounted for (e.g., stop-start 
technology, or high efficiency alternators) using the 1975-based 2-
cycle compliance dynamometer test procedures.\3010\ These fuel economy 
improvements are added to the 2-cycle performance results and are 
included in the calculation of a manufacturer's fuel economy in 
determining compliance relative to standards. In addition, for MYs 
2017--2021, there are also two levels of compliance incentives for 
full-size pickup trucks with mild-HEV or strong-HEV technology or that 
overperform standards by 15 percent or more, or by 20 percent or 
more.\3011\ This final rule removes this incentive starting in MY 2022, 
as discussed in more detail below. These fuel economy improvements are 
also included, for those model years and as earned, in the

[[Page 25189]]

calculation of a manufacturer's fuel economy.\3012\
---------------------------------------------------------------------------

    \3010\ EPA regulations provided an equivalent program beginning 
in MY 2012.
    \3011\ Manufacturers also must apply the technology to a minimum 
percentage of their full-size pickup truck production.
    \3012\ NHTSA characterizes any programmatic benefit 
manufacturers can use to comply with CAFE standards that fully 
accounts for fuel use as a ``flexibility'' (e.g., credit trading) 
and any benefit that counts less than the full fuel use as an 
``incentive'' (e.g., adjustment of alternative fuel vehicle fuel 
economy). NHTSA flexibilities and incentives are discussed further 
in Section IX.D.
---------------------------------------------------------------------------

    Some flexibilities and incentives are expressly provided for by 
statute, and some have been implemented by the agencies through 
regulations, consistent with the statutory scheme. Compliance 
flexibilities and incentives for the CAFE and CO2 programs 
have a great deal of theoretical attractiveness: If designed properly, 
they can help to reduce overall regulatory costs, while maintaining or 
improving programmatic benefits. If designed poorly, they may create 
significant potential for market distortion (for instance, when 
manufacturers--in response to an incentive to deploy a particular type 
of technology--produce vehicles for which there is no natural market, 
such vehicles must be discounted in order to sell).\3013\ 
Manufacturers' use of compliance flexibilities and incentives requires 
proper governmental and industry collaboration for manufacturers to 
achieve the most effective pathways to compliance.\3014\ Overly-
complicated flexibility and incentive programs can result in greater 
expenditure of both private sector and government resources to track, 
account for, and manage. Moreover, flexibilities or incentives that 
tend to favor specific technologies could distort the market. By these 
means, compliance flexibilities or incentives could create an 
environment in which entities are encouraged to invest in such favored 
technologies and, unless those technologies are independently supported 
by market forces, encourage rent seeking in order to protect, preserve, 
and enhance profits of companies that seek to take advantage of the 
distortions created by government mandate. Further, to the extent that 
there is a market demand for vehicles with lower CO2 
emissions and higher fuel economy, compliance flexibilities and 
incentives may cause some manufacturers to fall behind the industry's 
pace if they become overly reliant on them rather than simply improving 
the efficiency of their vehicles to meet that market demand.
---------------------------------------------------------------------------

    \3013\ While many manufacturers publicly discuss their 
commitment to certain technologies that reduce CO2 
emissions, consumer interest in them thus far remains low, despite 
often-large financial incentives from both manufacturers and the 
Federal and State governments in the form of tax credits (i.e., 
natural gas or fuel-cell vehicles). It is questionable whether 
continuing to provide significant compliance incentives for 
technologies that consumers appear not to want is an efficient means 
to achieve either compliance or national goals (see, e.g., Congress' 
phase-out of the AMFA dual-fueled vehicle incentive in EISA, 49 
U.S.C. 32906).
    \3014\ For these reasons, in this final rule, NHTSA is asking 
manufacturers to provide more detailed information on the new 
incentives allowed for A/C and off-cycle technologies and on credit 
trades for better collaboration in understanding the economic impact 
of these flexibilities and incentives and for the government to 
provide better oversight of the CAFE program.
---------------------------------------------------------------------------

    If standards are maximum feasible levels, as required by statute, 
then the need for extensive compliance flexibilities and incentives 
should be low. The agencies sought comments in the NPRM on whether and 
how each agency's existing flexibilities and incentives might be 
amended, revised, or deleted to avoid the inefficiencies and market 
distortions discussed above. Specifically, comments were sought on the 
appropriate level of compliance flexibility, including credit trading, 
in a program that is correctly designed to be maximum feasible, in 
accordance with the statute. Comments were also sought on whether to 
allow all incentive-based adjustments, except those that are mandated 
by statute, to expire, in addition to other possible simplifications to 
reduce market distortion, improve program transparency and 
accountability, and improve overall performance of the compliance 
programs. The agencies considered comments on those issues in preparing 
the final rule. A summary of all the flexibilities for the CAFE and 
CO2 programs finalized as a part of this final rule is 
provided in Table IX-1 though Table IX-4.
BILLING CODE 4910-59-P

[[Page 25190]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.746


[[Page 25191]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.747


[[Page 25192]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.748


[[Page 25193]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.749


[[Page 25194]]


BILLING CODE 4910-59-C
2. Light-Duty CAFE Compliance Data for MYs 2011-2019
    To understand manufacturers' potential approaches to using 
compliance flexibilities and incentives, CAFE compliance data for MYs 
2011 through 2019 is discussed in this section. NHTSA believes that 
providing these data is important because it gives the public a better 
understanding of current compliance trends and the potential impacts 
that increasing CAFE standards have had on those model years and future 
model years addressed by this rulemaking.
    NHTSA uses data from CAFE reports submitted by manufacturers to EPA 
or directly to NHTSA to evaluate compliance with the CAFE program. The 
data for MYs 2011 through 2017 include manufacturers' final compliance 
data that have been verified by EPA.\3015\ The data for MYs 2018 and 
2019 include the most recent projections from manufacturers' mid-model 
year and final-model year reports submitted to EPA and NHTSA, as 
required by 49 CFR part 537 and 40 CFR 600.512-12.\3016\ Because the 
projections do not reflect final vehicle production levels, the EPA 
verified final CAFE values may be slightly different than the 
manufacturers' projections. MY 2011 was selected as the start of the 
data because it represents the first compliance model year for which 
manufacturers were permitted to trade and transfer credits.\3017\ MY 
2019 is also important because it shows the projected performance of 
the fleet two years after manufacturers were allowed to use new 
flexibilities and incentives starting in MY 2017 to address increasing 
CAFE standards.
---------------------------------------------------------------------------

    \3015\ The data contain the latest information available from 
manufacturers except certain low volume manufacturers complying with 
standards under 49 CFR part 525.
    \3016\ MY 2018 data come from information received in 
manufacturers' final reports submitted to EPA according to 40 CFR 
600.512-12 and MY 2019 data come from information received in 
manufacturers' mid-model year CAFE reports submitted to NHTSA 
according to 49 CFR part 537.
    \3017\ 49 CFR 535.6(c).
---------------------------------------------------------------------------

    Figure IX-2 through Figure IX-5 provide a graphical overview of 
fuel economy performance and standards. Fuel economy performance 
includes three parts: (1) Measured performance, on the 2-cycle 
dynamometer test; (2) performance increases for alternative fueled 
vehicles, under the Alternative Motor Fuels Act of 1988 (AMFA); and (3) 
performance adjustments for improved A/C systems and off-cycle 
technologies.3018 3019 3020 These Figures do not account for 
credits earned or expected to be earned from overcompliance in prior or 
future model years that were used or are available for complying with 
CAFE standards. Graphs are included for the total fuel economy 
performance (the combination of all passenger cars and light trucks 
produced for sale during the model year) as a single fleet, and for 
each of the three CAFE compliance fleets: Domestic passenger car, 
import passenger car, and light truck fleets.
---------------------------------------------------------------------------

    \3018\ In the Figures, the label ``CAFE with Capped AMFA'' 
represents the maximum increase each year in the average fuel 
economy set to the limitation ``cap'' for manufacturers attributable 
to dual-fueled automobiles as prescribed in 49 U.S.C. 32906. The 
labels ``A/C'' and ``off-cycle'' represents the increase in the 
average fuel economy adjusted for A/C and off-cycle fuel consumption 
improvement values as prescribed by 40 CFR 600.510-12.
    \3019\ The Alternative Motor Fuels Act (AMFA) allows 
manufacturers to increase their fleet fuel economy performance 
values by producing dual-fueled vehicles. Incentives are available 
for building advanced technology vehicles such as hybrids and 
electric vehicles, compressed natural gas vehicles and for building 
vehicles able to run on dual-fuels such as E85 and gasoline. For MYs 
1993 through 2014, the maximum possible increase in CAFE performance 
is ``capped'' for a manufacturer attributable to dual-fueled 
vehicles at 1.2 miles per gallon for each model year and thereafter 
decreases by 0.2 miles per gallon each model year through MY 2019. 
49 U.S.C. 32906.
    \3020\ Consistent with applicable law, NHTSA established 
provisions starting in MY 2017 allowing manufacturers to increase 
fuel economy performance-based on fuel consumption benefits gained 
by technologies not accounted for during normal 2-cycle EPA 
compliance testing (called ``off-cycle technologies'' for 
technologies such as stop-start systems) as well as for A/C systems 
with improved efficiencies and for hybrid or electric full-size 
pickup trucks.
---------------------------------------------------------------------------

BILLING CODE 4910-59-P

[[Page 25195]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.750


[[Page 25196]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.751

    As shown in Figure IX-2, manufacturers' fuel economy performance 
for the total fleet was better than the overall CAFE standard through 
MY 2015. On average, the total fleet exceeded the overall CAFE 
standards by approximately 0.9 mpg for MYs 2011 to 2015. Comparatively, 
as shown in Figure IX-3 through Figure IX-5, for these same model 
years, domestic and import passenger cars exceeded standards on average 
by 2.1 mpg and 2.3 mpg, respectively. By contrast, for light trucks, 
manufacturers on average fell below standards by 0.3 mpg.
    For MYs 2016 through 2019, as shown in the Figures, NHTSA has 
determined that the combined CAFE performance, including all 
flexibilities and incentives, of the total fleet has or is expected to 
be worse than the applicable CAFE standards, and increasingly so. The 
domestic passenger car fleet is the only compliance category expected 
to continue to be better than CAFE standards through MY 2018. But even 
the overall domestic passenger car fleet

[[Page 25197]]

is expected to be worse than standards in MY 2019. The data show MYs 
2016 through 2019 standards involve significant compliance challenges 
for many vehicle manufacturers. This is evident in the fact that the 
total fleet falls below the applicable CAFE standards on average by 0.6 
mpg for these model years. Compliance challenges become even more 
substantial when observing individual compliance fleets. The largest 
individual performance shortfalls (i.e. the difference between CAFE 
performance values and standards) exist for import passenger car 
manufacturers, with an expected shortfall of 2.5 mpg in MY 2019, 
followed by light truck manufacturers, with a shortfall of 1.4 mpg in 
MY 2016.
    Table IX-5 provides the numerical final CAFE performance values and 
standards for MYs 2004 to 2017. Notably, there was an increase in total 
fleet fuel economy of only 0.1 mpg for MY 2014, and no increase for MY 
2016. In MY 2016, the total fleet's performance fell below the CAFE 
standard by 0.5 mpg. An increase in the total fleet's CAFE performance 
for MY 2017 was largely due to manufacturers gaining benefits from A/C 
and off-cycle technologies. For MY 2017, the total fleet's CAFE 
performance without A/C and off-cycle allowances increased by 0.1 mpg 
compared to MY 2016. However, even combined with new flexibilities, the 
total fleet's CAFE performance, for MY 2017, still falls below the CAFE 
standard by 0.4 mpg.
[GRAPHIC] [TIFF OMITTED] TR30AP20.752

BILLING CODE 4910-59-C
    Figure IX-6 provides a historical overview of the industry's use of 
CAFE compliance flexibilities for addressing performance 
shortfalls.\3021\ MY 2016 is the latest model year for which CAFE 
compliance determinations are complete, and credit application and 
civil penalty payment determinations made by the manufacturer. 
Historically, manufacturers have generally resolved credit shortfalls 
first by carrying forward any earned credits and then applying traded 
credits. In MYs 2014 and 2015, the amount of credit shortfalls is 
almost the same as the amount of carry-forward and traded credits. 
Manufacturers occasionally carryback credits or opt to transfer earned 
credits between their fleets to resolve performance shortfalls. Trading 
credits from another manufacturer and transferring them across fleets 
occurs far more frequently. Also, credit trading has generally taken 
the place of civil penalty payments for resolving performance 
shortfalls. Only a handful of manufacturers have made civil penalty 
payments since the implementation of the credit trading program.\3022\ 
NHTSA expects there may be sufficient credits in manufacturers' credit 
accounts to resolve all import passenger car and light truck 
performance shortfalls expected through MY 2019. By statute, 
manufacturers cannot use traded or transferred credits to address 
performance shortfalls for failing to meet the minimum domestic 
passenger car standards.\3023\ One domestic passenger car manufacturer 
paid civil penalties for failing to comply with the minimum domestic 
passenger car standards for MYs 2016 and 2017.\3024\ Additional 
manufacturers are

[[Page 25198]]

expected to pay civil penalty payments for failing to comply with the 
minimum domestic passenger cars standards for MYs 2018 through 2019.
---------------------------------------------------------------------------

    \3021\ The Figure includes all credits manufacturers have used 
in credit transactions to date. Credits contained in carryback plans 
yet to be executed or in pending enforcement actions are not 
included in the Figure.
    \3022\ Six manufacturers have paid CAFE civil penalties since 
credit trading began in 2011. Fiat Chrysler paid the largest civil 
penalty total over the period, followed by Jaguar Land Rover and 
then Volvo. See Summary of CAFE Civil Penalties Collected, CAFE 
Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Fines_LIVE.html.
    \3023\ Congress prescribed minimum domestic passenger car 
standards for domestic passenger car manufacturers and unique 
compliance requirements for these standards in 49 U.S.C. 32902(b)(4) 
and 32903(f)(2).
    \3024\ Fiat Chrysler paid $77,268,702.50 in civil penalties for 
MY 2016 and $79,376,643.50 for MY 2017 for failing to comply with 
the minimum domestic passenger car standards for those MYs.
[GRAPHIC] [TIFF OMITTED] TR30AP20.753

    The compliance data show that the rate at which industry has been 
increasing fuel economy, as shown by the actual fuel economy of the 
overall fleet, has not kept pace with the year-over-year increases in 
the stringency of the standards since MY 2010. The margin of CAFE 
overcompliance diminished steadily through MY 2015. In MY 2016, the 
fuel economy of the fleet was worse than standards, and the margin of 
the shortfall has or is projected to become worse through MY 2019. 
Manufacturers have increasingly used CAFE compliance flexibilities and 
paid more in civil penalties to address the growing CAFE shortfalls. 
The data show use of these flexibilities is likely to increase at least 
through 2019.
3. Shift in Sales Production From Passenger Cars to Light Trucks
    The notable trend in the stagnant growth in the automotive 
industry's CAFE performance is likely related to an increase in the 
purchase of light trucks beginning with MY 2013. Light trucks had a 
sharp spike in sales, increasing by a total of 5 percent from MYs 2013 
to 2014. In MY 2014, light trucks comprised approximately 41 percent of 
the total sales production volume of automobiles and has continued to 
grow ever since. In comparison, for model year 2014, domestic passenger 
cars represented 36 percent of the total fleet and import passenger 
cars represented 23 percent. Both domestic and import passenger car 
sales have continued to fall every year since MY 2013. Figure IX-7 
shows the sales production volumes of light trucks and domestic and 
import passenger cars for MYs 2004 to 2017. The proportion of light 
trucks in the fleet, being driven by consumer demand and lower fuel 
prices, raises some concern for the ability of that fleet to comply 
with future CAFE standards. Historically, light truck fleets have 
fallen below their associated CAFE standards and have had larger 
performance shortages than either import and domestic passenger car 
fleets. This trend is expected to continue, even with allowance for A/C 
and off-cycle flexibilities. For MY 2019, NHTSA expects even greater 
CAFE performance shortages in the light truck and import passenger car 
fleets than in prior model years, based upon manufacturer's MMY 
reports. The combined effect of these fuel economy shortages will 
require manufacturers to rely heavily on compliance flexibilities or 
pay civil penalties.
    Another important factor in automobile sales production impacting 
CAFE performance values involves increasing trends in the volume of 
small SUVs and pickup trucks. These vehicles as a percentage of total 
fleet increased from approximately 52 percent in MY 2012 to 63 percent 
in MY 2017. As shown in Figure IX-8, small SUVs, with 4WD and 2WD 
drivetrains, in particular have surpassed the sales production volumes 
of all other vehicle classes over these the given model years. The 
number of small and standard SUVs sold in the U.S. for MY 2017 nearly 
doubled compared to sales in the U.S. for MY 2012. During that same 
period, passenger car sales production as a total of vehicle sales 
production decreased by approximately 11 percent. The combination of 
low gas prices and the increased utility that SUVs provide may explain 
the shift in sales production. Nonetheless, if the sales of these small 
SUVs and pickup trucks continue to increase, NHTSA expects there will 
be continued stagnation in the CAFE performance of the overall fleet.
BILLING CODE 4910-59-P

[[Page 25199]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.754


[[Page 25200]]


[GRAPHIC] [TIFF OMITTED] TR30AP20.755

BILLING CODE 4910-59-C
4. Vehicle Classification
    Before manufactures can comply with CAFE and CO2 
standards, they must first determine how a vehicle is classified in 
accordance with 49 CFR part 523, ``Vehicle Classification.'' In EPCA, 
Congress designated some vehicles as passenger automobiles and some as 
non-passenger automobiles. Vehicle classification, for purposes of the 
light-duty CAFE and CO2 programs, refers to whether a 
vehicle is classified as a passenger automobile (car) or a non-
passenger automobile (light truck).3025 3026 As discussed 
previously, passenger cars and light trucks are subject to different 
fuel economy and CO2 standards, and light trucks have less 
stringent standards to accommodate their utility usage.
---------------------------------------------------------------------------

    \3025\ See 40 CFR 86.1803-01. For the MYs 2012-2016 standards, 
the MYs 2017-2025 standards, and this rule, EPA uses NHTSA's 
regulatory definitions for determining which vehicles would be 
subject to which CO2 standards.
    \3026\ EPCA uses the terms ``passenger automobile'' and ``non-
passenger automobile;'' NHTSA's regulation on vehicle 
classification, 49 CFR part 523, further clarifies the EPCA 
definitions and introduces the term ``light truck'' as a plainer 
language alternative for ``non-passenger automobile.''
---------------------------------------------------------------------------

    Under EPCA and NHTSA's current regulations, vehicles are classified 
as light trucks either on the basis of off-highway capability or on the 
basis of having truck-like (utility) 
characteristics.3027 3028 3029 Determining whether a vehicle 
is capable of ``off-highway operation'' is a two-part determination: 
First, does the vehicle either have 4-wheel drive or a gross vehicle 
weight rating (GVWR) over 6,000 pounds, and second, does the vehicle 
(that has either 4-wheel drive or over 6,000 pounds GVWR) also have ``a 
significant feature . . . designed for off-highway operation.'' \3030\ 
NHTSA's current regulations specify that this ``significant feature'' 
requires the vehicle to meet at least four out of five ground clearance 
dimensions.\3031\ Further, to be classified as a light truck on the 
basis of having truck-like characteristics instead, NHTSA regulations 
also require the vehicle to perform at least one of the following

[[Page 25201]]

functions: Carry more than 10 persons, provide temporary living 
quarters, have an open bed (i.e., a pickup truck), provide more cargo-
carrying volume than passenger-carrying volume, or permit expanded 
cargo volume capacity by the removal or stowing of rear seats.\3032\
---------------------------------------------------------------------------

    \3027\ 49 U.S.C. 32901(a)(18); 49 CFR part 523.
    \3028\ 49 CFR 523.5(b).
    \3029\ 49 CFR 523.5(a).
    \3030\ 49 U.S.C. 32901(a)(18).
    \3031\ The ground clearance dimensions are: (i) Approach angle 
of not less than 28 degrees; (ii) breakover angle of not less than 
14 degrees; (iii) departure angle of not less than 20 degrees; (iv) 
running clearance of not less than 20 centimeters; and/or (v) front 
and rear axle clearances of not less than 18 centimeters each.
    \3032\ By statute, vehicles that NHTSA, on behalf of the 
Secretary of DOT, ``decides by regulation [are] manufactured 
primarily for transporting not more than 10 individuals'' are 
passenger automobiles. 49 U.S.C. 32901(a)(18).
---------------------------------------------------------------------------

    Over time, NHTSA has revised its light truck vehicle classification 
regulations and issued legal interpretations to address changes in 
vehicle designs. Based upon agency observations of current vehicle 
design trends, compliance testing and evaluation, and discussions with 
stakeholders, NHTSA has become aware of certain additional design 
changes that further complicate light truck classification 
determinations for the CAFE and CO2 programs. NHTSA 
discussed several classification issues in the NPRM and sought comments 
on potential resolutions. Only a few comments were received, primarily 
from vehicle manufacturers, and they were aimed generally at requesting 
flexibility in how NHTSA applies the existing classification criteria. 
A summary of the comments received and NHTSA's responses for the final 
rule are explained in the following sections.
a) Classification Based on ``Truck-Like Characteristics''
    One of the ``truck-like characteristics'' that allows manufacturers 
to classify vehicles as light trucks is having at least three rows of 
seats as standard equipment, as long as the design also ``permit[s] 
expanded use of the automobile for cargo-carrying purposes or other 
non-passenger-carrying purposes through the removal or stowing of 
foldable or pivoting seats so as to create a flat, leveled cargo 
surface extending from the forwardmost point of installation of those 
seats to the rear of the automobile's interior.'' \3033\ Typically, 
most minivans qualify under the provision by expanding the cargo area 
through removable or stowable seats, and a small percentage of sports 
utility vehicles qualify through folding seats that use the seat backs 
to form a secondary ``raised'' cargo floor.\3034\ NHTSA identified two 
issues with this criterion that various manufacturers appear to be 
approaching differently. Both relate to how expanded cargo area is 
provided when seats are removed or stowed in the vehicle.
---------------------------------------------------------------------------

    \3033\ 49 CFR 523.5(a)(5)(ii).
    \3034\ All minivans and a small percentage of sports utility 
vehicles that qualify as light trucks do so by meeting the 
characteristic for third row seats. As more advanced seating designs 
are introduced in minivans, manufacturers that wish to retain this 
status will need to avoid losing the expanded cargo characteristics 
that are the basis for the allowing minivans to be qualified as 
light trucks.
---------------------------------------------------------------------------

    The first issue is how to identify the ``forwardmost point of 
installation'' and how the location impacts the available cargo floor 
area and volume behind the seats. Seating configurations have evolved 
considerably over the last twenty years, as minivan seats are now very 
complex in design, providing far more ergonomic functionality. For 
example, the market demand for increased rear seat leg room has 
resulted in adjustable second row seats mounted to sliding tracks. 
Earlier seating designs had fixed attachment points on the vehicle 
floor, and it was easy to identify the ``forwardmost point of 
installation'' because it was readily observable and did not change. 
When seats move forward and backward on sliding tracks, however, the 
``forwardmost point of installation'' is less readily identifiable. To 
avoid this complication, most manufacturers maintain light truck 
qualification by using adjustable seats that can be removed from the 
vehicle and having a flat floor rearward of the front seats.\3035\ For 
others, the qualification is not as apparent because new adjustable 
seats have been introduced that remain within vehicle to accommodate 
side airbags. Manufacturers designate various positions for the 
forwardmost point of installation in vehicles where the seat in the 
sliding track can be moved far enough forward to allow the entire seat 
to compress against the back of the front seat where it can be stowed 
beyond the forwardmost point of installation, while the seat cushion 
bottom folds towards the seatback. In some cases, manufacturers 
designate the forwardmost point of installation at a location in the 
sliding track where the seat is positioned at its rearmost position in 
the track. In others, the initial point of installation is designated 
at a location in the sliding track accommodating the seating position 
of a 75-percentile male test dummy. The amount of the flat floor 
surface area and cargo volume behind the seats can vary depending on 
which approach a manufacturer adopts.
---------------------------------------------------------------------------

    \3035\ NHTSA notes that to qualify as a light truck, a vehicle 
still requires a flat floor from the forwardmost point of 
installation of removable second row seats to the rear of the 
vehicle.
---------------------------------------------------------------------------

    NHTSA sought public comments in the NPRM to explore potential 
options for establishing the forwardmost point of installation for 
adjustable second row seats and to evaluate whether an additional 
classification criteria could be required, specifying a minimum amount 
of cargo volume behind the seats. Comments were received from the Auto 
Alliance and Fiat Chrysler.\3036\ Both the Auto Alliance and Fiat 
Chrysler commented that some flexibility is needed in determining the 
forwardmost point of installation that allows manufacturers to set the 
location of the seat attachment point to the sliding track in any 
manufacturer-designated position that allows for customer-ergonomics 
and safety, while still meeting the spirit of the expanded cargo-
carrying requirement.\3037\ The Auto Alliance further commented that 
the forwardmost attachment point of the seat structure to the floor is 
still a viable method of measurement, even when there is a sliding 
track between the floor attachment point and the seat.\3038\
---------------------------------------------------------------------------

    \3036\ The National Automobile Dealers Association commented 
generally that it does not support any substantial modifications to 
the existing passenger car and light truck fleet definitions.
    \3037\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3038\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    NHTSA did not propose any vehicle reclassifications and is not 
adopting a regulatory change at this time. Based on its review of the 
comments, NHTSA agrees that flexibility is warranted to accommodate 
safety and customer demand but clarifies that the regulation requires 
seats that are not removed to be stowed--that is, moved so as to form a 
cargo area behind the seats. Manufacturers can freely designate the 
seating location in the sliding track to establish the forwardmost 
point of installation. At that seat location, the forwardmost point of 
installation is the forwardmost attachment point of the seat structure 
(including any carriage structures) to the floor in the sliding track. 
Vehicles will be considered to meet the characteristic provided the 
rear of the seats can be moved forward beyond that point and the seats 
articulate to an unusable stowed position either in the floor of the 
vehicle or at the front perimeter of expanded cargo area.\3039\
---------------------------------------------------------------------------

    \3039\ The front perimeter of the cargo area is the plane formed 
behind the front seats and extending from one side of the vehicle to 
the other.
---------------------------------------------------------------------------

    The second issue concerns the ``flatness'' and ``levelness'' of 
folded rear seats that use the seat backs to form a raised cargo 
surface and whether the seats must form a continuous flat, leveled 
surface. Many SUVs have three rows of designated seating positions, 
where the second row has ``captain's seats'' (i.e., two independent 
bucket

[[Page 25202]]

seats), rather than the traditional bench-style seating more common 
when the provision was added to NHTSA's regulation. When captain's 
seats are folded down, the seatback can form a flat surface for 
expanded cargo-carrying purposes, but the surface of the seatbacks may 
be angled (i.e., at some angle slightly greater than 0[deg]), or may be 
at a different level with the rest of the cargo area (i.e., horizontal 
surface of folded seats is 0[deg] at a different height from horizontal 
surface of cargo area behind the seats). Captain's seats, when folded 
flat, may also leave significant gaps around and between the seats. 
Some manufacturers have opted to use plastic panels to level the 
surface and to covers the gaps between seats, while others have left 
the space open and the surface angled or at different levels. NHTSA 
sought comments in the NPRM on the following questions related to the 
requirement for a flat, leveled cargo surface:
     Does the cargo surface need to be flat and level in 
exactly the same plane, or does it fulfill the intent of the criterion 
and provide appropriate cargo-carrying functionality for the cargo 
surface to be other than flat and level in the same plane?
     Does the cargo surface need to be flat and level across 
the entire surface, or are (potentially large) gaps in that surface 
consistent with the intent of the criterion and providing appropriate 
cargo-carrying functionality? Should panels to fill gaps be required?
     Certain third row seats are located on top the rear axle 
causing them to sit higher and closer to the vehicle roof. When these 
seats fold flat the available cargo-carrying volume is reduced. Is 
cargo-carrying functionality better ensured by setting a minimum amount 
of useable cargo-carrying volume in a vehicle when seats fold flat?
    The Auto Alliance, Fiat Chrysler, Hyundai, Kia, and one individual, 
Walter Kreucher, commented on these seating issues. The Auto Alliance, 
Fiat Chrysler, and Walter Kreucher believed that the criteria for a 
``flat, leveled cargo surface'' should not be interpreted to mean that 
a cargo surface must be flat and level in exactly the same plane.\3040\ 
The comments noted that a surface that is not exactly flat and level in 
the same plane can still provide substantial cargo-carrying capacity, 
while allowing manufacturers to provide ergonomically comfortable seats 
that meet safety requirements.\3041\ The comments stated that NHTSA 
should not establish a minimum amount of cargo surface area for seats 
that remain within the vehicle.\3042\ Instead, they preferred that 
manufacturers should be allowed to determine the methodology for 
providing appropriate cargo-carrying functionality without NHTSA 
stipulating additional requirements for flat and level surfaces or gaps 
and gap-filling panels.\3043\
---------------------------------------------------------------------------

    \3040\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Walter 
Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
    \3041\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
0067-11943.
    \3042\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
    \3043\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
0067-11943.
---------------------------------------------------------------------------

    The Auto Alliance and Fiat Chrysler argued that area or volume 
requirements are not needed, as those attributes speak to overall 
vehicle size and shape, which should remain a consumer choice.\3044\ 
The requirements for expanded cargo- or other non-passenger-carrying 
purposes are fully met in the existing regulation, which requires a 
flat, leveled cargo surface with two rows of seats that are folded or 
stowed. Fiat Chrysler also commented that potential new requirements 
would likely be interpreted and executed differently across 
manufacturers and could narrow the choice of engineering solutions and 
negatively affect other important vehicle attributes.\3045\
---------------------------------------------------------------------------

    \3044\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3045\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    Hyundai and Kia commented that instead of requiring panels, NHTSA 
could limit the size of the gaps around and between folded seats.\3046\ 
In that case, manufacturers would have flexibility to use panels if 
they wish but could take other measures to narrow gaps. On the other 
hand, Walter Kreucher stated that NHTSA should allow gaps of any size 
and not require the use of panels to cover them.\3047\
---------------------------------------------------------------------------

    \3046\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411; 
Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
    \3047\ Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
---------------------------------------------------------------------------

    NHTSA is not adopting a regulatory change at this time. NHTSA 
agrees with commenters that it should not require a minimum amount of 
cargo surface area or volume for seats that remain within the vehicle, 
which could be difficult to meet for certain vehicle sizes and shapes 
that would otherwise be considered non-passenger vehicles. NHTSA agrees 
that the amount of cargo volume should be a consumer choice. Setting a 
minimum amount of cargo area or volume could have an adverse effect on 
potential new car buyers.
    NHTSA notes that there may also be safety considerations involved 
with the requirement to have a flat, leveled cargo surface area formed 
by seat backs. A flat, leveled cargo surface area could prevent objects 
from having a ramp-like surface to gain momentum in rolling backwards 
into the tailgate's interior surface, potentially causing stress or 
damage on the tailgate's latching mechanism. For these reasons, several 
standards exist in the industry for preventing objects from sliding, 
such as standards from the American Disability Act (ADA) that specify 
floor and ground design requirements for protecting wheelchair seated 
occupants. In addition, objects resting on the tailgate could become a 
hazard or source of injury for individuals opening the tailgate. At 
this time, NHTSA accepts the commenters' position that having a cargo 
surface area that is exactly flat and level in the same plane may not 
be necessary. Comments did not provide enough information for NHTSA to 
identify any changes to the existing requirements. Therefore, at this 
time, NHTSA will retain its existing provisions for the stowing of 
foldable or pivoting seats to create a flat, leveled cargo surface, but 
NHTSA may consider conducting research in the future regarding these 
issues. NHTSA has also determined that it should set not a limit on the 
size of the gaps between folded seats at this time, although it may 
consider adopting such limits in the future. NHTSA continues to 
encourage manufacturers to consider the safety implications of all 
aspects of their vehicle designs, including any angling of the seat 
back cargo surface and whether it is appropriate to offer panels as 
optional equipment for covering any large gap openings.
b) Issues That NHTSA Has Observed Regarding Classification Based on 
``Off-Road Capability''
(1) Measuring Vehicle Characteristics for Off-Highway Capability
    For a vehicle to qualify as off-highway capable, in addition to 
either having 4WD or a GVWR more than 6,000 pounds, the vehicle must 
have four out of five characteristics indicative of off-highway 
operation.\3048\ These characteristics are:
---------------------------------------------------------------------------

    \3048\ 49 CFR 523.5(b)(2).
---------------------------------------------------------------------------

     An approach angle of not less than 28 degrees
     A breakover angle of not less than 14 degrees
     A departure angle of not less than 20 degrees

[[Page 25203]]

     A running clearance of not less than 20 centimeters
     Front and rear axle clearances of not less than 18 
centimeters each
    NHTSA's regulations require manufacturers to measure these 
characteristics when a vehicle is at its curb weight, on a level 
surface, with the front wheels parallel to the automobile's 
longitudinal centerline, and the tires inflated to the manufacturer's 
recommended cold inflation pressure.\3049\ Given that the regulations 
describe the vehicle's physical position and characteristics at time of 
measurement, NHTSA previously assumed that manufacturers would use 
physical measurements of vehicles. In practice, NHTSA has instead 
received from manufacturers a mixture of angles and dimensions from 
design models (i.e., the vehicle as designed, not as actually produced) 
and/or physical vehicle measurements.\3050\ When appropriate, the 
agency will verify reported values by measuring production vehicles in 
the field. NHTSA currently requires that manufacturers use physical 
vehicle measurements as the basis for values reported to the agency for 
purposes of vehicle classification. NHTSA sought comment on whether 
regulatory changes are needed with respect to this issue.
---------------------------------------------------------------------------

    \3049\ Id.
    \3050\ NHTSA previously encountered a similar issue when 
manufacturers reported CAFE footprint information. In the October 
2012 final rule, NHTSA clarified manufacturers must submit footprint 
measurements based upon production values. 77 FR 63138 (October 15, 
2012).
---------------------------------------------------------------------------

(2) Approach, Breakover, and Departure Angles
    Approach angle, breakover angle, and departure angle are relevant 
to determining off-highway capability. Large approach and departure 
angles ensure the front and rear bumpers and valance panels have 
sufficient clearance for obstacle avoidance while driving off-road. The 
breakover angle ensures sufficient body clearance from rocks and other 
objects located between the front and rear wheels while traversing 
rough terrain. Both the approach and departure angles are derived from 
a line tangent to the front (or rear) tire static loaded radius arc 
extending from the ground near the center of the tire patch to the 
lowest contact point on the front or rear of the vehicle. The term 
``static loaded radius arc'' is based upon the definitions in SAE J1100 
and J1544. The term is defined as the distance from wheel axis of 
rotation to the supporting surface (ground) at a given load of the 
vehicle and stated inflation pressure of the tire (manufacturer's 
recommended cold inflation pressure).\3051\
---------------------------------------------------------------------------

    \3051\ 49 CFR 523.2.
---------------------------------------------------------------------------

    The static loaded radius arc is easy to measure, but the imaginary 
line tangent to the static loaded radius arc is difficult to ascertain 
in the field. The approach and departure angles are the angles between 
the line tangent to the static loaded radius arc and the level ground 
on which the test vehicle rests. Simpler measurements that provide good 
approximations for the approach and departure angles involve using 
either a line tangent to the outside diameter or perimeter of the tire 
or a line that originates at the geometric center of the tire contact 
patch and extends to the lowest contact point on the front or rear of 
the vehicle. The first method provides an angle slightly greater than, 
and the second method provides an angle slightly less than, the angle 
derived from the true static loaded radius arc. Both approaches can be 
used to measure angles in the field to verify data submitted by the 
manufacturers used to determine light truck classification decisions.
    NHTSA sought comment on what the effect would be if it replaced 
reference to the ``static loaded arc radius'' with a different term 
like ``outside perimeter of the tire'' or ``geometric center of the 
tire contact patch.'' The Auto Alliance and Fiat Chrysler offered 
comments. The Auto Alliance and Fiat Chrysler commented that only a 
measurement using the static loaded arc radius reasonably reflects the 
tire condition during off-road events that approach, breakover, and 
departure angles are quantifying. They also stated the static loaded 
arc radius best reflects the actual condition that exists versus the 
outside tire diameter.\3052\ Finally, the Auto Alliance commented the 
static loaded arc radius is easy to measure; therefore, the off-road 
criteria should remain tied to the static loaded arc radius.\3053\
---------------------------------------------------------------------------

    \3052\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3053\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    After reviewing the comments, NHTSA agrees that the static loaded 
arc radius is the most accurate way to account for the condition of the 
tire and the vehicle-to-ground interaction during off-road events. 
NHTSA has decided to accept the Auto Alliance's and Fiat Chrysler's 
views and will retain the existing definitions for off-road angles 
based upon the static loaded arc radius.
(3) Running Clearance
    NHTSA regulations define ``running clearance'' as ``the distance 
from the surface on which an automobile is standing to the lowest point 
on the automobile, excluding unsprung weight.'' \3054\ Unsprung weight 
includes the components (e.g., suspension, wheels, axles, and other 
components directly connected to the wheels and axles) that are 
connected and translate with the wheels. Sprung weight, on the other 
hand, includes all components fixed underneath the vehicle and 
translate with the vehicle body (e.g., mufflers and subframes). To 
clarify these requirements, NHTSA previously issued a letter of 
interpretation stating that certain parts of a vehicle--such as tire 
aero deflectors that are made of flexible plastic, bend without 
breaking, and return to their original position--would not count 
against the 20-centimeter running clearance requirement.\3055\ The 
agency explained that this does not mean a vehicle with less than 20-
centimeters running clearance could be elevated by an upward force that 
bends the deflectors and still be considered compliant with the running 
clearance criterion, as it would be inconsistent with the conditions 
listed in the introductory paragraph of 49 CFR 523.5(b)(2). Further, 
NHTSA explained that without a flexible component installed, the 
vehicle must meet the 20-centimeter running clearance along its entire 
underside. This 20-centimeter clearance is required for all sprung 
weight components.
---------------------------------------------------------------------------

    \3054\ Id.
    \3055\ See letter to Mark D. Edie, Ford Motor Company, July 30, 
2012, available at https://isearch.nhtsa.gov/files/11-000612%20M.Edie%20(Part%20523).htm.
---------------------------------------------------------------------------

    The agency is aware of vehicle designs that incorporate rigid 
(i.e., inflexible) air dams, valance panels, exhaust pipes, and other 
components, equipped as manufacturers' standard or optional equipment 
(e.g., running boards and towing hitches), that likely do not meet the 
20-centimeter running clearance requirement. Despite these rigid 
features, it appears manufacturers are not taking these components into 
consideration when making measurements. Additionally, NHTSA believes 
some manufacturers may provide dimensions for their base vehicles 
without considering optional or various trim level components that may 
reduce the vehicle's ground clearance. Consistent with our approach to 
other measurements, NHTSA believes that ground clearance, as well as 
all the other off-highway criteria for a light truck determination, 
should use the measurements from vehicles with all standard and 
optional equipment

[[Page 25204]]

installed, at the time of the first retail sale.\3056\ The agency 
reiterates that the characteristics listed in 49 CFR 523.5(b)(2) are 
characteristics indicative of off-highway capability. A fixed feature--
such as an air dam that does not flex and return to its original state 
or an exhaust that could detach--inherently interferes with the off-
highway capability of these vehicles. If manufacturers seek to classify 
these vehicles as light trucks under 49 CFR 523.5(b)(2) and the 
vehicles do not meet the four remaining characteristics to demonstrate 
off-highway capability, they must be classified as passenger cars.
---------------------------------------------------------------------------

    \3056\ See NHTSA's footprint test procedure for verifying CAFE 
standards uses vehicles equipped at time of first retail sale. See 
TP-537-01 located at https://www.nhtsa.gov/vehicle-manufacturers/test-procedures.
---------------------------------------------------------------------------

    In the NPRM, NHTSA sought public comments on how to consider 
components such as air dams, exhaust pipes, and other hanging component 
features--especially those that are inflexible--as relates to running 
clearance and whether the agency should consider amending its 
definition in Part 523 to account for these components. The Auto 
Alliance and three automobile manufacturers--Fiat Chrysler, Hyundai, 
and Kia--commented on the questions. The Auto Alliance and Fiat 
Chrysler commented that no change is needed for the 20-centimeter 
running clearance requirement for fixed features of the vehicle; all 
fixed components must have 20-centimeter of running clearance.\3057\ 
They agreed that flexible components that bend without breaking and 
return to their original position do not count against the 20-
centimeter running clearance requirement.\3058\ They disagreed with 
NHTSA's position that these requirements should apply to all vehicles 
with standard and optional equipment installed at the time of the first 
retail sale and proposed instead that the requirement should be ``as 
shipped to the dealer.'' \3059\ Additionally, the Auto Alliance asked 
NHTSA to make a specific allowance for vehicles that have adjustable 
ride height, such as air suspension, and permit the running clearance 
and other off-road clearance measurements to be made in the lifted or 
off- road mode.\3060\ Hyundai and Kia urged NHTSA not to modify the 
definition of ``running clearance,'' which currently is defined as 
``the distance from the surface on which an automobile is standing to 
the lowest point on the automobile, excluding unsprung weight.'' \3061\
---------------------------------------------------------------------------

    \3057\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3058\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3059\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3060\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3061\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411; 
Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
---------------------------------------------------------------------------

    Based upon the comments above, NHTSA has decided to retain its 
running clearance requirements for qualifying light trucks without 
change. First, running clearance means the distance from the surface on 
which an automobile is standing to all fixed components under the 
vehicle, excluding unsprung components, axle clearance components and 
flexible components that bend without breaking and returning to their 
original position as explained in NHTSA's previous interpretation. 
Second, NHTSA acknowledges that at this time, during validation testing 
for running clearance, a vehicle with optional equipment installed will 
only be tested ``as shipped to the dealer.'' NHTSA has found that 
optional equipment can impact a vehicle's ability to comply with 
running clearance requirements, while optional equipment must be 
considered for other light truck agency validation tests unless the 
equipment has no impact on the outcome of the test.
(4) Front and Rear Axle Clearance
    NHTSA regulations state that front and rear axle clearances of not 
less than 18 centimeters are another criterion that can be used for 
designating a vehicle as off-highway capable.\3062\ The agency defines 
``axle clearance'' as the vertical distance from the level surface on 
which an automobile is standing to the lowest point on the axle 
differential of the automobile.\3063\
---------------------------------------------------------------------------

    \3062\ 49 CFR 523.5(b)(2)(v).
    \3063\ 49 CFR 523.2.
---------------------------------------------------------------------------

    The agency believes this definition may be outdated because of 
vehicle design changes, including axle system components and 
independent front and rear suspension components. In the past, 
traditional light trucks with and without 4WD systems had solid rear 
axles with center- mounted differentials on the axle. For these trucks, 
the rear axle differential was closer to the ground than any other axle 
or suspension system component. This traditional axle design still 
exists today for some trucks with a solid chassis (also known as body-
on-frame configuration). Today, however, many SUVs and CUVs that 
qualify as light trucks are constructed with a unibody frame and have 
unsprung (e.g., control arms, tie rods, ball joints, struts, shocks, 
etc.) and sprung components (e.g., the axle subframes) connected 
together as a part of the axle assembly.\3064\ These unsprung and 
sprung components are located under the axles, making them lower to the 
ground than the axles and the differential, and were not contemplated 
when NHTSA established the definition and the allowable clearance for 
axles. The definition also did not originally account for 2WD vehicles 
with GVWRs greater than 6,000 pounds that had one axle without a 
differential, such as the model year 2018 Ford Expedition. Vehicles 
with axle components that are low enough to interfere with the 
vehicle's ability to perform off-road would seem inconsistent with the 
regulation's intent of ensuring off-highway capability, as Congress 
required.\3065\
---------------------------------------------------------------------------

    \3064\ Unibody frames integrate the frame and body components 
into a combined structure.
    \3065\ 49 U.S.C. 32901(a)(18)(A).
---------------------------------------------------------------------------

    In light of these issues, comments were sought in the NPRM on 
whether (and if so, how) to revise the definition of axle clearance. 
NHTSA sought comments on what unsprung axle components should be 
considered when determining a vehicle's axle clearance. The agency 
questioned whether the definition for axle clearance should be modified 
to account for axles without differentials. NHTSA also sought comment 
on whether the axle subframes surrounding the axle components but 
affixed directly to the vehicle unibody as sprung mass (lower to the 
ground than the axles) should be considered in the allowable running 
clearance discussed above. Finally, NHTSA sought comments on whether it 
should consider replacing both the running and axle clearance criteria 
with a single ground clearance criterion that considers all components 
underneath the vehicle that impact a vehicle's off-road capability.
    Comments were received from the Auto Alliance, Fiat Chrysler, 
Hyundai, and Kia. All the manufacturers that commented claimed no 
change is needed to the current definition, regardless of whether the 
axle components are sprung or unsprung masses, as the bottom of the 
differential is the vulnerable component.\3066\ The Auto Alliance also 
stated there is no

[[Page 25205]]

need to further modify the definition to account for axles without 
differentials. Further, the Auto Alliance does not think a single 
criterion that considers all components under the axle is needed and 
prefers to keep the existing regulation.\3067\ Fiat Chrysler and the 
Auto Alliance also recommended that 2WD SUVs and CUVs be reclassified 
back into the truck fleet, where they had been placed prior to the 2011 
MY. Their position is that 2WD SUVs are designed to meet the ``off-
road-capable'' definition in NHTSA's rules by having the required 
running and/or axle clearances as well as meeting other off-road 
dimensional criteria.\3068\ Hyundai stated that changing the point of 
measurement now would have significant development and economic 
impacts.\3069\ Kia stated that it has designed its vehicles and 
developed product plans in reliance on the current definitions, and 
those designs and product plans cannot be modified cheaply or 
quickly.\3070\
---------------------------------------------------------------------------

    \3066\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Hyundai, 
Detailed Comments, EPA-HQ-OAR-2018-0283-4411; Kia, Detailed 
Comments, EPA-HQ-OAR-2018-0283-4195.
    \3067\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3068\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; 
Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3069\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411.
    \3070\ Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
---------------------------------------------------------------------------

    NHTSA already addressed the comments on 2WD SUVs in a previous 
rulemaking, and NHTSA has no additional response at this time.\3071\ 
Upon review of other comments, manufacturers did not clearly 
distinguish which parts of the axle sub-frames should be considered as 
sprung masses in order for NHTSA to understand if modifications are 
needed to its axle clearance requirements. Therefore, at this time, 
NHTSA is retaining its axle clearance requirements as currently 
specified. However, NHTSA still believes it is beneficial to continue 
efforts at defining those axle components that are sprung or unsprung 
masses before considering any changes to its regulatory provisions. In 
addition, NHTSA needs to understand any significant developmental and 
economic impacts that might be associated with any possible changes to 
its requirements. Therefore, NHTSA will consider collecting further 
information on these issues and may take further action related to this 
issue in the future.
---------------------------------------------------------------------------

    \3071\ No new arguments have been raised beyond those already 
considered in the April 6, 2006, final rule (see 71 FR 17566).
---------------------------------------------------------------------------

B. EPA Compliance and Enforcement

1. Overview of the EPA Compliance Process
    EPA established comprehensive vehicle certification, compliance, 
and enforcement provisions for the GHG standards as part of the 
rulemaking establishing the initial GHG standards for MY 2012-2016 
vehicles.\3072\ Manufacturers have been following these provisions 
since MY 2012 and EPA did not propose or seek comments on changing its 
compliance and enforcement program.
---------------------------------------------------------------------------

    \3072\ See 75 FR 25468-25488 and 77 FR 62884-62887 for a 
description of these provisions. See also ``The 2018 EPA Automotive 
Trends Report, Greenhouse Gas Emissions, Fuel Economy, and 
Technology since 1975,'' EPA-420-R-19-002 March 2019 for additional 
information regarding EPA compliance determinations.
---------------------------------------------------------------------------

a) What Compliance Flexibilities and Incentives are Currently Available 
Under the CO2 Program and How Do Manufacturers Use Them?
    Under EPA's regulations, manufacturers can use credit flexibilities 
to comply with CO2 standards for passenger car or light 
truck compliance fleets. Similar to the CAFE program, manufacturers 
gain credits when the performance of a fleet exceeds its required 
CO2 fleet average standard which can be carried forward for 
five years. EPA also allows a one-time credit carry-forward exceeding 5 
years, allowing MY 2010-2015 to be carried forward through MY2021. A 
manufacturer's fleet performance that does not meet the fleet average 
standard generates a credit deficit. Manufacturers can carry credit 
deficits forward up to three model years before having to resolve the 
shortfall.
    NHTSA's program continues the 5-year carry-forward and 3-year 
carryback, as required by statute. Credit ``transfer'' means the 
ability of manufacturers to move credits from their passenger car fleet 
to their light truck fleet, or vice versa. As part of the EISA 
amendments to EPCA, NHTSA was required to establish by regulation a 
CAFE credit transferring program, now codified at 49 CFR part 536, to 
allow a manufacturer to transfer credits between its car and truck 
fleets to achieve compliance with the standards. For example, credits 
earned by over-compliance with a manufacturer's car fleet average 
standard could be used to offset debits incurred because the 
manufacturer did not meet the truck fleet average standard in a given 
year.
    Under Section 202(a) of the CAA, there is no statutory limitation 
on car/truck credit transfers, and EPA's CO2 program allows 
unlimited credit transfers across a manufacturer's car and light truck 
fleets to meet CO2 standards.
    EPA requested comment on a variety of ``enhanced flexibilities'' 
whereby EPA could make adjustments to current incentives and credit 
provisions and potentially add new flexibility opportunities to expand 
the means by which manufacturers may satisfy standards. Some of these 
additional flexibilities would not result in a reduction in program 
stringency, while others would incentivize technologies that could 
realize greater CO2 emissions reductions over a longer term, 
but would result in a loss of emission benefits in the short-term, as 
discussed below. EPA requested comments on these topics to support the 
increased application of technologies that the automotive industry is 
developing and deploying that could potentially lead to further long-
term emissions reductions and allow manufacturers to comply with 
standards while reducing costs.
    EPA explained that one category of flexibilities, such as off-cycle 
credits and credit banking, involve credits that are based on real 
world emissions reductions and do not represent a loss of overall 
emissions benefits or a reduction in program stringency, yet offer 
manufacturers potentially lower-cost or more efficient path to 
compliance. Another category of flexibilities, such as incentives for 
battery electric vehicles, hybrid technologies, and alternative fuels, 
do result in a loss of emissions benefit and represent a reduction in 
the effective stringency of the standards to the extent the incentives 
are used by manufacturers. These incentives would help manufacturers 
meet a numerically more stringent standard, but would not reduce real-
world CO2 emissions in the short term compared to a lower 
stringency option with fewer such incentives. EPA's policy rationale 
for providing such incentives, as articulated in the 2012 rulemaking, 
was that such programs could incentivize the development and deployment 
of advanced technologies with the potential to lead to greater 
CO2 emissions reductions in the longer-term, where such 
technologies today are limited by higher costs, market barriers, 
infrastructure, and consumer awareness.\3073\ Such incentive approaches 
would also result in rewarding automakers who invest in certain 
technological pathways, rather than being technology neutral.
---------------------------------------------------------------------------

    \3073\ See 77 FR 62810-62826 (Oct. 15, 2012).
---------------------------------------------------------------------------

    Prior to the proposal, automakers and other stakeholders expressed 
support for

[[Page 25206]]

this type of compliance flexibility. For example, in March 2018, Ford 
stated, ``We support increasing clean car standards through 2025 and 
are not asking for a rollback. We want one set of standards nationally, 
along with additional flexibility to help us provide more affordable 
options for our customers.'' \3074\ Honda, in April 2018, also 
expressed its support for an approach that retained the existing 
standards while extending the advanced technology multipliers for 
electrified vehicles, eliminated automakers' responsibility for the 
impact of upstream emissions from the electric grid, and accommodated 
more off-cycle technologies.\3075\
---------------------------------------------------------------------------

    \3074\ ``A Measure of Progress'' Bill Ford, Executive Chairman, 
Ford Motor Company, and Jim Hackett, President and CEO, Ford Motor 
Company, March 27, 2018, https://medium.com/cityoftomorrow/a-measure-of-progress-bc34ad2b0ed.
    \3075\ Honda Release ``Our Perspective--Vehicle Greenhouse Gas 
and Fuel Economy Standards,'' April 20, 2018, http://news.honda.com/newsandviews/pov.aspx?id=10275-en.
---------------------------------------------------------------------------

    EPA's request for comments was largely based on its consideration 
of input from automakers and other stakeholders, including suppliers 
and alternative fuels industries, supporting a variety of program 
flexibilities.\3076\ The following provides an overview of EPA's 
request for comments on several flexibility concepts, the comments EPA 
received, and the agency's response to those comments. After 
considering comments, EPA is not adopting new incentives in the areas 
of credit multipliers (with the exception of multipliers for natural 
gas vehicles), new incentives for hybrid vehicles, incentives for 
autonomous or connected vehicles, or alternative fueled vehicles other 
than natural gas, as part of this final rule. EPA is finalizing program 
changes for the treatment of upstream emissions for electric vehicles, 
the treatment of natural gas vehicles, the treatment of hybrid and 
target-beating full-size pickup trucks, and off-cycle credits, as 
discussed below.
---------------------------------------------------------------------------

    \3076\ Memorandum to docket EPA-HQ-OAR-2018-0283 regarding 
meetings with the Alliance of Automobile Manufacturers on April 16, 
2018 and Global Automakers on April 17, 2018. EPA-HQ-OAR-2018-0283-
0022.
---------------------------------------------------------------------------

(1) Credit Flexibilities
    Under the EPA program, CO2 credits may be carried 
forward, or banked, for a period of five years, with the exception that 
MY 2010-2015 credits may be carried forward and used through MY 2021. 
CO2 credits may also be traded between manufacturers and 
transferred between passenger car and light truck fleets similar to the 
CAFE program, but without any adjustment for fuel savings. Under 
Section 202(a) of the CAA, there is no statutory limitation on credit 
transfers between a manufacturer's passenger car and light truck 
fleets, and EPA's CO2 program allows unlimited credit 
transfers across a manufacturer's passenger car and light truck fleets 
to comply with CO2 standards. This flexibility is based on 
the expectation that it will help facilitate manufacturer compliance 
with CO2 standards in the lead time provided, and allow 
CO2 emissions reductions to be achieved in the most cost 
effective way.
    Automakers suggested, prior to the NPRM proposal, a variety of ways 
in which CO2 credit life could be extended under the CAA, 
like allowing automakers to carry-forward MY 2010 and later banked 
credits to MY 2025, extending the life of credits beyond five years, or 
even unlimited credit life where credits would not expire. EPA 
requested comments in the NPRM on extending credit carry-forward under 
the CO2 program beyond the current five years, including 
unlimited credit life.
    General comments were received in response to the NPRM from the 
National Automobile Dealers Association and Volkswagen. They commented 
that credit carry-forward and carryback options help with annual 
compliance with the CO2 program.\3077\ They stated that 
these mechanisms allow manufacturers to become compliant over the 
course of the time a credit is usable in the market.\3078\ Toyota, 
General Motors, Fiat Chrysler, the Auto Alliance, and the Global 
Automakers each commented that CO2 credits earned by 
manufacturers need a longer life so they may be carried forward further 
than the current five-year limitation.\3079\ They asked for an 
unlimited period for using CO2 credits without restrictions, 
since they argue that automakers have earned those credits and should 
be allowed to use them however they see fit.\3080\ They also stated 
that this would incentivize manufacturers to make early reductions in 
CO2 emissions.\3081\ Furthermore, it was noted that credits 
are earned when manufacturers achieve lower CO2 fleet 
average emissions than otherwise required by regulation in any given 
model year. They stated that this typically results from actions taken 
by a manufacturer to deploy specific models or more efficient 
technology than required, often at a higher cost. Such technologies 
reduce the amount of CO2 emissions released into the 
atmosphere over the life of the vehicle, which could be over several 
decades. Therefore, the resulting credit earned by a manufacturer for 
having made the product or technology investment that resulted in the 
reduced emissions should not be limited to five years.
---------------------------------------------------------------------------

    \3077\ National Automobile Dealers Association, Detailed 
Comments, NHTSA-2018-0067-12064; Volkswagen, Detailed Comments, 
NHTSA-2017-0069-0583.
    \3078\ See, e.g., National Automobile Dealers Association, 
NHTSA-2018-0067-12064.
    \3079\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; General 
Motors, Detailed Comments, NHTSA-2018-0067-11858; Fiat Chrysler, 
Detailed Comments, NHTSA-2018-0067-11943; Auto Alliance, Detailed 
Comments, NHTSA-2018-0067-12073; Global Automakers, Detailed 
Comments, NHTSA-2018-0067-12032.
    \3080\ See, e.g., Global Automakers, Detailed Comments, NHTSA-
2018-0067-12032.
    \3081\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
0067-11858.
---------------------------------------------------------------------------

    Global Automakers, the Auto Alliance, Fiat Chrysler, and Toyota 
requested a one-time expiration date extension through 2026 for 
CO2 credits earned in MYs 2010-2015.\3082\ They asserted 
that earned credits represent actual CO2 reductions and 
increasing their lifespan will allow for better compliance. Conversely, 
Honda disagreed with the extension of MY 2010-2015 credits through 2026 
because they have been selling their credits under the assumption that 
they would expire.\3083\ Honda stated that shorter life (soon to 
expire) credits are worth less than longer life credits, leading to a 
disadvantage for manufacturers who have already sold these credits at a 
lower price. Honda asserted that the one-time extension would benefit 
only a few automakers.\3084\ However, Honda did agree that a one-time 
extension through 2026 for MYs 2016-2020 CO2 credits would 
assist with compliance because these credits have yet to be involved in 
trades.\3085\
---------------------------------------------------------------------------

    \3082\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032; Alliance, Detailed Comments, NHTSA-2018-0067-12073; Fiat 
Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Toyota Detailed 
Comments, NHTSA-2018-0067-12150.
    \3083\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3084\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3085\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
---------------------------------------------------------------------------

    In sum, commenters requested either unlimited allowances to carry-
forward surplus credits without any expiration date, a one-time 
expiration date extension through 2026 for CO2 credits 
earned from MY 2010 and later, or consideration for extending credit 
life longer than the current five-year provision. After considering the 
comments received, EPA has decided not to change its credit carry-
forward provisions at this time, and will retain the credit carry-
forward period under the CO2 program at five years for 
credits

[[Page 25207]]

generated in MYs 2016 and later. EPA does not believe any changes to 
its credit carry-forward provisions are warranted. EPA notes that 
NHTSA's CAFE program is constrained by statute to a five-year carry-
forward so if EPA adopted a longer carry-forward period, it might be of 
limited use since the level of stringency of the CO2 and 
CAFE standards is similar across the programs. Also, the analysis on 
which the tailpipe CO2 emissions standards finalized today 
are based, assumed a five-year carry-forward period for credits.
    Another reason for denying manufacturers' requests is the potential 
inequitable advantage a longer credit life could have for manufacturers 
with surplus credits, especially those with significant amounts of 
credits currently banked for multiple model years. Manufacturers 
without credits, or manufacturers who have already sold their credits 
at current market values based on the present five-year carry-forward 
credit lifespan, as Honda discussed, will be significantly 
disadvantaged.\3086\ These manufacturers are unlikely to be able to 
renegotiate the price of credit trades already made. Manufacturers with 
large amounts of credits would clearly be advantaged and able to 
distort the market in ways unfavorable to the goal of reducing 
emissions. EPA is concerned that these manufacturers will be able to 
create uncertainties in the market by being able to infuse large 
volumes of credits into future model years where it may even be 
possible to delay some cost-effective technologies from entering 
production because manufacturers are relying upon these credits as an 
alternative pathway to compliance.
---------------------------------------------------------------------------

    \3086\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
---------------------------------------------------------------------------

(2) Advanced Technology Incentives
    The existing EPA CO2 program provides incentives for 
electric vehicles, fuel-cell vehicles, plug-in hybrid vehicles, and 
natural gas vehicles. The 2012 rulemaking allowed manufacturers to use 
a 0 grams/mile emissions factor for all electric powered vehicles 
rather than having to account for the CO2 emissions 
associated with upstream electricity generation, up to a per-
manufacturer cumulative production cap for MYs 2022-2025. The program 
also includes multiplier incentives that allow manufacturers to count 
advanced technology vehicles as more than one vehicle in the compliance 
calculations. The multipliers began with MY 2017 and end after MY 
2021.\3087\ Prior to the proposal, stakeholders suggested that these 
incentives should be expanded to support further the production of 
advanced technologies by allowing manufacturers to continue to use the 
0 grams/mile emissions factor for electric powered vehicles rather than 
having to account for upstream electricity generation emissions and by 
extending and potentially increasing the multiplier incentives.
---------------------------------------------------------------------------

    \3087\ The multipliers are for EV/FCVs: 2017-2019--2.0, 2020--
1.75, 2021--1.5; for PHEVs and dedicated and dual-fuel CNG vehicles: 
2017-2019--1.6, 2020--1.45, 2021--1.3.
---------------------------------------------------------------------------

    First, EPA requested comments on extending the use of 0 grams/mile 
emissions factor for electric powered vehicles.
    The Auto Alliance, Global Automakers, and several manufacturers 
commented that upstream utility emissions come from power plants, not 
vehicle tailpipes, and manufacturers have no control over the feedstock 
used by those power plants and should not be held responsible for their 
upstream electricity emissions.\3088\ The Auto Alliance further 
commented that removing upstream accounting is not an incentive for 
advanced technology vehicles; rather, it should be seen as a correction 
to remove responsibility for emissions over which the automakers have 
no control.\3089\ Fiat Chrysler commented that ``requiring upstream 
accounting could impede development of BEVs or PHEVs, as accounting of 
upstream emissions degrades the CO2 performance of BEVs to 
the level of PHEVs, and PHEVs to the level of a conventional hybrid 
electric vehicle. This, in effect, disincentivizes the technology.'' 
\3090\
---------------------------------------------------------------------------

    \3088\ See, e.g., Volkswagen, Detailed Comments, NHTSA-2017-
0069-0583.
    \3089\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3090\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    Several other commenters also supported not counting upstream 
emissions and instead only counting electric powered vehicle tailpipe 
emissions of 0 grams/mile.\3091\ These commenters included NCAT, SAFE, 
BorgWarner, CALSTART, Eaton, and Edison Electric Institute.
---------------------------------------------------------------------------

    \3091\ See, e.g., NCAT, NHTSA-2018-0067-11969.
---------------------------------------------------------------------------

    API did not support continuing the 0 grams/mile emission factor for 
electricity use, commenting that by failing to factor the real 
contribution of upstream CO2 emissions from electric 
generation, the regulatory agencies would distort the market for 
developing transportation fuel alternatives.\3092\ API commented that 
EPA should not ignore the environmental burden of upstream emissions in 
granting production incentives to automakers.
---------------------------------------------------------------------------

    \3092\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
---------------------------------------------------------------------------

    Manufacturers of Emission Controls Association (MECA) commented 
that ``with the growing emphasis on real-world emission reductions, it 
becomes increasingly important to consider all emissions to the 
environment, including upstream emissions. Numerous studies have shown 
that in many parts of the country, the temporary 0 grams/mile upstream 
emissions factor is not delivered in the real-world . . . MECA believes 
that EPA should continue to set performance-based standards that assess 
technology pathways based on delivering the intended emission 
reductions over the full well-to-wheels vehicle life cycle in the real-
world.'' \3093\ Motor & Equipment Manufacturers Association (MEMA) also 
supported a well-to-wheel fuel lifecycle approach, commenting that 
without this type of comprehensive assessment on the fuel impacts and 
comprehensive CO2 costs, policies improperly ``slant toward 
preferred technologies.'' \3094\ Nonetheless, MEMA commented that it is 
not opposed to continuing to allow 0 grams/mile emissions factor for 
electric powered vehicles through 2026.
---------------------------------------------------------------------------

    \3093\ MECA, Detailed Comments, NHTSA-2018-0067-11994.
    \3094\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

    The Union of Concerned Scientists (UCS) commented that not 
accounting for upstream emissions combined with the multipliers has a 
significant impact on the efficacy of the standard, and extending these 
regulatory incentives is more likely to result in a credit giveaway 
than to drive additional deployment of electric vehicles.\3095\ UCS 
further commented that, to date, more than half of the electric 
vehicles sold have been in California and the states that have adopted 
California's ZEV standards; however, UCS asserted, federal standards 
ignore the upstream emissions for all vehicles sold.
---------------------------------------------------------------------------

    \3095\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    After carefully considering the wide range of comments on whether 
to include upstream emissions associated with electricity use in the 
compliance calculations for electrified vehicles, EPA has decided to 
allow the continued use of the 0 grams/mile emissions factor with no 
per-manufacturer production caps or other limitations. EPA is revising 
its regulations to remove the production caps and related provisions. 
When EPA initially adopted a production cap for manufacturers that

[[Page 25208]]

use the 0 grams/mile emissions factor, in the rulemaking to establish 
CO2 standards for MY 2012-2016 vehicles, there were no 
controls in place for CO2 emissions from electricity 
production.\3096\ This was also the case when EPA extended the 0 grams/
mile upstream provision and revised the production caps in the rule 
establishing MY 2017-2025 standards.\3097\ However, since then, EPA has 
adopted a program to control CO2 emissions from power 
plants.\3098\ Emissions from the power sector have been declining and 
that trend is projected to continue.\3099\ For these reasons, EPA no 
longer views the upstream emissions factor as an incentive in the same 
way it views a multiplier incentive which provides bonus credits. EPA 
agrees that, at this time, manufacturers should not account for 
upstream utility emissions. Therefore, EPA is adopting regulatory 
changes consistent with its historical practice of basing compliance 
with vehicle emissions standards on tailpipe emissions through model 
year 2026. EPA may choose to reconsider this decision in a future 
CO2 rulemaking, and will reexamine the issue when 
establishing standards commencing with the 2027 model year.\3100\
---------------------------------------------------------------------------

    \3096\ 75 FR 25341, May 7, 2010.
    \3097\ 77 FR 62816, October 15, 2012.
    \3098\ 84 FR 32520, July 8, 2019.
    \3099\ 84 FR 32561.
    \3100\ By comparison, the CAFE program uses an energy efficiency 
metric instead of an emissions metric, and standards that are 
expressed in miles per gallon. For PHEVs and BEVs, to determine 
gasoline the equivalent fuel economy for operation on electricity, a 
Petroleum Equivalency Factor (PEF) is applied to the measured 
electrical consumption. The PEF for electricity was established by 
the Department of Energy, as required by statute, and includes an 
accounting for upstream energy associated with the production and 
distribution for electricity relative to gasoline. Therefore, the 
CAFE program includes upstream accounting based on the metric that 
is consistent with the fuel economy metric. The PEF for electricity 
also includes an incentive that effectively counts only 15 percent 
of the electrical energy consumed.
---------------------------------------------------------------------------

    Second, EPA requested comments on extending or increasing advanced 
technology incentives, including multiplier incentives, with 
multipliers in the range of 2.0-4.5. EPA received a wide range of 
comments both for and against increasing the multiplier incentives. The 
MY 2017-2025 CO2 program finalized in 2012 included 
incentive multipliers for certain advanced technologies for MY 2017-
2021 vehicles.
    The Auto Alliance, Global Automakers, and several individual 
manufacturers commented in support of continued and increased 
multipliers. The Auto Alliance commented that EPA should extend and 
significantly expand multipliers ``to encourage a transition to these 
technologies while cost, range, and infrastructure challenges are 
addressed to encourage ongoing investments in advanced technologies.'' 
\3101\ Global Automakers commented that multipliers should be included 
through MY 2026, set at values that encourage ongoing investment in 
advanced technologies, without diluting overall efficiency improvements 
in the program.\3102\ NCAT, Eaton, Plug-in America, Alliance to Save 
Energy, SAFE, and MEMA also supported additional multiplier incentives 
to encourage further the production and sale of advanced technology 
vehicles.\3103\
---------------------------------------------------------------------------

    \3101\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3102\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
    \3103\ NCAT, Detailed Comments, NHTSA-2018-0067-11969; Eaton, 
Detailed Comments, EPA-HQ-OAR-2018-0283-5068; Plug-In America, 
Detailed Comments, NHTSA-2018-0067-12028; Alliance to Save Energy, 
Detailed Comments, NHTSA-2018-0067-11837; SAFE, Detailed Comments, 
NHTSA-2018-0067-11981; see https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

    EPA also received comments against extending the multiplier 
credits. UCS commented that reducing the stringency of the standards 
lessens the need for the adoption of these vehicles and undermines the 
initial rationale for these credits, resulting in a significant bank of 
credits which would further erode the benefits of these 
standards.\3104\ American Council for an Energy-Efficient Economy 
(ACEEE) commented that providing multiplier incentives for any longer 
period, or at a greater rate than those currently in place, would 
create windfall credits for manufacturers given the industry's current 
product plans.\3105\ Fiat Chrysler commented generally in support of a 
multiplier incentive, but noted that since multipliers are a 
CO2--only flexibility not present in the CAFE program, 
greater use of multipliers would result in further disharmonizing the 
programs.\3106\ API commented against multipliers, stating that the 
program should be technology neutral and that regulatory agencies 
should not incentivize either producer or consumer investments in 
government-selected technologies applied to government-selected vehicle 
categories.\3107\
---------------------------------------------------------------------------

    \3104\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3105\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
    \3106\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3107\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
---------------------------------------------------------------------------

    In this final rule, EPA is neither adopting any additional EV or 
FCV multipliers nor extending the existing multipliers scheduled to 
phase out after MY 2021 for EVs, PHEVs, and FCVs. EPA is concerned that 
additional multiplier incentives beyond those already in place for 
these vehicles which are currently available to consumers would reduce 
the emissions benefits associated with the program. As discussed below 
in section IX.B.1.a.(3)(b), EPA is providing an additional multiplier 
for dedicated and dual-fuel NGVs, which are not currently produced by 
auto manufacturers, for MYs 2022-2026. The CO2 program 
already provides a significant incentive for PHEVs, EVs, and FCVs by 
only counting tailpipe emissions (not accounting for upstream 
emissions).
(3) Special Considerations
(a) Incentives for Connected or Automated Vehicles
    Connected and automated (including autonomous) vehicles have the 
potential to impact significantly vehicle emissions in the future, with 
their aggregate impact being either positive or negative, depending on 
a large number of vehicle-specific and system-wide factors. EPA noted 
in the proposal that connected or automated vehicles would be eligible 
for credits under the off-cycle program if a manufacturer provides data 
sufficient to demonstrate the real-world emissions benefits of such 
technology applied to its vehicles. However, demonstrating the 
incremental real-world benefits of these emerging technologies will be 
challenging. Prior to the proposal, stakeholders suggested that EPA 
should consider an incentive for these technologies without requiring 
individual manufacturers to demonstrate real-world emissions benefits 
of the technologies. A number of stakeholders also requested that EPA 
consider credits for automated and connected vehicles that are placed 
in ridesharing or other high mileage applications, where any potential 
environmental benefits could be multiplied due to the high utilization 
of these vehicles. EPA requested comment on such incentives as a way to 
facilitate increased use of these technologies, including some level of 
assurance that they will lead to future additional emissions 
reductions. For example, EPA stated in the proposal that any near-term 
incentive program should include some demonstration that the 
technologies will be both truly new and have some connection to overall 
environmental benefits. EPA further outlined and sought comment on 
several approaches

[[Page 25209]]

to incentivize automated and connected vehicle technologies.
    EPA received comments supporting and opposing incentives for 
automated and connected vehicles. The Auto Alliance commented that the 
agencies should incentivize the adoption of these technologies and 
provide for possibly additional credit once the benefits beyond the 
credit values have been confirmed.\3108\ It further commented that a 
growing body of modeling results, as well as real-world driving 
statistics, show that current automated driving technologies improve 
real-world fuel efficiency and reduce CO2 emissions. SAFE 
commented that connected automated vehicles have tremendous potential 
to save lives, and when combined with ride-sharing and electric 
powertrains, they can also increase efficiencies and save fuel.\3109\ 
SAFE argued that an initial review of the literature shows the 
potential for these technologies to improve fuel economy by up to 25 
percent when they are optimized and aggregated alongside other 
traditional efficiency technologies. Toyota commented that automated 
vehicles, and possibly new mobility models such as ridesharing, can 
help attain societal goals concerning climate change, energy security, 
traffic congestion, and safety.\3110\ Ford commented that it is 
supportive of credits for future connected and automated vehicles and 
that autonomous vehicles are considered the future of personal 
mobility, with many manufacturers announcing plans to release 
autonomous-capable vehicles in the near term.\3111\ Ford added that 
these vehicles have the potential to not only provide meaningful real-
world CO2 and fuel economy benefits, but also add true 
societal benefit for the public good by providing transportation to 
those who would otherwise not have access. General Motors and Jaguar 
Land Rover commented in favor of additional credits for vehicles placed 
in ride-sharing or high mileage applications.\3112\
---------------------------------------------------------------------------

    \3108\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3109\ SAFE, Detailed Comments, NHTSA-2018-0067-11981.
    \3110\ Toyota, Detailed Comments, NHTSA-2018-0067-12150.
    \3111\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
    \3112\ General Motors, Detailed Comments, NHTSA-2018-0067-11858; 
Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-11916.
---------------------------------------------------------------------------

    SAFE commented that autonomous vehicles will lead to new jobs and 
better worker productivity. It stated that these vehicles will also 
reduce congestion and lead to safer travel.\3113\
---------------------------------------------------------------------------

    \3113\ SAFE, Detailed Comments, NHTSA-2018-0067-11981.
---------------------------------------------------------------------------

    Other commenters opposed incentives for automated and connected 
vehicles, generally commenting that while the technologies are 
promising, the impacts of the technologies remain highly uncertain and 
therefore incentives are not appropriate. ACEEE commented that EPA 
should not incentivize technologies such as automated vehicle 
technology or ridesharing services, unless and until it can be 
demonstrated that such an incentive will result in emissions reduction 
benefits and will not undermine the existing standards.\3114\ ACEEE 
believes that there currently exists no real-world data to justify 
granting of off-cycle credits for automated vehicle technologies, and 
that providing automakers credits for deploying technologies which are 
driven by demands other than fuel savings and emissions reduction only 
allows them to make fewer real-world emissions reductions elsewhere. 
ACEEE further stated that while automated vehicles promise all-new 
possibilities and efficiencies in transportation and the use of 
infrastructure, the net impact on transportation sector energy use and 
emissions is unknown.
---------------------------------------------------------------------------

    \3114\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
---------------------------------------------------------------------------

    UCS commented that the ``evidence to-date does not warrant 
incentivizing such technologies--there is no provable environmental 
benefit of such technologies, and the agencies have previously 
correctly acknowledged that any such potential impacts would be related 
to indirect benefits, which raise serious concerns about compliance and 
enforcement to ensure the integrity of the program.'' \3115\ Honda 
commented that there remains considerable uncertainty in the literature 
regarding the energy and environmental benefits (or negative benefits) 
of connected/automated vehicle technology.\3116\ Honda commented that 
if technology benefits can be verified under robust, repeatable 
conditions, they should warrant off-cycle credits under the existing 
off-cycle program. Honda does not believe credits should be granted for 
application of technology alone.
---------------------------------------------------------------------------

    \3115\ U.S.C., Detailed Comments, NHTSA-2018-0067-12039.
    \3116\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
---------------------------------------------------------------------------

    CARB commented that new compliance flexibilities (or off-cycle 
credit categories) for automated vehicles are not appropriate at this 
time.\3117\ CARB believes that, although the technology is widely 
expected to provide safety and mobility benefits, automakers are 
expected to bring the technology to market regardless, so incentives 
are unnecessary, and it is not established that these technologies will 
reduce emissions given their potential for high annual mileage. 
Resources for the Future commented they do not see a rationale for 
providing special credits to automated vehicles since such vehicles 
could increase or decrease emissions.\3118\ Competitive Enterprise 
Institute (CEI) commented that some connected and/or automated vehicle 
technology applications--namely platooning--may improve fuel efficiency 
through improved aerodynamics and thus reduce CO2 emissions; 
however, such applications to date are limited to heavy-vehicle 
prototypes beyond the scope of this rulemaking and in any event should 
be subject to verification prior to any award of off-cycle 
credits.\3119\ CEI commented further: ``We urge EPA to preserve the 
existing off-cycle program requirement that manufacturers demonstrate 
CO2 emissions reductions prior to the award of credits, 
rather than picking technology winners and losers that have nothing to 
do with fuel economy or emissions.'' National Association of Truck Stop 
Operators (NATSO) commented against incentives, stating that although 
automated vehicles have the potential positively to transform 
transportation (and indeed day-to-day life) in the U.S., there are also 
a number of complexities and potential costs associated with 
them.\3120\
---------------------------------------------------------------------------

    \3117\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
    \3118\ Resources for the Future, Detailed Comments, NHTSA-2018-
0067-11789.
    \3119\ CEI, Detailed Comments, EPA-HQ-OAR-2018-0283-4166.
    \3120\ NATSO, Detailed Comments, EPA-HQ-OAR-2018-0283-5484.
---------------------------------------------------------------------------

    EPA is not adopting new incentives for automated and connected 
vehicles. While EPA agrees there may be potential for such technologies 
to reduce emissions long-term, depending on how the technologies are 
developed, implemented, and used, EPA remains concerned about the high 
degree of uncertainty regarding the impacts of the technologies and 
potential loss of emissions reductions associated with such incentives. 
EPA agrees with the comments that, at this time, it is more appropriate 
for manufacturers to seek credits through the existing off-cycle 
credits program where manufacturers would be required to provide data 
demonstrating direct emissions improvements for the technologies.

[[Page 25210]]

(b) Natural Gas Vehicle (NGV) Credits
    Vehicles that are able to run on compressed natural gas (CNG) are 
eligible for an advanced technology multiplier credit for MYs 2017-
2021, as discussed in the Advanced Technology Incentives section above. 
Dual-fueled natural gas vehicles, which can run either on natural gas 
or on gasoline, also may use utility factors higher than 0.5 when 
weighting tailpipe emissions measured over the test procedures while 
operating on natural gas and gasoline test fuels if the vehicles meet 
minimum design criteria, including minimum CNG range requirements. 
Prior to the proposal, EPA received input from several industry 
stakeholders that supported expanding these incentives to stimulate 
production of vehicles capable of operating on natural gas, including 
treating incentives for natural gas vehicles on par with those for 
electric vehicles and other advanced technologies, and adjusting or 
removing the minimum range requirements for dual-fueled CNG vehicles. 
EPA requested comments on these potential additional incentives for 
natural gas fueled vehicles.
    Among comments received regarding incentives for NGVs, Ariel 
Corporation and VNG together commented that NGVs can be effectively 
promoted by providing a level playing field and regulatory parity with 
EVs.\3121\ They stated, ``an effective alternative compliance pathway 
for NGVs can be established with a few simple changes to the 
regulations including applying the '0.15 divisor' to emissions 
calculations, which would harmonize EPA's regulations with the 
statutory CAFE program, and recognize the real-world emissions benefits 
of RNG [renewable natural gas], and provide NGVs with reasonable parity 
with EVs.'' Ariel and VNG commented also that EPA should offer advanced 
technology production multipliers for NGVs on par with EVs and FCVs, 
with NGVs receiving these incentives at the same level and for the same 
duration as electric and fuel-cell vehicles. These commenters believe 
that while NGVs have lower technology hurdles than these vehicles, they 
face similar infrastructure challenges and offer similar or superior 
emissions benefits through the use of RNG.
---------------------------------------------------------------------------

    \3121\ Joint Submission from Ariel Corp. and VNG.co, Detailed 
Comments, NHTSA-2018-0067-7573.
---------------------------------------------------------------------------

    Coalition for Renewable Natural Gas, NGVAmerica, the American Gas 
Association, and the American Public Gas Association commented in a 
joint submission that NHTSA and EPA should use this rulemaking 
opportunity to expand incentives for NGVs and thereby increase the 
availability of NGVs in the light-duty sector, particularly for pickup 
trucks, work vans, and sport utility vehicles.\3122\ These commenters 
also submitted comments supporting additional incentives for full-size 
pickup NGVs and incentives for vehicles equipped to be converted to 
operate on natural gas. Coalition for Renewable Natural Gas, et al., 
commented that allowing 0 grams/mile accounting for electricity use is 
favorable to electric vehicles because it allows electric vehicle 
manufacturers to take credit for anticipated improvements in emissions 
associated with the electric grid resulting from increased use of 
natural gas and renewable energy.\3123\ It further commented that given 
the significant amount of renewable natural gas currently being used 
and projected to be used in future years, using a factor of 0.15 or 
even greater to offset NGV emissions is warranted because RNG use 
reduces carbon dioxide emissions by 85 percent or more in most cases. 
Ingevity similarly commented in support of EPA including a 0.15 
multiplier incentive for purposes of CO2 compliance parity 
between natural gas and electric dual-fuel vehicles as necessary and 
critical to promote the commercialization of light-duty natural gas 
vehicles and stimulate the increased utilization of RNG. Ingevity added 
that growth in the natural gas vehicle market is necessary to meet 
future RFS obligations.\3124\
---------------------------------------------------------------------------

    \3122\ Joint Submission from the Coalition for Renewable Natural 
Gas, NGVAmerica, the American Gas Association, and the American 
Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
    \3123\ Joint Submission from the Coalition for Renewable Natural 
Gas, NGVAmerica, the American Gas Association, and the American 
Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
    \3124\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666.
---------------------------------------------------------------------------

    United States Senator James M. Inhofe commented that ``even if all 
current incentives for EVs are eliminated, EVs still have a compliance 
advantage going forward. This is because the policy and technical 
approaches underlying the [CO2] regulations embedded 
preferential treatment for the previous administration's favored 
technology. I respectfully ask you not to give NGVs preferential 
treatment, but to level the playing field to allow the marketplace to 
determine the future of NGV adoption and not the federal bureaucracy. 
To achieve this parity, reinstating the 0.15 [CO2] 
multiplier is essential.'' \3125\
---------------------------------------------------------------------------

    \3125\ James M. Inhofe, Detailed Comments, EPA-HQ-OAR-2018-0283-
7456.
---------------------------------------------------------------------------

    In addition to supporting the application of a 0.15 factor, some in 
the natual gas industry also commented in support of production 
multipliers for NGVs. Ariel and VNG commented that EPA should offer 
advanced technology production multipliers for NGVs on par with EVs and 
FCVs, with NGVs receiving these incentives at the same level and for 
the same duration as electric and fuel cell vehicles. Ingevity 
commented that dual-fuel and dedicated NGV multipliers should be 
extended through 2025 as an effective way to promote the 
commercialization of these kinds of vehicles by the automakers. NGV 
America et al. commented that ``NGVs, both dedicated and dual-fuel, 
should be provided with the same vehicle production multiplier credits 
as have previously been, and continue to be, provided to EVs and FCVs. 
Given that the expected and likely range capabilities of NGVs will 
generally exceed EV ranges (including natural gas dual-fuel vehicles 
that significantly outperform the range capabilities of PHEVs which 
justifiably enjoy a lower multiplier as compared to EVs), the vehicle 
production multipliers that are used for EVs should be applied to NGVs, 
including dual fuel NGVs. Specifically, dedicated and dual-fuel NGVs 
(or all covered advanced technology vehicles) should receive a base 
multiplier of 2.0 (or any such higher multiplier afforded to EVs/FCVs) 
for at least model years 2019 through 2021 and the same multipliers 
afforded to EVs/FCVs thereafter through 2025.''
    National Association of Convenience Stores (NACS) and the Society 
of Independent Gasoline Marketers of America (SIGMA) commented, ``the 
Associations urge you to treat all fuels and technologies equally, 
including NGVs, EVs, and petroleum-based motor fuels. It is the role of 
the Agencies to set performance specifications via notice-and-comment 
rulemaking to ensure that they are appropriate. Once the specifications 
are set, however, it should be up to the market to determine how best 
to meet them.'' \3126\
---------------------------------------------------------------------------

    \3126\ Joint submission on behalf of NACS and SIGMA, Detailed 
Comments, EPA-HQ-OAR-2018-0283-5824.
---------------------------------------------------------------------------

    UCS commented that natural gas is a potent greenhouse gas, and any 
direct emissions of methane pose a significant threat to any effort to 
limit climate change.\3127\ UCS stated, ``these direct emissions 
upstream significantly

[[Page 25211]]

undermine any potential benefit that could come from the pump-to-wheel 
benefits of displacing gasoline or diesel with natural gas.'' UCS also 
commented, ``furthermore, the technology underpinning any natural gas-
powered vehicle is exceptionally mundane--natural gas has been deployed 
previously in vehicles like the Honda Civic, and aftermarket CNG 
conversions have long been available on the market. Again, there is no 
critical hurdle to overcome with CNG powered vehicles, and there is 
little if any benefit to any such incentives. We strongly recommend 
that EPA eliminate all incentives for natural gas vehicles and instead 
ensure such vehicles are credited commensurate with their impact on the 
environment.'' CARB also commented that new compliance flexibilities 
for NGVs are not appropriate at this time.\3128\
---------------------------------------------------------------------------

    \3127\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3128\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    The Natural Gas Vehicles of America (NGVAmerica) commented that 
there is no incentive under existing EPA and NHTSA regulations for an 
automaker to sell vehicles equipped to be converted to operate on 
natural gas (so-called ``gaseous-prep vehicles''), even though selling 
such vehicles often results in the increased availability of 
alternative fuel vehicles. Today, most alternative fuel conversions are 
performed on newly manufactured gaseous-prep vehicles or vehicles that 
have been equipped by the original equipment manufacturers with 
hardened valves, valve seats, pistons, and piston rings. As an example, 
most of Ford's commercial truck line-up is available as gaseous-prep, 
and many such vehicles are converted to natural gas or propane by 
qualified vehicle manufacturers. Converting these vehicles, producing 
an assembly-line gaseous-prep vehicle, and sharing diagnostic 
information are critical to ensuring that aftermarket conversions 
perform well in-use and do not degrade the vehicle's emission control 
equipment. Given the complexity of today's automobiles, it is virtually 
impossible to legally convert new vehicles without this level of 
cooperation from vehicle manufacturers.
    NGVAmerica further commented that providing a regulatory incentive 
for automakers to sell these vehicles would expand the availability of 
gaseous-prep vehicles and increase consumer choice for alternative fuel 
vehicles. EPA, therefore, should provide a credit for selling such 
vehicles if the automaker can verify that the vehicles were 
subsequently upfitted or converted using an EPA certified alternative 
fuel system. Given the significant cost associated with certifying 
vehicles and installing natural gas tanks, there is very little 
likelihood that such an incentive would be abused by automakers. As 
with credits for original equipment manufactured vehicles, the utility 
factor for these vehicles would be based on the range of the vehicle 
when operating on natural gas. In this way, vehicles with larger range 
would earn more credit and vehicles with reduced range would earn less 
credit.
    Regarding comments that EPA should provide additional credits to 
auto manufacturers for the potential use of RNG due to upstream 
benefits associated with the production of RNG by applying a 0.15 
factor, EPA disagrees because auto manufacturers would not be required 
to ensure such fuels are used in the vehicles they produce over the 
life of those vehicles. Commenters provided a rationale for why they 
believe all NGVs produced in the future will be fueled with RNG, but 
EPA believes there is no assurance that this would be the case. If 
fossil fuel-based natural gas is used in the vehicles, the 
environmental benefits asserted by the commenters would not exist and 
the substantial vehicle incentives recommended by the commenters would 
result in a loss of environmental benefits. EPA does not believe it is 
appropriate to attribute most or all of the potential benefits of the 
production and use of RNG to the vehicle manufacturer. EPA's Renewable 
Fuel Standards (RFS) already appropriately credit RNG use as compared 
to fossil fuel-based natural gas. The RFS program provides a 
substantial incentive for RNG production, and those incentives may lead 
to even lower fuel pricing and greater demand for RNG as vehicle fuel, 
and for NGVs in the future. The RFS program also can provide incentives 
for liquid cellulosic fuels, advanced bio-diesel, and other types of 
renewable transportation fuels. Consistent with EPA's decision not to 
include upstream emissions associated with electricity use for EVs and 
PHEVs discussed above, EPA believes it is appropriate at this time to 
maintain the focus of the light-duty vehicle GHG standards on the 
capabilities of the vehicle to control emissions, and not rely on 
lifecycle fuel characteristics as a basis for developing specific 
vehicle incentives, particularly where those fuels are already 
incentivized by the RFS program.
    After considering comments regarding incentive multipliers for NGVs 
and the current lack of light-duty NGV offerings by OEMs in the market, 
EPA has decided to include a multiplier incentive of 2.0 for MY 2022-
2026 dedicated and dual-fuel NGVs. This multiplier will go into effect 
when the previously established multipliers expire, thus extending the 
mulipler for NGVs for 5 years beyond those previously established for 
NGVs. While other alternative fuel vehicles that were provided 
multiplier incentives are increasingly available in the light-duty 
marketplace, no OEM is currently offering light-duty NGVs. Since Honda 
ended production of the CNG version of the Honda Civic at the end of MY 
2015, there have been no OEM NGV offerings available to consumers. EPA 
continues to believe that NGVs could be an important part of the 
overall light-duty vehicle fleet mix, and such offerings would enhance 
the diversity of potentially cleaner alternative fueled vehicles 
available to consumers.\3129\ EPA believes it is appropriate to extend 
the availability of a production multiplier through MY 2026 for both 
dual-fuel and dedicated NGVs to potentially help spur their re-
introduction by OEMs in the light-duty vehicle market.
---------------------------------------------------------------------------

    \3129\ The CNG Honda Civic had approximately 20 percent lower 
CO2 than the gasoline Civic in MY 2015.
---------------------------------------------------------------------------

    EPA also received comments on the application of the regulatory 
utility factor. For dual-fuel vehicles, emissions are measured on both 
fuels (e.g., gasoline and natural gas) and weighted using a factor 
referred to in the regulations as a utility factor. To use a utility 
factor for natural gas greater than 0.5, a dual-fuel NGV must meet 
design criteria requiring the vehicle to have a natural gas to gasoline 
driving range of 2:1. The vehicle must also preferentially operate on 
natural gas until the natural gas tank is empty. EPA adopted these 
design criteria as part of the 2012 final rule to help ensure vehicles 
using a utility factor of higher than 0.5 would likely be fueled with 
and use natural gas most of the time on the road. At that time, EPA was 
concerned that natural gas refueling may be much more inconvenient for 
drivers relative to electric charging for PHEVs due to a lack of CNG 
refueling stations (or home refueling, compared to the availability of 
home chargers for many PHEVs) and, therefore, dual-fuel vehicles with 
limited driving range on natural gas would likely frequently operate on 
gasoline.
    EPA received comments regarding the design criteria. Ingevity 
commented that it has developed a low-pressure (900 psi) adsorbed 
natural gas (ANG) fuel storage technology that allows vehicles to be 
refueled using an affordable and reliable low-pressure natural gas 
fueling

[[Page 25212]]

appliance.\3130\ Ingevity commented that ANG will allow for a 
distributed refueling network at users' homes and businesses, just like 
electrical recharging equipment has been installed for PHEVs over the 
last several years. Ingevity commented that the design criteria for 
dual-fuel NGVs that were established in the MYs 2017-2025 final rule 
``make it impossible to reasonably and affordably manufacture a dual-
fuel NGV that can fully utilize the utility factor (UF) approach for 
determining fuel economy and [CO2] emissions.'' Ingevity 
recommended that the design criteria for dual-fuel NGVs be removed and 
that the utility factor be based only on the range of the NGV on 
natural gas, equivalent to the treatment of PHEVs. MECA submitted 
similar comments regarding ANG technology.\3131\
---------------------------------------------------------------------------

    \3130\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666.
    \3131\ See MECA, Detailed Comments, NHTSA-2018-0067-11999.
---------------------------------------------------------------------------

    Ariel and VNG also commented that design criteria imposed on dual-
fuel NGVs add unnecessary costs and complexity, and currently are 
arbitrarily applied only to dual-fuel NGVs, and not to their dual-fuel 
hybrid counterparts.\3132\ NACS, SIGMA, and NATSO also recommended that 
EPA remove eligibility requirements associated with the utility 
factor.\3133\
---------------------------------------------------------------------------

    \3132\ Joint Submission from Ariel Corp. and VNG, Detailed 
Comments, NHTSA-2018-0067-7573.
    \3133\ Joint submission on behalf of NACS and SIGMA, Detailed 
Comments, EPA-HQ-OAR-2018-0283-5824; NATSO, Detailed Comment, EPA-
HQ-OAR-2018-0283-5484.
---------------------------------------------------------------------------

    After considering the comments, EPA is removing the design criteria 
from the regulations and thereby allowing higher utility factors to be 
used for dual-fuel natural gas vehicles based solely on driving range 
on natural gas, as is the case for PHEVs. The utility factor represents 
a reasonable way of weighting the emissions of a dual-fuel vehicle on 
each fuel to derive a single emissions value when including the dual-
fuel vehicles in a manufacturer's fleet average compliance 
determination. Ideally, the utility factor would match the use of each 
fuel in real-world vehicle operation. The utility factor is not meant 
to incentivize the adoption of a particular technology, so it differs 
fundamentally from incentives such as multipliers. With the development 
of low-pressure natural gas vehicle fueling system technology since the 
2012 final rule, EPA's concerns regarding limited fueling 
infrastructure that led the agency to adopt the design criteria in the 
2012 rule are significantly diminished. EPA believes that low-pressure 
fueling is a new advancement that offers the potential for more 
convenient refueling for individuals or businesses similar to that for 
PHEVs. EPA expects owners of dual-fuel CNG vehicles preferentially to 
seek to refuel and operate on CNG fuel as much as possible, both 
because the owner would have to pay a higher vehicle price for the 
dual-fuel capability, and because CNG fuel is considerably cheaper than 
gasoline. With the opportunity for relatively low-cost on-site 
refueling at homes or businesses, EPA expects such vehicles to be 
refueled with natural gas similar to how people refuel PHEVs. Vehicle 
purchasers that choose high pressure vehicle systems over low pressure 
systems would likely do so only if they have ready access to a high 
pressure refueling system, for example, at a fleet's central fueling 
location. Removing the design criteria for dual-fuel natural gas 
vehicles also addresses the concerns of some commenters regarding the 
differing treatment of PHEVs and dual-fuel NGVs.
    EPA believes that with the advancement of technology offering the 
potential for more flexible refueling of NGVs, removing the design 
criteria is a reasonable change to the regulations. This regulatory 
change will apply starting with MY 2021. MY 2021 will provide 
sufficient time for orderly implementation and EPA is not aware of any 
dual-fuel NGVs emissions certified for MYs 2019-2020 that would 
otherwise be affected if this change were to be implemented sooner.
    EPA received comments that vehicle conversions and ``gaseous-prep'' 
vehicles should be eligible for credits. In response to comments on 
vehicle conversions, alternative fuel converters are not required to 
meet fleet average standards but instead may comply with 40 CFR part 85 
subpart F regulations providing a tampering exemption. Fleet average 
standards are generally not appropriate for fuel conversion 
manufacturers because the ``fleet'' of vehicles to which a conversion 
system may be applied has already been accounted for under the OEM's 
fleet average standard. Alternative fuel converters are not 
manufacturing new vehicles, but are converting existing vehicles that 
have already been certified by the OEM. CO2 credits are 
available to OEMs based on fleet emissions performance compared to the 
fleet average standards and therefore conversions are not eligible for 
these credits. EPA did not propose to change and is not changing the 
exemption process promulgated in 40 CFR part 85 subpart F. Because fuel 
conversions are not required to meet the fleet average standards, 
credits generated under those standards are not available. Regarding 
gaseous-prep vehicles, these vehicles are not NGVs at initial sale and 
therefore are not eligible for NGV incentives. Instead, they are 
included in the OEM's fleet as gasoline-only vehicles. EPA disagrees 
with the commenters that such vehicles should be eligible for NGV 
incentives at time of initial sale if the vehicle is later converted to 
natural gas since the OEM does not measure the emissions of the vehicle 
on natural gas at time of certification and is not responsible for the 
emissions performance of the vehicle on natural gas over the life of 
the vehicle.

C. NHTSA Compliance and Enforcement

1. Overview of the NHTSA Compliance Process
    Consumer choice drives the mixture of automobiles on the road. 
Manufacturers largely produce a mixture of vehicles to meet consumer 
demand and address compliance with CAFE standards though the 
application of fuel economy improving technologies to those vehicles, 
and by using compliance flexibilities and incentives that are available 
in the CAFE program. As discussed earlier in this notice, each vehicle 
manufacturer is subject to separate CAFE standards for passenger cars 
and light trucks, and for the passenger car standards, a manufacturer's 
domestically-manufactured and imported passenger car fleets are 
required to comply separately.\3134\ Additionally, domestically-
manufactured passenger cars are subject to a statutory minimum 
standard.\3135\ CAFE program flexibilities are largely provided for in 
statute. Credits for air conditioning efficiency, off-cycle, and pickup 
truck advanced technologies are not expressly specified by CAFE 
statute, but are ``implemented consistent with EPCA's provisions 
regarding calculation of fuel economy'' as discussed in section C.2 
below.
---------------------------------------------------------------------------

    \3134\ 49 U.S.C. 32904(b).
    \3135\ 49 U.S.C. 32902(b)(4).
---------------------------------------------------------------------------

    Compliance with the CAFE program begins with manufacturers 
submitting required reports to NHTSA in advance and during the model 
year that contain information, specifications, data, and projections 
about their fleets.\3136\ Manufacturers report early product 
projections to NHTSA describing their efforts to comply with CAFE 
standards per EPCA's reporting requirements.\3137\ Manufacturers' early 
projections are required to identify any of the

[[Page 25213]]

flexibilities and incentives manufacturers plan to use for air-
conditioning (A/C) efficiency, off-cycle and, through MY 2021, full-
size pickup truck advanced technologies. EPA consults with NHTSA when 
reviewing and considering manufacturers' requests for fuel consumption 
improvement values for A/C and off-cycle technologies that improve fuel 
economy. NHTSA evaluates and monitors the performance of the industry 
using the information provided. NHTSA also audits manufacturers' 
projected data for conformance and verifies vehicle design data through 
testing to ensure manufacturers are complying as projected. After the 
model year ends, manufacturers submit final reports to EPA, including 
final information on all the flexibilities and incentives allowed or 
approved for the given model year.\3138\ EPA then calculates the fuel 
economy level of each fleet produced by each manufacturer, and 
transmits that information to NHTSA.\3139\
---------------------------------------------------------------------------

    \3136\ 49 U.S.C. 32907(a); 49 CFR 537.7.
    \3137\ 49 U.S.C. 32907(a).
    \3138\ For example, alternative fueled vehicles get special 
calculations under EPCA (49 U.S.C. 32905-06), and fuel economy 
levels can also be adjusted to reflect air conditioning efficiency 
and ``off-cycle'' improvements, as discussed below.
    \3139\ 49 U.S.C. 32904(c)-(e). EPCA granted EPA authority to 
establish fuel economy testing and calculation procedures; EPA uses 
a two-year early certification process to qualify manufacturers to 
start selling vehicles, coordinates manufacturer testing throughout 
the model year, and validates manufacturer-submitted final test 
results after the close of the model year.
---------------------------------------------------------------------------

    NHTSA notes that some manufacturers have submitted and/or 
resubmitted requests for A/C and off-cycle benefits after EPA final 
reports are completed or nearly completed and, in those cases, such 
submissions are causing considerable delays in EPA's ability to 
finalize CAFE reports. Late and revised submissions can place 
significant burdens on the government in order to reassess a 
manufacturer's CAFE performances and standards and can also cause 
significant impacts on previous compliance model years. In the 
following sections, EPA and NHTSA are incorporating regulatory 
modifications or providing guidance to help manufacturers expedite 
approvals and to facilitate the governments processing of the 
flexibilities and incentives.
    NHTSA determines each manufacturer's obligation to comply with 
applicable model year's CAFE standards and notifies the manufacturer if 
any of its fleet performances fall below standards. Manufacturers must 
submit plans detailing the compliance flexibilities to be used to 
resolve any possible noncompliances or may pay civil penalties to 
address any deficits for falling below standards. NHTSA periodically 
releases data and reports to the public through its CAFE Public 
Information Center (PIC) based on information in the EPA final reports 
for the given compliance model year, and based on the projections 
manufacturers provide to NHTSA for the next two model years.\3140\
---------------------------------------------------------------------------

    \3140\ NHTSA CAFE Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
---------------------------------------------------------------------------

2. NHTSA's CAFE Program Compliance
    EPCA and EISA specify several flexibilities and incentives that are 
available to help manufacturers comply with CAFE standards. Some 
flexibilities are defined, and sometimes limited by statute--for 
example, while Congress allowed manufacturers to transfer credits 
earned for over-compliance from their car fleet to their truck fleet 
and vice versa, Congress also limited the amount by which manufacturers 
could increase their CAFE levels using those transfers.\3141\ 
Consistent with the limits Congress placed on certain statutory 
flexibilities and incentives, NHTSA crafted and implements the credit 
transfer and trading regulations authorized by EISA to help ensure that 
total fuel savings are preserved when manufacturers exercise statutory 
compliance flexibilities.
---------------------------------------------------------------------------

    \3141\ See 49 U.S.C. 32903(g).
---------------------------------------------------------------------------

    NHTSA and EPA have previously developed other compliance 
flexibilities and incentives for the CAFE program consistent with the 
statutory provisions regarding EPA's calculation of manufacturers' fuel 
economy levels. As discussed previously, NHTSA finalized in the 2012 
final rule, for MYs 2017 and later, an approach for manufacturers' 
``credits'' under EPA's program to be applied as fuel economy 
``adjustments'' or ``improvement values'' under NHTSA's program for: 
(1) Technologies that cannot be measured or cannot be fully measured on 
the 2-cycle test procedure, i.e., ``off-cycle'' technologies; and (2) 
A/C efficiency improvements that also improve fuel economy but cannot 
be measured on the 2-cycle test procedure. Additionally, both agencies' 
programs give manufacturers compliance incentives through MY 2021 for 
utilizing specified technologies on pickup trucks, such as pickup truck 
hybridization.
    The following sections outline how NHTSA determines whether 
manufacturers are in compliance with the CAFE standards for each model 
year, and how manufacturers may use compliance flexibilities, or 
address noncompliance by paying civil penalties. As addressed above, 
some compliance flexibilities are expressly prescribed in statute and 
some are implemented consistent with EPCA's provisions regarding 
calculation of fuel economy. NHTSA proposed new language updating and 
clarifying existing regulatory text in this area as part of the NPRM. 
NHTSA also sought comments in the NPRM on these changes, as well as on 
the general efficacy of these flexibilities in the fuel economy and 
CO2 programs.
    Moreover, the following sections explain how manufacturers submit 
data and information to the agency. As part of the NPRM, NHTSA proposed 
to implement a new standardized template for manufacturers to use to 
submit CAFE data to the agency, as well as a standardized template for 
reporting credit transactions. Additionally, NHTSA proposed adding 
requirements that specify the precision of the fuel savings adjustment 
factor in 49 CFR 536.4. These new requirements are intended to 
streamline reporting and data collection from manufacturers, in 
addition to helping the agency use the best available data to inform 
CAFE program decision makers. The comments received to these proposals 
are included in Section IX.C.2.a)(2)(d) along with NHTSA's responses to 
the comments and final resolutions established in the final rule.
    NHTSA also sought comments on removing or modifying certain CAFE 
program flexibilities. The comments received and NHTSA's responses to 
those comments are discussed below.
a) How does NHTSA determine compliance?
(1) Manufacturers Submit Data to NHTSA and EPA and the Agencies 
Validate Results
    EPCA, as amended by EISA, requires a manufacturer to submit reports 
to the Secretary of Transportation explaining whether the manufacturer 
will comply with an applicable CAFE standard for the model year for 
which the report is made; the actions a manufacturer has taken or 
intends to take to comply with the standard; and other information the 
Secretary requires by regulation.\3142\ A manufacturer must submit a 
report containing the above information during the 30-day period before 
the beginning of each model year, and during the 30-day period 
beginning the 180th day of the model year.\3143\ When a manufacturer 
determines it is unlikely to comply with a CAFE standard, the 
manufacturer must report additional

[[Page 25214]]

actions it intends to take to comply and include a statement about 
whether those actions are sufficient to ensure compliance.\3144\
---------------------------------------------------------------------------

    \3142\ 49 U.S.C. 32907(a).
    \3143\ Id.
    \3144\ Id.
---------------------------------------------------------------------------

    To implement these reporting requirements, NHTSA issued 49 CFR part 
537, ``Automotive Fuel Economy Reports,'' which specifies three types 
of CAFE reports that manufacturers must submit. A manufacturer must 
first submit a pre-model year (PMY) report containing the 
manufacturer's projected compliance information for that upcoming model 
year. By regulation, the PMY report must be submitted in December of 
the calendar year prior to the corresponding model year.\3145\ 
Manufacturers must then submit a mid-model year (MMY) report containing 
updated information from manufacturers based upon actual and projected 
information known midway through the model year. By regulation, the MMY 
report must be submitted by the end of July for the applicable model 
year.\3146\ Finally, manufacturers must submit a supplementary report 
to supplement or correct previously submitted information, as specified 
in NHTSA's regulation.\3147\
---------------------------------------------------------------------------

    \3145\ 49 CFR 537.5(b).
    \3146\ Id.
    \3147\ 49 CFR 537.8.
---------------------------------------------------------------------------

    If a manufacturer wishes to request confidential treatment for a 
CAFE report, it must submit both a confidential and redacted version of 
the report to NHTSA. CAFE reports submitted to NHTSA contain estimated 
sales production information, which may be protected as confidential 
until the termination of the production period for that model 
year.\3148\ NHTSA temporarily protects each manufacturer's competitive 
sales production strategies, but does not permanently exclude sales 
production information from public disclosure. Sales production volumes 
are part of the information NHTSA routinely makes publicly available 
through the CAFE PIC.
---------------------------------------------------------------------------

    \3148\ 49 CFR part 512, appx. B(2).
---------------------------------------------------------------------------

    The manufacturer reports provide information on light-duty 
automobiles such as projected and actual fuel economy standards, fuel 
economy performance values, and production volumes, as well as 
information on vehicle design features (e.g., engine displacement and 
transmission class) and other vehicle attribute characteristics (e.g., 
track width, wheelbase, and other off-road features for light trucks). 
Beginning with MY 2017, to obtain credit for fuel economy improvement 
values attributable to additional technologies, manufacturers must also 
provide information regarding A/C systems with improved efficiency, 
off-cycle technologies (e.g., stop-start systems, high-efficiency 
lighting, active engine warm-up), and full-size pickup trucks with 
hybrid technologies or with emissions/fuel economy performance that is 
better than footprint-based targets by specified amounts. This includes 
identifying the makes and model types equipped with each technology, 
the compliance category those vehicles belong to, and the associated 
fuel economy improvement value for each technology.\3149\ In some 
cases, NHTSA may require manufacturers to provide supplementary 
information to justify or explain the benefits of these technologies 
and their impact on fuel consumption or to evaluate the safety 
implication of the technologies. These details are necessary to 
facilitate NHTSA's technical analyses and to ensure the agency can 
perform enforcement audits as appropriate.
---------------------------------------------------------------------------

    \3149\ NHTSA collects model type information based upon the EPA 
definition for ``model type'' in 40 CFR 600.002.
---------------------------------------------------------------------------

    NHTSA uses manufacturer-submitted PMY, MMY, and supplementary 
reports to assist in auditing manufacturer compliance data and 
identifying potential compliance issues as early as possible. 
Additionally, as part of its footprint validation program, NHTSA 
conducts vehicle testing throughout the model year to confirm the 
accuracy of the track width and wheelbase measurements submitted in the 
reports.\3150\ These tests help the agency better understand how 
manufacturers may adjust vehicle characteristics to change a vehicle's 
footprint measurement, and ultimately its fuel economy target. NHTSA 
also includes a summary of manufacturers' PMY and MMY data in an annual 
fuel economy performance report made publicly available on its PIC.
---------------------------------------------------------------------------

    \3150\ U.S. Department of Transportation, NHTSA, Laboratory Test 
Procedure for 49 CFR part 537, Automobile Fuel Economy Attribute 
Measurements (Mar. 30, 2009), available at http://www.nhtsa.gov/DOT/NHTSA/Vehicle%20Safety/Test%20Procedures/Associated%20Files/TP-537-01.pdf.
---------------------------------------------------------------------------

    NHTSA uses EPA-verified final-model year (FMY) data to evaluate 
manufacturers' compliance with CAFE program requirements, and draws 
conclusions about the performance of the industry. After manufacturers 
submit their FMY data, EPA verifies the information, accounting for 
NHTSA and EPA testing, and subsequently forwards the final verified 
data to NHTSA.
(2) Changes to CAFE Reporting Requirements Made by This Final Rule
    NHTSA proposed changes to its CAFE reporting requirements with the 
intent of streamlining data collection and reporting for manufacturers 
while helping the agency obtain the best available data to inform CAFE 
program decision-makers. The agency developed two new standardized 
reporting templates for manufacturers and proposed to start using the 
templates beginning in the 2019 compliance model year. In the NPRM, 
NHTSA sought comments on the templates. NHTSA's responses to the 
comments received and the changes to the templates for the final rule 
are presented below.
(a) Standardized CAFE Reporting Template
    When NHTSA received and reviewed manufacturers' projection reports 
for MYs 2013--2015, the agency observed that most did not conform to 
the requirements specified in 49 CFR part 537. For example, NHTSA 
identified several instances where manufacturers' CAFE reports included 
a ``yes'' or ``no'' response to a request for a vehicle's numerical 
ground clearance values. In a 2015 notice of proposed rulemaking, NHTSA 
proposed to amend 49 CFR part 537 to require a new data format for 
manufacturers' light-duty vehicle CAFE projection reports.\3151\ In 
response to the proposal, some manufacturers commented that the 
previous changes in reporting requirements generated confusion and led 
to reporting errors. NHTSA recognized that the modification to the base 
tire definition in the 2012 final rule for MYs 2017 and later seemed to 
make some manufacturers uncertain about what footprint data was 
required in the reports.\3152\ Specifically, certain manufacturers did 
not understand that the modified base tire definition required them to 
provide estimated attribute-based target standards for each unique 
model type/footprint combination beginning with MY 2013. NHTSA 
discovered cases where manufacturers only provided target or vehicle 
data for certified vehicle configurations, and did not report 
information for each of the unique model type/footprint combinations 
for their available production vehicles in the market. However, NHTSA 
did not adopt the proposed data format from the 2015 proposed rule 
after receiving

[[Page 25215]]

adverse comments from manufacturers.\3153\
---------------------------------------------------------------------------

    \3151\ 80 FR 40540 (Jul. 13, 2015).
    \3152\ 49 CFR 523.2.
    \3153\ 81 FR 73958 (Oct. 25, 2016).
---------------------------------------------------------------------------

    Since the issuance of the final rule in 2016, NHTSA has continued 
to receive projection reports that contain inaccurate and/or missing 
data. These noncompliant reports impede NHTSA's ability to audit 
manufacturer compliance data, identify potential compliance issues, and 
analyze industry trends. Problems with inaccurate or missing data has 
become an even greater issue for manufacturers reporting on the new MY 
2017 incentives for efficient A/C systems, off-cycle technologies, and 
full-size pickup trucks with hybrid technologies/improved exhaust 
emission performance.\3154\ These incentives are explained in Section 
IX.C.2.c). Manufacturers seeking to take advantage of these new 
benefits must provide information at the model-type level; however, 
many manufacturers did not submit the required information in their PMY 
reports for MYs 2017, 2018, and 2019. This caused NHTSA's Office of 
Enforcement to send letters reminding manufacturers of their obligation 
to submit accurate and complete CAFE reports. NHTSA will continue to 
monitor the accuracy, completeness, and timeliness of manufacturers' 
CAFE reports and may take additional action as appropriate.
---------------------------------------------------------------------------

    \3154\ NHTSA allows manufacturers to use these flexibilities and 
incentives for complying with standards starting in MY 2017; the 
FCIV for full-size pickup trucks with hybrid technologies/improved 
exhaust emission performance applies only through MY 2021, as 
discussed further below.
---------------------------------------------------------------------------

    In the NPRM, NHTSA proposed a new standardized template for 
reporting PMY and MMY information, as specified in 49 CFR 537.7(b) and 
(c), as well as supplementary information required by 49 CFR 537.8. The 
template allows manufacturers to build out the required confidential 
versions of CAFE reports specified in 49 CFR part 537 and to produce 
automatically the required non-confidential versions by clicking a 
button within the template. While NHTSA recognizes that modifications 
to the reporting requirements may initially be a slight inconvenience 
to manufacturers, the number of noncompliant reports the agency 
continues to receive justifies development of a uniform reporting 
method to help ensure compliance with CAFE regulations. Adopting a 
standardized template will assist manufacturers in providing the agency 
with all necessary data, thereby helping manufacturers to ensure they 
are complying with CAFE regulations. The template organizes the 
required data in a manner consistent with NHTSA and EPA regulations and 
simplifies the reporting process by incorporating standardized 
responses consistent with those provided to EPA. The template collects 
the relevant data, calculates intermediate and final values in 
accordance with EPA and NHTSA methodologies, and aggregates all the 
final values required by NHTSA regulations in a single summary 
worksheet. Thus, NHTSA believes that the standardized templates will 
benefit both the agency and manufacturers by helping to avoid reporting 
errors, such as data omissions and miscalculations, and will ultimately 
simplify and streamline reporting.
    NHTSA proposed to require that manufacturers use the standardized 
template for all PMY, MMY, and supplementary CAFE reports. NHTSA 
observed that a significant number of manufacturers submit their MMY 
reports as updated PMY reports--using the same amount of information, 
despite fewer data requirements. To conform with this method, NHTSA 
designed the template based on one standardized format that uses the 
same data requirements for all CAFE reports. This approach will further 
simplify CAFE projection reporting for manufacturers. The template 
contains a few additional data fields for certain vehicle 
characteristics; however, the inclusion of model type indexes will 
limit the number of required entries by populating a number of pre-
entered data fields based on one value.
    The standardized template will also allow NHTSA to modify its 
existing compliance database to accept and import uniform data and 
automatically aggregate manufacturers' data. This will allow NHTSA to 
execute its regulatory obligations more efficiently and effectively. 
Overall, the template will help to ensure compliance with data 
requirements under EPCA/EISA and drastically reduce the industry and 
government's burden for reporting in accordance with the Paperwork 
Reduction Act.\3155\ NHTSA made the template available through its 
docket as well as its PIC, and sought comment on the regulatory changes 
to the reporting process.
---------------------------------------------------------------------------

    \3155\ 44 U.S.C. 3501 et seq.
---------------------------------------------------------------------------

    Comments on the template were received from the Auto Alliance, 
Global Automakers, Ford, Mercedes-Benz, Toyota, Volvo and Volkswagen. 
The Auto Alliance, Toyota, and Volkswagen opposed adopting the proposed 
template; however, Global Automakers agreed with the appropriateness of 
a standardized template that combines credit trading information with a 
data reporting template.\3156\ Global Automakers also made two 
recommendations: (1) Combine EPA's AB&T template with NHTSA's CAFE 
Projections Reporting Template to streamline reporting and reduce 
burden; and (2) add an FMY report requirement as an update to the MMY 
report submission.\3157\
---------------------------------------------------------------------------

    \3156\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Toyota, Detailed Comments, NHTSA-2018-0067-12150; Volkswagen, 
Detailed Comments, NHTSA-2017-0069-0583.
    \3157\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
---------------------------------------------------------------------------

    Mercedes-Benz, Ford, and Volkswagen commented about data fields 
they believed were outdated, or not relevant to fuel economy testing or 
projecting fuel economy performance.\3158\ Mercedes-Benz stated that 
some required data fields are not currently collected as a part of the 
fuel economy testing process, and their capture would require 
additional burden.\3159\ Mercedes-Benz believes those data fields 
should be an optional requirement. Additionally, Mercedes-Benz 
recommended that NHTSA omit certain data fields, and stated that it 
would be helpful if NHTSA clarified its intention for the information 
in others.\3160\ The specific data fields mentioned by Mercedes-Benz 
are in Table IX-6. Ford stated that many of the data fields are 
outdated, have no bearing on compliance assessments, and are misaligned 
with the current reporting structure, which is dictated by model type 
index.\3161\ Similarly, Volkswagen stated that the proposed reporting 
template is populated with many fields that do not immediately appear 
relevant to projecting CAFE performance, align with the existing 
requirements in 49 CFR 537.7, or seem relevant in the space of 
automotive technology.\3162\
---------------------------------------------------------------------------

    \3158\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
0283-4182; Ford, Detailed Comments, NHTSA-2018-0067-11928; 
Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
    \3159\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
0283-4182.
    \3160\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
0283-4182.
    \3161\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
    \3162\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.

---------------------------------------------------------------------------

[[Page 25216]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.756

    The Auto Alliance and Mercedes-Benz noted the differences in how 
NHTSA and EPA request data on A/C efficiency and off-cycle 
technologies. Mercedes-Benz highlighted the difficulty in predicting 
the projected sales production of the technologies, and the Auto 
Alliance cautioned that the number of reporting entries would increase 
by a factor of ten or more.\3163\ The Auto Alliance stated its belief 
that the change in reporting requirements would cost its members more 
than $1 million in information technology changes and that the changes 
could not be completed prior to MY 2021.\3164\ Likewise, Ford contended 
that an implementation date for MY 2019 is aggressive and does not 
provide manufacturers with adequate lead time.\3165\
---------------------------------------------------------------------------

    \3163\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-0283-4182.
    \3164\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3165\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    The Auto Alliance emphasized that the templates lack common 
reporting standardization with submissions to EPA.\3166\ The Auto 
Alliance, Global Automakers, Toyota, and Volvo all requested that NHTSA 
and EPA accept a single, common reporting format to satisfy reporting 
for both agencies.\3167\ Mercedes-Benz and Volkswagen requested 
stakeholder workshops to review the template with agency staff, with 
the former recommending that NHTSA host the workshops in partnership 
with EPA.\3168\
---------------------------------------------------------------------------

    \3166\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3167\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Global Automakers, Detailed Comments, NHTSA-2018-0067-12032; Toyota, 
Detailed Comments, NHTSA-2018-0067-12150; Volvo, Detailed Comments, 
NHTSA-2018-0067-12036.
    \3168\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
0283-4182; Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    Ford requests that NHTSA re-examine the proposed required 
submission methods and reconsider current electronic submission 
methods.\3169\ Ford expressed concern about the efficiency and security 
issues involved in submitting data on a CD through the mail containing 
confidential business information.\3170\ Ford identified what it 
believes are better available avenues for submission, such as secured 
email or online portals like EPA's Central Data Exchange.\3171\
---------------------------------------------------------------------------

    \3169\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
    \3170\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
    \3171\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    NHTSA disagrees with many of the manufacturers' assertions. 
Differences in EPA and NHTSA regulations prevent establishing a single 
reporting format for CAFE purposes. For example, EPA only needs early 
model year information for manufacturers' applications for 
certification required under 40 CFR 86.1843-01. Manufacturers submit a 
single application with extensive details for each certified vehicle 
within a test group (i.e., the certified vehicle represents all the 
vehicles within the test group with similar technologies and 
performance characteristics). In comparison, NHTSA's required early 
model year information is far less detailed and is aggregated for model 
types and compliance categories. However, NHTSA and EPA already share 
all the relevant CAFE FMY information pursuant to an interagency 
agreement. This arrangement not only benefits manufacturers but also 
reduces the burden on the Federal government. Since much of the 
required data in NHTSA's projections template is already contained in 
EPA final reports, manufacturers would not be required to generate 
additional information but simply to provide estimates along the way to 
finalizing the data. NHTSA plans to release a data matrix that maps 
data elements between the CAFE template and the EPA final CAFE reports. 
NHTSA will notify the public when the matrix will be available on its 
website. Consequently, there is no need to create an additional final 
report as an updated version of NHTSA's MMY report, as suggested by 
Global Automakers. Once NHTSA configures its CAFE database to accept 
the reporting template via file upload, the agency will be able to use 
the model type index data field to connect data values from the 
template to corresponding values in EPA's final CAFE report. 
Manufacturers should note that CAFE reports are estimated projections 
of the EPA final CAFE compliance data. Contrary to Mercedes concerns 
about the difficulty in predicting the projected sales production of 
the technologies, NHTSA only expects manufacturers to provide the most 
up-to-date information available 30 days before a report is required to 
be submitted to the

[[Page 25217]]

Administrator as specified in 49 CFR part 537.5(d). While manufacturer 
PMY reports may be limited in certain instances (excluding vehicles 
already in sales distribution), the MMY reports should be more 
inclusive and closer to the final values reported to EPA. Manufacturers 
should also be submitting supplementary reports to NHTSA if they 
believe there will be significant differences between CAFE MMY reports 
and the EPA final reports.
    Commenters also stated that the A/C and off-cycle information 
reported in the NHTSA template is inconsistent with the EPA EV-
CIS.\3172\ NHTSA notes that the inconsistency between the agencies is 
intentional and necessary. NHTSA's off-cycle and A/C information must 
be collected in greater detail than that reported to the EPA EV-CIS. 
NHTSA collects detailed information on A/C and off-cycle technologies 
for determining penetration rates of specific technologies in the 
market, as well as analyzing the types of technologies as equipped on 
specific model types. In comparison, EPA aggregates the data for 
calculating credits, which allows for combining the benefits for all 
the technologies equipped on a model type. NHTSA also will use the 
detailed information for public disclosure and for auditing purposes. 
However, NHTSA acknowledges the Auto Alliance's concerns about the 
burden placed on the industry for providing more detailed data and 
therefore will not require manufacturers to start using the templates 
for reporting until MY 2023. NHTSA also agrees with Ford that it is 
important to consider the issues of security and efficiency with 
respect to the submission of confidential information to the agency, 
and the agency will consider possible changes to its procedures 
relating to the receipt and handling of confidential information to 
ensure streamlined, secure, and efficient submission of confidential 
information, including CAFE reports.\3173\
---------------------------------------------------------------------------

    \3172\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
    \3173\ See 49 CFR part 512, 537.5.
---------------------------------------------------------------------------

    Secondly, NHTSA agrees with Mercedes-Benz and Volkswagen that 
workshops will aid in implementing the templates by providing 
instruction on how to complete them. NHTSA plans to host a workshop for 
manufacturers to discuss the implementation process. NHTSA believes 
finalizing the template in this rulemaking is important to address 
continuing concerns with reporting noncompliance (i.e., missing, 
incomplete, or inaccurate submissions) with the existing provisions in 
Part 537. Ultimately, establishing the new templates and holding 
educational workshops will be more effective in achieving industry 
compliance than imposing penalties on a case-by-case basis for failure 
to comply with reporting provisions.
    Finally, NHTSA is also adopting changes to the proposed template in 
response to comments from Mercedes-Benz, Ford, and Volkswagen. NHTSA 
made changes to several of the data fields discussed by Mercedes-Benz. 
NHTSA does not agree with Mercedes-Benz's recommendation to omit the 
``Type of Overdrive'' or ``Type of Torque Converter'' data fields; 
however, the agency does believe the proposed data to be inserted into 
those fields may be too specific for CAFE purposes. Therefore, the 
agency is finalizing a requirement that manufacturers identify whether 
vehicles are equipped with overdrive or a torque converter by selecting 
``Yes'' or ``No'' from a dropdown list. The agency has also changed the 
``Calibration'' field to ``Other Calibration'' to clarify the data 
being requested, and changed the ``Auxiliary Emission Control Device'' 
in the ``Fuel Economy'' worksheets to a dropdown that allows users to 
select multiple emission control systems. NHTSA believes that adding 
dropdown lists in the template creates uniformity in the reported 
information and makes the information more relevant to current 
vehicles.
    The agency agrees with the essence of Volkswagen's assertion that 
some of the required data fields may no longer be as common on 
contemporary vehicles, and therefore, may not apply to all 
manufacturers. As suggested by Mercedes-Benz, NHTSA has decided to make 
the ``Catalyst Usage,'' ``Distributor Calibration,'' ``Choke 
Calibration,'' and ``Other Calibration'' data fields optional with a 
default value of ``N/A.'' NHTSA does not agree with Mercedes-Benz's 
recommendation that NHTSA provide a better understanding of its 
intention for the information in certain data fields. ``Electric 
Traction Motor, Motor Controller,'' ``Battery Configuration,'' 
``Electrical Charging System,'' and ``Energy Storage Device'' are the 
data fields that characterize the basic powerplant for electric 
vehicles. Basic Engine, along with Carline and Transmission Class, make 
up a model type for light-duty vehicles. Therefore, those five fields 
are used to group vehicles by model type in accordance with EPA 
regulations. Fuel economy performance is calculated by 
Subconfiguration, which is a subset of a model type. As such, those 
five data fields are an integral part of grouping vehicles for fuel 
economy testing purposes in accordance with EPA regulations. NHTSA also 
does not agree with Volkswagen's assertion that the template is 
populated with many fields that do not appear relevant to projecting 
CAFE performance. As previously mentioned, many of the data fields are 
used to arrange vehicles into groups for calculating fuel economy 
performance in accordance with 49 CFR 537.7.
    Furthermore, NHTSA has re-engineered the template in a few areas to 
include additional supporting data elements used in calculating other 
data fields required by Part 537. These fields may not directly align 
with the existing requirements in Part 537 but are necessary for 
validation purposes. For this reason, NHTSA is also finalizing its 
proposal in the NPRM to remove the optional provisions for reporting 
the data fields for determining the CAFE model type target standards, 
making the information mandatory in the template. Additional changes 
have been made to the template to improve fuel economy calculations. 
NHTSA edited the template to include the calculation procedure for 
alternative-fuel vehicles and corrected the test procedure adjustment 
(TPA) calculation to align the fleet average fuel economy calculation 
methodology with 40 CFR 600.510-12. Several expanded worksheets and 
functional features were also added to the template to improve the 
usability of the templates for manufacturers. These changes include 
modifications such as adding the estimated credits and a minimum 
domestic passenger shortfall calculator as the last fields to the 
``Summary'' worksheet. Other functional changes include protecting 
users from changing the formatting or data validation in each cell and 
allowing columns to be widened by users.
(b) Standardized Credit Documents
    A credit ``[t]rade'' is defined in 49 CFR 536.3 as ``the receipt by 
NHTSA of an instruction from a credit holder to place its credits in 
the account of another credit holder.'' \3174\ ``Traded credits are 
moved from one credit holder to the recipient credit holder within the 
same compliance category for which the credits were originally earned. 
If a credit has been traded to another credit holder and is 
subsequently traded back to the originating manufacturer, it will be 
deemed not to have been traded for compliance purposes.'' \3175\ NHTSA 
does not administer trade negotiations between manufacturers and when a

[[Page 25218]]

trade document is received the agreement must be issued jointly by the 
current credit holder and the receiving party.\3176\ NHTSA does not 
settle contractual or payment issues between trading manufacturers.
---------------------------------------------------------------------------

    \3174\ 49 CFR 536.3(b).
    \3175\ Id.
    \3176\ See 49 CFR 536.8(a).
---------------------------------------------------------------------------

    NHTSA created its CAFE database to maintain credit accounts for 
manufacturers and to track all credit transactions. A credit account 
consists of a balance of credits in each compliance category and 
vintage held by the holder. While maintaining accurate credit records 
is essential, it has become a challenging task for the agency given the 
recent increase in credit transactions. Manufacturers have requested 
that NHTSA approve trade or transfer requests not only in response to 
end-of-model year shortfalls, but also, during the model year, when 
purchasing credits to bank.
    To reduce the burden on all parties, encourage compliance, and 
facilitate quicker NHTSA credit transaction approval, the agency 
proposed in the NPRM to add a required template to standardize the 
information parties submit to NHTSA in reporting a credit transaction. 
Presently, manufacturers are inconsistent in submitting the information 
required by 49 CFR 536.8, creating difficulty for NHTSA in processing 
transactions. The template NHTSA proposed is a simple spreadsheet that 
trading parties fill out. When completed, parties will be able to click 
a button on the spreadsheet to generate a credit transaction summary 
and if applicable credit trade confirmation, the latter of which shall 
be signed by both trading entities. The credit trade confirmation 
serves as an acknowledgement that the parties have agreed to trade 
credits. The completed credit trade summary and a PDF copy of the 
signed trade confirmation must be submitted to NHTSA. Using the 
template simplifies CAFE compliance aspects of the credit trading 
process, and helps to ensure that trading parties follow the 
requirements for a credit transaction in 49 CFR 536.8(a).\3177\
---------------------------------------------------------------------------

    \3177\ Submitting a properly completed template and accompanying 
transaction letter will satisfy the trading requirements in 49 CFR 
part 536.
---------------------------------------------------------------------------

    Additionally, the credit trade confirmation includes an 
acknowledgement of the ``error or fraud'' provisions in 49 CFR 
536.8(f)-(g), and the finality provision of 49 CFR 536.8(g). NHTSA 
sought comment on this approach, as well as on any changes to the 
template that may be necessary to facilitate manufacturer credit 
transaction requests. The agency uploaded the proposed template to the 
NHTSA's docket and the CAFE PIC site for manufacturers to download and 
review.
    Only Global Automakers commented on the proposed credit transaction 
template, and Global Automakers supported adopting a uniform template. 
Global Automakers stated that, in theory, it agrees that a standardized 
template with credit trading information is appropriate, and a similar 
template is already in use for these types of reporting requirements by 
its members that could be integrated into the end of the year EPA final 
report. Global Automakers believes the use of similar templates have 
been well-established, and such a template could be implemented across 
multiple agencies (i.e. NHTSA and EPA) with very little lag time in 
learning.\3178\ No comments were received on the transaction letter 
generated by the template.
---------------------------------------------------------------------------

    \3178\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
---------------------------------------------------------------------------

    For the final rule, NHTSA is finalizing the proposed requirements 
for its credit templates to be incorporated into provisions for Part 
536. NHTSA understands that manufacturers may be using similar credit 
reporting templates as part of their current business processes but has 
decided to adopt the template proposed in the NPRM. The NHTSA credit 
templates are an integral part of a long-range technology deployment 
that is already underway and will automate the NHTSA's CAFE database 
and web portal systems. When complete, the systems and portals will 
receive information directly from manufacturers and enable 
manufacturers, independently, to confirm credit trades and receive 
real-time credit balances. For this reason, diverging from the proposed 
templates for the final rule would impose unnecessary costs upon NHTSA. 
In the interest of accommodating the transition by manufacturers from 
other standardized templates, the agency will delay mandatory use of 
the CAFE credit template until January 1, 2021. Manufacturers may 
deviate from the generated language in the NHTSA credit trade 
confirmation by adding additional qualifications but, at a minimum, 
must include the core information generated by the template.
(c) Credit Transaction Information
    Credit trading among entities commenced in the CAFE program 
starting in MY 2011.\3179\ To date, NHTSA has received numerous credit 
trades from manufacturers but has only made limited information 
publicly available.\3180\ As discussed earlier, NHTSA maintains an 
online CAFE database with manufacturer and fleetwide compliance 
information that includes year-by-year accounting of credit balances 
for each credit holder. While NHTSA maintains this database, the 
agency's regulations currently state that it does not publish 
information on individual transactions, and NHTSA has not previously 
required trading entities to submit information regarding the 
compensation (whether financial, or other items of value) manufacturers 
receive in exchange for credits.3181 3182 Thus, NHTSA's PIC 
offers sparse information to those looking to determine the value of a 
credit.
---------------------------------------------------------------------------

    \3179\ 49 CFR 536.6(c).
    \3180\ Manufacturers may generate credits, but non-manufacturers 
may also hold or trade credits. Thus, the word ``entities'' is used 
to refer to those that may be a party to a credit transaction.
    \3181\ 49 CFR 536.5(e)(1).
    \3182\ NHTSA understands that not all credits are exchanged for 
monetary compensation. The proposal that NHTSA is adopting in this 
final rule requires entities to report compensation exchanged for 
credits, and is not limited to reporting monetary compensation.
---------------------------------------------------------------------------

    The lack of information regarding credit transactions means 
entities wishing to trade credits have little, if any, information to 
determine the value of the credits they seek to buy or sell. It is 
widely assumed that the civil penalty for noncompliance with CAFE 
standards largely determines the upper value of a credit, because it is 
logical to assume that manufacturers would not purchase credits if it 
cost less to pay civil penalties instead, but it is unknown how other 
factors affect the value. For example, a credit nearing the end of its 
five-model-year lifespan would theoretically be worth less than a 
credit within its full five-model-year lifespan. In the latter case, 
the credit holder would likely value the credit more, as it can be used 
for compliance purposes for a longer period of time.
    In the interest of facilitating a transparent and efficient credit 
trading market, NHTSA stated in the NPRM that consideration is being 
given to modifying its regulations for credit trade information. NHTSA 
sought comment in the NPRM about the feasibility of requiring more 
information disclosure around trades, including price information, 
noting that neither the public, shareholders, competitors, nor even the 
agencies themselves know the price of credit transactions. More 
specifically, NHTSA proposed requiring trading parties to submit 
information disclosing the identities of the parties to credit trades, 
the number of credits traded, and the amount of compensation exchanged 
for credits. Furthermore, NHTSA proposed that regulations

[[Page 25219]]

would also permit the agency to publish information about specific 
transactions on the PIC.
    NHTSA received comments from Volkswagen, Honda, Fiat Chrysler, 
Toyota, Global Automakers, the Auto Alliance, UCS, and from one private 
citizen, Mr. Jason Schwartz, regarding the scope of available credit 
information. All auto associations and manufacturers requested that 
NHTSA maintain the confidentiality of credit trades and transactions. 
The remaining commenters felt increased transparency would benefit the 
market.
    Global Automakers, the Auto Alliance, Fiat Chrysler, and Volkswagen 
stated that credit trades are business-to-business, contain internal 
information and can involve both financial and non-financial 
compensation between parties.\3183\ They stated credit transactions 
should be viewed as being similar to other competitive purchase 
agreements, which include non-disclosure terms and strict 
confidentiality with regard to cost and compensation.\3184\ They 
contended that negotiations must remain confidential to protect the 
sensitive business practices for both the buyer and seller, and that 
revealing purchasing terms could result in a competitive disadvantage 
for both.\3185\ Further, it was stated that certain transactions may 
not happen if they are publicized for fear of public criticism, making 
the program less efficient.\3186\
---------------------------------------------------------------------------

    \3183\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032; Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; Fiat 
Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Volkswagen, 
Detailed Comments, NHTSA-2017-0069-0583.
    \3184\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
    \3185\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
    \3186\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
0067-11943.
---------------------------------------------------------------------------

    Honda added that disclosing trading terms may not be as simple as a 
spot purchase at a given price.\3187\ Honda explained that it has 
undertaken a number of transactions for both CAFE and CO2 
credits, and there has been a range of complexity in these transactions 
due to numerous factors that are reflective of the marketplace, such as 
the volume of credits, compliance category, credit expiration date, a 
seller's compliance strategy, and even the CAFE penalty rate in effect 
at that time.\3188\ In addition, Honda stated that automakers have a 
range of partnerships and cooperative agreements with their own 
competitors.\3189\ Honda commented that credit transactions can be an 
offshoot of these broader relationships, and difficult to price 
separately and independently.\3190\ Thus, Honda believes there may not 
be a reasonable, or even meaningful, presentation of ``market'' 
information in a transaction ``price.'' \3191\ Finally, Honda concluded 
by stating that information on pricing terms and business partner 
pairings is highly competitive and, if made public, could divulge to 
competitors a buyer's and/or seller's future compliance strategy.\3192\ 
For these reasons, Honda believes it is appropriate to maintain the 
confidentiality of trade terms, pricing information, and of trading 
partner identification.\3193\
---------------------------------------------------------------------------

    \3187\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3188\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3189\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3190\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3191\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3192\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3193\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
---------------------------------------------------------------------------

    Fiat Chrysler stated that revealing credit transaction information 
would reveal highly confidential business information.\3194\ It stated 
that credit transaction information may reveal the technology that is 
most valued by a company and the value of putting certain technology 
into a vehicle.\3195\ It believed that credit trades are complex 
business transactions made at arm's length.\3196\ As such, they may 
include monetary and non-monetary compensation, non-disclosure 
provisions, and other sensitive terms.\3197\ Fiat Chrysler commented 
that publicizing such sensitive information could stifle the credit 
market and potentially result in uncompetitive outcomes, and could also 
decrease the efficiency in the credit trading marketplace.\3198\ Fiat 
Chrysler further stated that the NPRM's justifications for requiring 
the disclosure of credit transaction information is unfounded and the 
government has no need of this information in the regular course of 
doing business.\3199\
---------------------------------------------------------------------------

    \3194\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3195\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3196\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3197\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3198\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3199\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    The Auto Alliance, Honda, Toyota, and Volkswagen argued against 
NHTSA publishing credit movements each model year on its PIC. They 
stated that detailed credit banks by account holder are available to 
the public or entities wishing to engage in the credit market and that 
information is already sufficient.\3200\ Global Automakers further 
contended that the agencies know which companies are trading and how 
those credits are being used, which is all that should be required for 
administering the program.\3201\ The Auto Alliance argued that in 
private markets, trades and prices often are not made public; this 
privacy does not mean that the markets operate any less effectively, 
nor that the public at large does not benefit from the transactions 
that lower costs for all parties.\3202\
---------------------------------------------------------------------------

    \3200\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Honda, Detailed Comments, NHTSA-2018-0067-11818; Toyota, Detailed 
Comments, NHTSA-2018-0067-12150; Volkswagen, Detailed Comments, 
NHTSA-2017-0069-0583.
    \3201\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
    \3202\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Volkswagen further commented that revealing confidential purchase 
terms has no precedent in the automotive industry. Volkswagen's 
position is that it does not disclose contract pricing for purchasing 
fuel saving technologies from suppliers, such as for turbochargers or 
battery packs. Therefore, Volkswagen does not believe it is appropriate 
to disclose the purchase price for CAFE credits.\3203\
---------------------------------------------------------------------------

    \3203\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    Opposite views from those expressed by automobile manufacturers 
were received in the comments from UCS and Jason Schwartz. Both 
commenters strongly supported an increase in information regarding 
credit trading in the CAFE program.\3204\ They argued that more 
information will allow manufacturers to make better informed decisions 
and lead to greater industry efficiency in general.\3205\ UCS added 
that while the PIC does have some information, it is difficult to 
discern how the manufacturers are dividing credits to offset 
shortfalls.\3206\ It requested NHTSA disclose at least as much 
information as EPA provides from its program, if not providing more 
information on transaction price and

[[Page 25220]]

compliance category.\3207\ Jason Schwartz had similar arguments for 
more transparency. Mr. Schwartz added that the agencies can assume that 
credits may be traded at prices similar to the civil penalty rate for 
noncompliance under the CAFE standards, but not knowing the actual 
prices greatly complicates the agencies' estimations of the costs of 
complying with the standards.\3208\ Schwartz used several examples to 
explain and justify the need for making data on credit transactions, 
prices, and holdings publicly available to help the agency and the 
public assess the efficacy of the program.\3209\ He also explained that 
such information will enable the smooth operation of the credit market 
by enabling credit buyers to better evaluate the value of credits and 
placing all players on equal informational footing which facilitates 
price discovery, and assists buyers and sellers in reaching 
terms.\3210\ He added that regulators should require greater 
transparency to facilitate oversight.\3211\ He asserted his belief that 
greater transparency in tracking transactions and credits helps 
regulators detect fraud, manipulation, market power, abuse, and to 
enforce compliance.\3212\
---------------------------------------------------------------------------

    \3204\ UCS, Detailed Comments, NHTSA-2018-0067-12039; Jason 
Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
    \3205\ See, e.g., UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3206\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3207\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3208\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
    \3209\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
    \3210\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
    \3211\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
    \3212\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
---------------------------------------------------------------------------

    In response to these comments, NHTSA has decided not to share 
detailed information on credit transactions or the cost of individual 
credit transactions with the public. NHTSA agrees with manufacturers 
that revealing confidential purchase terms could result in a 
competitive disadvantage for both credit buyers and sellers, as well as 
harm to companies revealing highly confidential business materials. 
However, NHTSA believes that greater government oversight is needed 
over the CAFE credit market. NHTSA needs to understand more information 
surrounding trades, including costing information. As Honda recognized 
in its comments, NHTSA needs to understand the full range of complexity 
in transactions, monetary and non-monetary, in addition to the range of 
partnerships and cooperative agreements between credit account 
holders--which may impact the price of credit trades.\3213\ NHTSA also 
believes, as mentioned by commenters, that disclosure of information 
concerning credit trades is important for facilitating government 
oversight for protecting against fraud, manipulation, market power, and 
abuse which may occur in the credit market.
---------------------------------------------------------------------------

    \3213\ Honda, Detailed Comments, NHTSA-2018-0067-11819.
---------------------------------------------------------------------------

    NHTSA is adopting new reporting provisions in this final rule. 
Starting January 1, 2021, manufacturers will be required to submit all 
credit trade contracts, including costing and transactional 
information, to the agency. This information may be submitted 
confidentially, in accordance with 49 CFR part 512.\3214\ NHTSA will 
use this information to determine the true cost of compliance for all 
manufacturers. This information will allow NHTSA to assess better the 
impact of its regulations on the industry, and provide more insightful 
information to use in developing future rulemakings. This confidential 
information will be held by secure electronic means in NHTSA's database 
systems. As for public information, NHTSA will include more information 
on the PIC on aggregated credit transactions, such as the combined 
flexibilities all manufacturers used for compliance as shown in Figure 
IX-6, or information comparable to the credit information EPA makes 
available to the public. In the future, NHTSA will consider what 
information, if any, can be meaningfully shared with the public on 
credit transactional details or costs, while accounting for the 
concerns raised by the automotive industry.
---------------------------------------------------------------------------

    \3214\ See also 49 U.S.C. 32910(c).
---------------------------------------------------------------------------

(d) Precision of the CAFE Credit Adjustment Factor
    EPCA, as amended by EISA, required the Secretary of Transportation 
to establish an adjustment factor to ensure total oil savings are 
preserved when manufacturers trade credits.\3215\ The adjustment factor 
applies to credits traded between manufacturers and to credits 
transferred across a manufacturer's compliance fleets.
---------------------------------------------------------------------------

    \3215\ 49 U.S.C. 32903(f)(1).
---------------------------------------------------------------------------

    In establishing the adjustment factor, NHTSA did not specify the 
exact precision of the output of the equation in 49 CFR 536.4(b). 
NHTSA's standard practice has been round to the nearest four decimal 
places (e.g., 0.0001) for the adjustment factor. However, in the 
absence of a regulatory requirement, many manufacturers have contacted 
NHTSA for guidance, and NHTSA has had to correct several credit 
transaction requests. In some instances, manufacturers have had to 
revise signed credit trade documents and submit additional trade 
agreements to properly address credit shortfalls.
    NHTSA proposed in the NPRM to add requirements to 49 CFR 536.4 
specifying the precision of the adjustment factor by rounding to four 
decimal places (e.g., 0.0001). NHTSA has also included equations for 
the adjustment factor in its proposed credit transaction report 
template, mentioned above, with the same level of precision. NHTSA 
sought comment on this approach but received no comments, and therefore 
is finalizing this approach in this final rule.
(3) NHTSA Then Analyzes EPA-Certified CAFE Values for Compliance
    After manufacturers complete certification testing and submit their 
final compliance values to EPA, EPA verifies the data and issues final 
CAFE reports to manufacturers and NHTSA. NHTSA then evaluates whether 
the manufacturers' compliance categories (i.e., domestic passenger car, 
imported passenger car, and light truck fleets) meet the applicable 
CAFE standards. NHTSA uses EPA-verified data to compare fleet average 
standards with actual fleet performance values in each compliance 
category. Each vehicle a manufacturer produces has a fuel economy 
target based on its footprint (footprint curves are discussed above in 
Section II.C), and each compliance category has a CAFE standard 
measured in miles per gallon (mpg). The manufacturer's fleet average 
CAFE standard is calculated based on the fuel economy target value and 
production volume of each vehicle model. The CAFE performance is 
calculated based on the compliance value and production volume of each 
vehicle model. A manufacturer complies with the CAFE standard if its 
fleet average performance is greater than or equal to its required 
standard, or if it is able to use available compliance flexibilities, 
described below in Section IX.C.2.c. to resolve any shortfall.
    If the average fuel economy level of the vehicles in a compliance 
category falls below the applicable fuel economy standard, NHTSA 
provides written notification to the manufacturer that it has not met 
that standard. The manufacturer is then required to confirm the 
shortfall and either submit a plan indicating how it will allocate 
existing credits, or if it does not have sufficient credits available 
in that fleet, how it will earn, transfer, and/or acquire credits, or 
pay the appropriate civil penalty. The manufacturer must submit a 
credit allocation plan or payment within 60 days of receiving agency 
notification.

[[Page 25221]]

    NHTSA approves a credit allocation plan unless it finds the 
proposed credits are unavailable or that it is unlikely that the plan 
will result in the manufacturer earning sufficient credits to offset 
the projected shortfall. If a plan is approved, NHTSA revises the 
manufacturer's credit account accordingly. If a plan is rejected, NHTSA 
notifies the manufacturer and requests a revised plan or payment of the 
appropriate civil penalty. Similarly, if the manufacturer is delinquent 
in submitting a response within 60 days, NHTSA takes action to collect 
a civil penalty. If NHTSA receives and approves a manufacturer's plan 
to carryback future earned credits within the following three years in 
order to comply with current regulatory obligations, NHTSA will defer 
levying civil penalties for noncompliance until the date(s) when the 
manufacturer's approved plan indicates that the credits will be earned 
or acquired to achieve compliance. If the manufacturer fails to acquire 
or earn sufficient credits by the plan dates, NHTSA will initiate 
noncompliance proceedings to collect civil penalties.\3216\
---------------------------------------------------------------------------

    \3216\ See generally 49 CFR part 536.
---------------------------------------------------------------------------

(4) Civil Penalties for Noncompliance
    In the event that a manufacturer does not comply with a CAFE 
standard, EPCA provides that the manufacturer is potentially liable for 
a civil penalty.\3217\ The manufacturer determines whether to use 
available credits to reduce or offset its potential penalty.\3218\ This 
penalty rate is $5.50 for each tenth of a mpg that a manufacturer's 
average fuel economy falls short of the standard for a given model year 
multiplied by the total volume of those vehicles in the affected 
compliance category manufactured for that model year.\3219\ A person 
(or manufacturer) that violates 49 U.S.C. 32911(a), including general 
CAFE violations other than those for failing to comply with CAFE 
standards (i.e., fuel economy labeling violations), is also liable to 
the United States Government for a civil penalty of not more than 
$42,530 for each violation. A separate violation occurs for each day 
the violation continues. All penalties are paid to the U.S. Treasury 
and not to NHTSA.\3220\
---------------------------------------------------------------------------

    \3217\ 49 U.S.C. 32911-12.
    \3218\ See 49 U.S.C. 32912.
    \3219\ NHTSA finalized a retaining the $5.50 civil penalty rate 
in an April 2018 NPRM. See 83 FR 13904 (Apr. 2, 2018).
    \3220\ 49 U.S.C. 32912(e) allows for fiscal year 2008 and each 
fiscal year thereafter, the total amount deposited in the general 
fund of the Treasury during the preceding fiscal year from fines, 
penalties, and other funds obtained through enforcement actions 
conducted pursuant to EISA and EPCA (including funds obtained under 
consent decrees), the Secretary of the Treasury, subject to the 
availability of appropriations, shall: (1) transfer 50 percent of 
such total amount to the account providing appropriations to the 
Secretary of Transportation for the administration of this chapter, 
which shall be used by the Secretary to support rulemaking under 
this chapter; and (2) transfer 50 percent of such total amount to 
the account providing appropriations to the Secretary of 
Transportation for the administration of this chapter, which shall 
be used by the Secretary to carry out a program to make grants to 
manufacturers for retooling, reequipping, or expanding existing 
manufacturing facilities in the United States to produce advanced 
technology vehicles and components.

Potential Civil Penalty = $5.50 x (Avg. FE Performance-Avg. FE 
---------------------------------------------------------------------------
Standard) x 10 x Total Production

    Since the inception of the CAFE program, the U.S. Treasury has 
collected a total of $1,049,355,116 in CAFE civil penalty payments. 
Generally, import manufacturers have paid significantly more in civil 
penalties than domestic manufacturers, with the majority of payments 
made by import manufacturers for passenger cars and not light trucks. 
Over the total program lifetime, import manufacturers paid a total of 
$1,048,896,676 in CAFE penalties while domestic manufacturers paid a 
total of $458,440.\3221\
---------------------------------------------------------------------------

    \3221\ These totals include penalties associated with all fleets 
for these manufacturers. For example, the total penalties paid by 
import manufacturers includes penalties associated with shortfalls 
in those manufacturers' domestic passenger car fleets.
---------------------------------------------------------------------------

    Prior to the CAFE credit trade and transfer program, several 
manufacturers opted to pay civil penalties instead of complying with 
CAFE standards. Since NHTSA introduced trading and transferring, 
manufacturers have largely traded or transferred credits to achieve 
compliance, rather than paying civil penalties for noncompliance. NHTSA 
therefore assumes that buying and selling credits is a more cost-
effective strategy for manufacturers than paying civil penalties, in 
part, because it seems logical that the price of a credit is directly 
related to the civil penalty rate and decreases as a credit's life 
diminishes.\3222\ Prior to trading and transferring, on average, 
manufacturers paid $28,073,281.93 in civil penalty payments annually (a 
total of $814,125,176 from MYs 1982 to 2010). Since trading and 
transferring began, manufacturers now pay an average of $26,136,660 
each model year. The agency notes that six manufacturers have paid 
civil penalties since 2011 totaling $235,229,940; Fiat Chrysler paid a 
civil penalty in MY 2016 equal to $77,268,720.50 and in MY 2017 equal 
to $79,376,643.50 for for failing to meet the minimum domestic 
passenger car standards for those MYs. NHTSA expects that, over the 
next several years, manufacturers will face challenges in avoiding 
paying further civil penalties as standards increase in stringency. 
Compared to the current $5.50 CAFE civil penalty rate, a rate of $14 
would cause manufacturers that do not comply with CAFE to pay 
significantly higher civil penalties, potentially in the magnitude of 
hundreds of millions of dollars annually beyond current projections. 
Additionally, although NHTSA has not historically been privy to the 
monetary terms of credit trades, NHTSA expects that the price of 
credits would increase in line with any increase in the CAFE civil 
penalty rate.
---------------------------------------------------------------------------

    \3222\ See 49 CFR 536.4 for NHTSA's regulations regarding CAFE 
credits.
---------------------------------------------------------------------------

b) What Exemptions and Exclusions Does NHTSA Allow?
(1) Emergency and Law Enforcement Vehicles
    Under EPCA, manufacturers are allowed to exclude emergency 
vehicles, which include law enforcement vehicles, from their CAFE 
fleet.\3223\ All manufacturers that produce emergency vehicles have 
historically done so. NHTSA did not propose any changes to this 
exclusion and therefore is retaining the provision without change for 
the final rule.
---------------------------------------------------------------------------

    \3223\ 49 U.S.C. 32902(e).
---------------------------------------------------------------------------

(2) Small Volume Manufacturers
    Per 49 U.S.C. 32902(d), NHTSA established requirements for exempted 
small volume manufacturers in 49 CFR part 525, ``Exemptions from 
Average Fuel Economy Standards.'' The small volume manufacturer 
exemption is available for any manufacturer whose projected or actual 
combined sales (whether in the U.S. or not) are fewer than 10,000 
passenger automobiles in the model year two years before the model year 
for which the manufacturer seeks an exemption.\3224\ The manufacturer 
must submit a petition with information stating that the applicable 
CAFE standard is more stringent than the maximum feasible average fuel 
economy level that the manufacturer can achieve.\3225\ NHTSA must then 
issue by Federal Register notice, a proposed decision granting or 
denying the petition and inviting public comment.\3226\ If the agency 
proposed to grant the petition, the notice includes an alternative 
average fuel economy standard for the passenger automobiles 
manufactured by the manufacturer.\3227\ After conclusion of the public 
comment period, the agency publishes a final

[[Page 25222]]

decision in the Federal Register.\3228\ If the agency grants the 
petition, it establishes an alternative standard, which is the maximum 
feasible average fuel economy level for the manufacturers to which the 
alternative standard applies.\3229\ NHTSA did not propose and is not 
making any changes to the small volume manufacturer provision or 
alternative standards regulations in this rulemaking.
---------------------------------------------------------------------------

    \3224\ 49 CFR 525.5.
    \3225\ 49 CFR 525.7(h).
    \3226\ 49 CFR 525.8(c).
    \3227\ Id.
    \3228\ 49 CFR 525.8(e).
    \3229\ 49 U.S.C. 32902(d)(2); 49 CFR 525.8(e).
---------------------------------------------------------------------------

c) What Compliance Flexibilities and Incentives Are Currently Available 
Under the CAFE Program and How Do Manufacturers Use Them?
    There are several compliance flexibilities and incentives that 
manufacturers can use to achieve compliance with CAFE standards beyond 
applying fuel economy-improving technologies. Some compliance 
flexibilities and incentives are statutorily mandated by Congress 
through EPCA and EISA. These specifically include program credits 
generated from overcompliance, including the ability to carry-forward, 
carryback, trade and transfer credits, and special fuel economy 
calculations for dual- and alternative-fueled vehicles (discussed in 
turn, below). However, 49 U.S.C. 32902(h) expressly prohibits NHTSA 
from considering the availability of statutorily established credits 
(either for building dual- or alternative-fueled vehicles or from 
accumulated transfers or traders) in setting the level of the 
standards. Thus, NHTSA may not raise CAFE standards because 
manufacturers have enough credits to meet higher standards, or because 
alternative fuel vehicles (including electric vehicles) are available 
to help manufacturers achieve compliance. This is an important 
difference from EPA's authority under the CAA, which does not contain 
such a restriction, and which flexibility EPA has utilized in the past 
in determining appropriate levels of stringency for its program.
    Generating, trading, transferring, and applying CAFE credits is 
governed by statute.\3230\ Program credits are generated when a vehicle 
manufacturer's fleet over-complies with its standard for a given model 
year, meaning its vehicle fleet achieved a higher corporate average 
fuel economy value than the amount required by the CAFE program for 
that fleet in that model year. Conversely, if the fleet average CAFE 
level does not meet the standard, the fleet would incur debits (also 
referred to as a shortfall). A manufacturer whose fleet generates a 
credit shortfall in a given model year can resolve its shortfall using 
any one or combination of several credits flexibilities, including 
credit carryback, credit carry-forward, credit transfers, and credit 
trades.
---------------------------------------------------------------------------

    \3230\ 49 U.S.C. 32903.
---------------------------------------------------------------------------

    NHTSA also has promulgated compliance flexibilities and incentives 
consistent with EPCA's provisions regarding calculation of fuel economy 
levels for individual vehicles and for fleets.\3231\ These compliance 
flexibilities and incentives, which were first adopted in the 2012 rule 
for MYs 2017 and later, include A/C efficiency improvement and off-
cycle adjustments, and adjustments for advanced technologies in full-
size pickup trucks, including adjustments for mild and strong hybrid 
electric full-size pickup trucks and performance-based incentives in 
full-size pickup trucks. The fuel consumption improvement benefits of 
these technologies measured by various testing methods can be used by 
manufacturers to increase the CAFE performance of their fleets. As 
discussed below, the adjustments for advanced technologies in full-size 
pickup trucks will no longer be available beginning in MY 2022.
---------------------------------------------------------------------------

    \3231\ 49 U.S.C. 32904.
---------------------------------------------------------------------------

    Under NHTSA regulations, credit holders (including, but not limited 
to manufacturers) have credit accounts with NHTSA where they can, as 
outlined above, hold credits, and use them to achieve compliance with 
CAFE standards, by carrying forward, carrying back, or transferring 
credits across compliance categories. Manufacturers with excess credits 
in their accounts can also trade credits to other manufacturers, who 
may use those credits to resolve a shortfall currently or in a future 
model year. A credit may also be cancelled before its expiration date 
if the credit holder so chooses. Traded and transferred credits are 
subject to an ``adjustment factor'' to ensure total oil savings are 
preserved.\3232\ Credits earned before MY 2011 may not be traded or 
transferred.\3233\
---------------------------------------------------------------------------

    \3232\ 49 CFR 536.4(c).
    \3233\ 49 CFR 536.6(c).
---------------------------------------------------------------------------

    Credit ``carryback'' means that manufacturers are able to use 
credits to offset a deficit that had accrued in a prior model year, 
while credit ``carry-forward'' means that manufacturers can bank 
credits and use them towards compliance in future model years. EPCA, as 
amended by EISA allows manufacturers to carryback credits for up to 
three model years, and to carry-forward credits for up to five model 
years.\3234\ Credits expire the model year after which the credits may 
no longer be used to achieve compliance with fuel economy 
regulations.\3235\ Manufacturers seeking to use carryback credits must 
have an approved carryback plan from NHTSA demonstrating their ability 
to earn sufficient credits in future MYs that can be carried back to 
resolve the current MY's credit shortfall.
---------------------------------------------------------------------------

    \3234\ 49 U.S.C. 32903(a).
    \3235\ 49 CFR 536.3(b).
---------------------------------------------------------------------------

    Credit ``trading'' refers to the ability of manufacturers or 
persons to sell credits to, or purchase credits from, one another. EISA 
gave NHTSA discretion to establish by regulation a CAFE credit trading 
program, to allow credits to be traded between vehicle manufacturers, 
now codified at 49 CFR part 536.\3236\ EISA prohibited manufacturers 
from using traded credits to meet the minimum domestic passenger car 
CAFE standard.\3237\
---------------------------------------------------------------------------

    \3236\ 49 U.S.C. 32903(f).
    \3237\ 49 U.S.C. 32903(f)(2).
---------------------------------------------------------------------------

    As mentioned previously, the agencies sought comments in the NPRM 
on whether and how each agency's existing flexibilities and incentives 
might be amended, revised, or deleted to avoid the inefficiencies and 
market distortions as discussed earlier. NHTSA was concerned with the 
potential for unintended consequences. Specifically, comments were 
sought on the appropriate level of compliance flexibilities, including 
credit trading, in a program that is correctly designed to follow 
statutory direction to create maximum feasible fuel economy standards. 
Given that the credit trading program is discretionary under EISA, 
NHTSA also sought comments on whether the credit trading provisions in 
49 CFR part 536 should cease to apply beginning in MY 2022. Comments 
were sought on whether to allow all incentive-based adjustments, except 
those that are mandated by statute, to expire, in addition to other 
possible simplifications to reduce market distortion, improve program 
transparency and accountability, and improve overall performance of the 
compliance programs.
    The comments received from the public and NHTSA's responses to 
those comments are discussed below. A summary of all the flexibilities 
and incentives, and information on whether they were either retained or 
modified for the final rule, is presented in Table IX-1 through Table 
IX-4.

[[Page 25223]]

(1) Credit Carry-Forward and Back
    Under the CAFE program, when the average fuel economy of a 
compliance fleet manufactured in a particular model year exceeds its 
applicable average fuel economy standard, the manufacturer earns 
credits.\3238\ The credits may be applied to: (1) Any of the 3 
consecutive model years immediately before the model year for which the 
credits are earned; and (2) any of the 5 consecutive model years 
immediately after the model year for which the credits are earned. For 
example, a credit earned for exceeding model year 2017 standards will 
be usable for compliance purposes through and including the 2022 
compliance model year. NHTSA did not seek comment on or propose changes 
to any of the aspects of its lifespan for CAFE credits because of the 
existing statutory limitation set forth by Congress. The public offered 
no comments on such flexibilities under the CAFE program.
---------------------------------------------------------------------------

    \3238\ 49 U.S.C. 32903 and 49 CFR 536.
---------------------------------------------------------------------------

(2) Credit Trading
    All commenters responding to the NPRM on this issue favored 
retaining the existing CAFE credit trading program. Comments on credit 
trading were received from Volkswagen, Honda, General Motors, CARB, 
BorgWarner, Jaguar Land Rover, Fiat Chrysler, Global Automakers, the 
Auto Alliance, the Institute for Policy Integrity, Toyota, and academic 
commenters, Jeremy Michalek and Jason Schwartz. No comments were 
received supporting the idea of changing the existing credit trading 
program.
    In general, manufacturers' comments centered around problems in 
predicting whether consumers will purchase the fuel efficient vehicles 
necessary for manufacturers to meet their compliance obligations. They 
stated that continuing the credit trading program allows manufacturers 
to address uncertainty in the market better.\3239\ The Auto Alliance, 
Volkswagen, Fiat Chrysler, and Honda commented that credit 
flexibilities allow manufacturers to comply with the program even when 
faced with market uncertainties.\3240\ Honda stated that credit trading 
allows the government to set reasonable standards without fear of 
having to cater to the least-capable manufacturer.\3241\ Jaguar Land 
Rover stated the removal of NHTSA's credit trading programs would 
increase and intensify the dis-harmonization between the CO2 
and CAFE programs.\3242\
---------------------------------------------------------------------------

    \3239\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
0067-11943.
    \3240\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Volkswagen, Detailed Comments, NHTSA-2017-0069-0583-22; Fiat 
Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Honda, Detailed 
Comments, NHTSA-2018-0067-11818.
    \3241\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3242\ Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-
11916-9.
---------------------------------------------------------------------------

    Global Automakers, Fiat Chrysler, Jason Schwartz, and Jeremy 
Michalek each commented that the credit trading program allows for a 
more efficient compliance process given that more fuel-efficient 
manufacturers can sell their credits to manufacturers who fall 
short.\3243\ These commenters and BorgWarner stated that the program 
lowers the overall cost of reducing fuel consumption.\3244\ Likewise, 
Jaguar Land Rover, Fiat Chrysler, and General Motors argued compliance 
flexibilities, like trading, increase the ability to achieve higher 
fuel economy and reduced CO2 emissions. They found that the 
credit trading flexibility allows them to invest more money in 
technologies that will lead to future increases in their fuel 
economy.\3245\ Similarly, CARB argued credit flexibilities have been 
shown to be successful in reducing emissions and spurring innovation. 
It saw no reason to remove a successful program.\3246\
---------------------------------------------------------------------------

    \3243\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; 
Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162; Jeremy 
Michalek, Detailed Comments, NHTSA-2018-0067-11903.
    \3244\ BorgWarner, Detailed Comments, NHTSA-2018-0067-11895.
    \3245\ Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-
11916; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; 
General Motors, Detailed Comments, NHTSA-2018-0067-11858.
    \3246\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    Fiat Chrysler stated that credit trading allows manufacturers to 
provide more choices for consumers since manufacturers are not required 
to meet the standard exactly, but rather, they can purchase traded 
credits and then provide vehicles the public is demanding while still 
complying with fleet average standards.\3247\ They stated that this 
leads to the overall compliance of the U.S. fleet while allowing for 
more consumer choices. They further added that if the program is 
removed, manufacturers that currently generate credits from their fuel-
efficient fleet may find it more profitable to begin producing less 
fuel-efficient vehicles, perhaps even halting the current improvements 
in fuel efficiency across the industry.\3248\
---------------------------------------------------------------------------

    \3247\ General Motors, Detailed Comments, NHTSA-2018-0067-11943.
    \3248\ General Motors, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    Honda commented that regulatory flexibilities, such as credit 
trading, built into the CO2 and CAFE programs have become 
critical elements to the programs' success, especially in the face of 
product cadences with uneven sales that do not always match compliance 
obligations.\3249\ General Motors stated its belief that program 
flexibilities will continue to play an increasingly important role in 
reducing CO2 emissions and increasing fuel economy through 
technologies and innovations.\3250\ CARB stated that existing 
flexibilities create consistency in compliance planning for automakers 
for model years in the existing program.\3251\ Fiat Chrysler added that 
each of the CAFE and CO2 programmatic tools and 
flexibilities should be retained, improved and strengthened. Fiat 
Chrysler opined that this is a chance for the agencies to make better 
policies that work more efficiently and as intended, and cautioned that 
eliminating them now could have the serious negative impact of making 
the standards more stringent and costlier for manufacturers.\3252\
---------------------------------------------------------------------------

    \3249\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3250\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
    \3251\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
    \3252\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    NHTSA is not making changes to its credit trading provisions in the 
final rule. NHTSA sought comments on removing the optional credit 
trading program to explore public views on market distortions or 
windfalls that occur as a result of the credit trading program. 
However, commenters consistently opined that removing existing 
flexibilities might result in manufacturers not building certain types 
of vehicles. This could adversely impact compliance plans over multiple 
model years. NHTSA concurs with those views, and since this final rule 
adopts CAFE standards that continuously increase through MY 2026, 
understands the importance of allowing for credit trading to provide 
additional means of achieving compliance for manufacturers who face 
varying degrees of difficulty in achieving the standards the agencies 
are finalizing today. With increasing standards, credit trading 
flexibilities help to compensate for the possibility of an uneven sales 
mix of vehicle types and to aid with compliance planning.

[[Page 25224]]

Final sales volumes, as presented earlier, show a shift over the past 
several years in consumers purchasing more small SUVs subject to 
passenger car standards, and these vehicles are less fuel efficient 
than the compact and mid-sized passenger cars that previously dominated 
the market. The need to ensure consumer choice is adequately considered 
drives the need for NHTSA to provide credit trading flexibility to 
manufacturers. For example, even with increasing standards, a 
manufacturer could continue to sell certain types of vehicles with 
lower mpg performance over a longer period of time to satisfy its 
consumers by purchasing credits or carrying credits back from future 
model years to address the mpg fleet shortages caused by these 
vehicles, before ultimately having to introduce more fuel-efficient 
technologies. NHTSA believes that these types of scenarios are 
consistent with the purpose of the CAFE credit program, as adopted by 
Congress.
(3) Credit Transferring
    Credit ``transfer'' means the ability of manufacturers to move 
credits from their passenger car fleet to their light truck fleet, or 
vice versa. As part of the EISA amendments to EPCA, NHTSA was required 
to establish by regulation a CAFE credit transferring program, now 
codified at 49 CFR part 536, to allow a manufacturer to transfer 
credits between its car and truck fleets to achieve compliance with the 
standards.\3253\ For example, credits earned by overcompliance with a 
manufacturer's car fleet average standard may be used to offset debits 
incurred because of that manufacturer's failed to meet the truck fleet 
average standard in a given year. However, EISA imposed a cap on the 
amount by which a manufacturer could raise its CAFE performance through 
transferred credits: 1 mpg for MYs 2011-2013; 1.5 mpg for MYs 2014-
2017; and 2 mpg for MYs 2018 and beyond.\3254\ These statutory limits 
will continue to apply to the determination of compliance with CAFE 
standards. EISA also prohibits the use of transferred credits to meet 
the minimum domestic passenger car fleet CAFE standard.\3255\
---------------------------------------------------------------------------

    \3253\ See 49 U.S.C. 32903(g)(1).
    \3254\ 49 U.S.C. 32903(g)(3).
    \3255\ 49 U.S.C. 32903(g)(4).
---------------------------------------------------------------------------

    In the NPRM, NHTSA responded to the 2016 petition for rulemaking 
from the Auto Alliance and Global Automakers (Alliance/Global or 
Petitioners) asking to amend the regulatory definition of ``transfer'' 
as it pertains to compliance flexibilities.\3256\ In particular, 
Alliance/Global requested that NHTSA add text to the definition of 
``transfer'' stating that the statutory transfer cap in 49 U.S.C. 
32903(g)(3) applies when the credits are transferred. Alliance/Global 
assert that adding this text to the definition is consistent with 
NHTSA's prior position on this issue in the MYs 2012-2016 final rule, 
in which NHTSA stated:
---------------------------------------------------------------------------

    \3256\ Auto Alliance and Global Automakers Petition for Direct 
Final Rule with Regard to Various Aspects of the Corporate Average 
Fuel Economy Program and the Greenhouse Gas Program (June 20, 2016) 
at 13, available at https://www.epa.gov/sites/production/files/2016-09/documents/petition_to_epa_from_auto_alliance_and_global_automakers.pdf 
[hereinafter Alliance/Global Petition].

    NHTSA interprets EISA not to prohibit the banking of transferred 
credits for use in later model years. Thus, NHTSA believes that the 
language of EISA may be read to allow manufacturers to transfer 
credits from one fleet that has an excess number of credits, within 
the limits specified, to another fleet that may also have excess 
credits instead of transferring only to a fleet that has a credit 
shortfall. This would mean that a manufacturer could transfer a 
certain number of credits each year and bank them, and then the 
credits could be carried forward or back `without limit' later if 
and when a shortfall ever occurred in that same fleet.\3257\
---------------------------------------------------------------------------

    \3257\ 75 FR 25666 (May 7, 2010).

    NHTSA clarified in the NPRM, based upon a previous interpretation, 
that the transfer cap from EISA does not limit how many credits may be 
transferred in a given model year, but it does limit the application of 
transferred credits to a compliance category in a model year.\3258\ The 
interpretation concludes by stating that, ``Thus, manufacturers may 
transfer as many credits into a compliance category as they wish, but 
transferred credits may not increase a manufacturer's CAFE level beyond 
the statutory limits.'' \3259\
---------------------------------------------------------------------------

    \3258\ See, letter from O. Kevin Vincent, Chief Counsel, NHTSA 
to Tom Stricker, Toyota (July 5, 2011), available at https://isearch.nhtsa.gov/files/10-004142%20-%20Toyota%20CAFE%20credit%20transfer%20banking%20-%205%20Jul%2011%20final%20for%20signature.htm (last accessed Apr. 
18, 2018).
    \3259\ Id.
---------------------------------------------------------------------------

    NHTSA maintains its views that the transfer caps in 49 U.S.C. 
32903(g)(3) are properly read to apply to the application of credits. 
As NHTSA explained in the NPRM, it understands that the language in the 
MYs 2012-2016 final rule could be read to suggest that the transfer cap 
applies at the time credits are transferred. However, NHTSA believes 
its existing interpretation--that the transfer cap applies at the time 
the credits are used--is a more appropriate, plain language reading of 
the statute. While manufacturers have approached NHTSA with various 
interpretations that would essentially allow them to circumvent the 
EISA transfer cap, NHTSA believes such interpretations are improper 
because they would not give effect to the statutory transfer cap. 
Therefore, NHTSA proposed in the NPRM to deny Alliance/Global's 
petition to revise the definition of ``transfer'' in 49 CFR 536.3, and 
is now finalizing that denial.
    In response to the tentative denial of the petition above in the 
NPRM, comments were received from the Global Automakers and Toyota 
asking NHTSA to reconsider applying the transfer cap of 2.0 mpg per 
year when credits are transferred rather than when they are 
applied.\3260\ They reiterated that imposing the cap when applying the 
credits is overly burdensome, but did not provide any new information 
that has persuaded NHTSA to change its view that the petition should be 
denied. The Auto Alliance also stated that NHTSA should revise its 
definition of ``transfer'' to be more consistent with EPA.\3261\
---------------------------------------------------------------------------

    \3260\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032; Toyota, Detailed Comments, NHTSA-2018-0067-12150.
    \3261\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Other more general comments to the NPRM were also received from 
Walter Kreucher, Jeremy Michalek, Global Automakers, the Auto Alliance, 
and Toyota, regarding the use of the credit transfer flexibility. These 
commenters generally appreciated the transfer flexibility for its 
ability to reduce compliance costs.\3262\ More specifically, Walter 
Kreucher commented that the ability to transfer credits between 
compliance categories was beneficial for manufacturers and allowed for 
efficiency in the markets and reduce compliance costs.\3263\
---------------------------------------------------------------------------

    \3262\ See, e.g., Global Automakers, Detailed Comments, NHTSA-
2018-0067-12032.
    \3263\ Walter Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
---------------------------------------------------------------------------

    For the final rule, NHTSA is not making any changes to the existing 
provisions regarding transferring credits. NHTSA's position remains 
unchanged that the transfer cap in 49 U.S.C. 32903(g)(1) clearly limits 
the amount of performance increase for a manufacturer's fleet that 
fails to achieve the prescribed standards. The same statutory provision 
prevents NHTSA from changing its definition for transfer to be 
consistent with EPA. Consequently, NHTSA is not changing its definition 
or its previous interpretation that the application of transfer caps 
applies at the time the credits are used and not when

[[Page 25225]]

transferred. Therefore, NHTSA is finalizing its decision to deny the 
Auto Alliance and Global Automakers petition.
(4) Minimum Domestic Passenger Car Standard
    EPCA, as amended by EISA, addresses the minimum domestic passenger 
car standard (MDPCS), clearly stating that any manufacturer's 
domestically-manufactured passenger car fleet must meet the greater of 
either 27.5 mpg on average, or 92 percent of the average fuel economy 
projected by the Secretary for the combined domestic and non-domestic 
passenger automobile fleets manufactured for sale in the U.S. by all 
manufacturers in the model year, which projection shall be published in 
the Federal Register when the standard for that model year is 
promulgated in accordance with 49 U.S.C. 32902(b).\3264\ Since that 
requirement was added to the statute, NHTSA has always calculated the 
``92 percent'' as greater than 27.5 mpg. NHTSA published the 92 percent 
MDPCS for MYs 2017-2025 at 49 CFR 531.5(d) as part of the 2012 final 
rule. 49 CFR 531.5(e) explains that the published MDPCS for MYs 2022-
2025 are not final and may change when NHTSA sets standards for those 
model years. This is consistent with the statutory requirement that the 
92 percent standards must be determined at the time an overall 
passenger car standard is promulgated and published in the Federal 
Register.\3265\ Any time NHTSA establishes or changes a passenger car 
standard for a model year, the MDPCS for that model year must also be 
evaluated or re-evaluated and established accordingly. Thus, this final 
rule establishes the applicable MDPCS for MYs 2021-2026.
---------------------------------------------------------------------------

    \3264\ 49 U.S.C. 32902(b)(4).
    \3265\ 49 U.S.C. 32904(b)(4)(B).
---------------------------------------------------------------------------

    NHTSA considered comments received about the MDPCS, and discusses 
the comments and the agency's assessment in Section VIII.B.1.b).
    Table IX-7 lists the minimum domestic passenger car standards and 
compares them to standards that would correspond to each of the other 
regulatory alternatives considered. NHTSA has updated these to reflect 
its overall analysis and resultant projection for the CAFE standards 
finalized today, highlighted below as ``Preferred (Alternative 3),'' 
and has calculated what those standards would be under the no action 
alternative (as issued in 2012, as updated for the NPRM, and as further 
updated by today's analysis) and under the other alternatives described 
and discussed further in Section V, above.
[GRAPHIC] [TIFF OMITTED] TR30AP20.757

(5) Fuel Savings Adjustment Factor
    Under NHTSA's credit trading regulations, a fuel savings adjustment 
factor is applied when trading occurs between manufacturers or when a 
manufacturer transfers credits between its fleets, but not when a 
manufacturer carries credits forward or carries back credits within the 
same fleet.\3266\ The Alliance/Global requested in their 2016 petition 
that NHTSA require manufacturers to apply the fuel savings adjustment 
factor when credits are carried forward or carried back within the same 
fleet, including for existing, unused credits.
---------------------------------------------------------------------------

    \3266\ See 49 CFR 536.4(c).
---------------------------------------------------------------------------

    Per EISA, total oil savings must be preserved in NHTSA's credit 
trading program.\3267\ The statutory provisions for credit transferring 
within a manufacturer's fleet do not explicitly include the same 
requirement; however, NHTSA prescribed a fuel savings adjustment factor 
that applies to both credit trades between manufacturers and credit 
transfers between a manufacturer's compliance fleets. 
3268 3269
---------------------------------------------------------------------------

    \3267\ 49 U.S.C. 32903(f)(1).
    \3268\ 49 U.S.C. 32903(g).
    \3269\ See 49 CFR 536.5; see also 74 FR 14430 (Mar. 30, 2009) 
(Per NHTSA's final rule for MY 2011 Average Fuel Economy Standards 
for Passenger Cars and Light Trucks, ``There is no other clear 
expression of congressional intent in the text of the statute 
suggesting that NHTSA would have authority to adjust transferred 
credits, even in the interest of preserving oil savings. However, 
the goal of the CAFE program is energy conservation; ultimately, the 
U.S. would reap a greater benefit from ensuring that fuel oil 
savings are preserved for both trades and transfers. Furthermore, 
accounting for traded credits differently than for transferred 
credits does add unnecessary burden on program enforcement. Thus, 
NHTSA will adjust credits both when they are traded and when they 
are transferred so that no loss in fuel savings occurs.'').
---------------------------------------------------------------------------

    When NHTSA initially considered the preservation of oil savings, 
the agency

[[Page 25226]]

explained how one credit is not necessarily equal to another. For 
example, the fuel savings lost if the average fuel economy of a 
manufacturer falls one-tenth of an mpg below the level of a relatively 
low standard are greater than the average fuel savings gained by 
raising the average fuel economy of a manufacturer one-tenth of a mpg 
above the level of a relatively high CAFE standard.\3270\ The effect of 
applying the adjustment factor is to increase the numerical value of 
credits for compliance accounting that are earned for exceeding a CAFE 
standard, that are applied to a compliance category with a higher CAFE 
standard. Likewise, the adjustment factor has the effect of decreasing 
the numerical value of credits for compliance accounting that are 
earned for exceeding a CAFE standard, that are applied to a compliance 
category with a lower CAFE standard. While applying the adjustment 
factor impacts the compliance accounting value of credits which are 
denominated in miles per gallon, the adjustment maintains the real 
world value of credits from the perspective of the actual amount of 
fuel consumed or saved.
---------------------------------------------------------------------------

    \3270\ 74 FR 14432 (Mar. 30, 2009).
---------------------------------------------------------------------------

    Alliance/Global stated, in its 2016 petition, that while carry-
forward and carryback credits have been used for many years, the CAFE 
standards did not change during the Congressional CAFE freeze, meaning 
credits earned during those years were associated with the same amount 
of fuel savings from year to year.\3271\ Alliance/Global suggest that 
because there is no longer a Congressional CAFE freeze, NHTSA should 
apply the adjustment factor when moving credits within a manufacturer's 
fleet (i.e. carry-forward or carryback) beginning retroactively in MY 
2011.\3272\
---------------------------------------------------------------------------

    \3271\ Alliance/Global Petition at 10.
    \3272\ Alliance/Global Petition at 4.
---------------------------------------------------------------------------

    In the NPRM, NHTSA tentatively denied Alliance/Global's request to 
apply the fuel savings adjustment factor to credits that are carried 
forward or carried back within the same fleet to the extent that the 
request would impact credits carried forward or back retroactively 
within manufacturers' compliance fleets (i.e., credits that were 
generated prior to MY 2021 when the standards set by this rule first 
apply). NHTSA tentatively determined that applying the adjustment 
factor to credits earned in prior model years would be inequitable to 
apply retroactively. There would be an advantage for manufacturers 
carrying credits into future model years with higher CAFE standards. 
Manufacturers have historically planned compliance strategies based, at 
least in part, on the existing rules for how credits could be carried 
forward and back, including the lack of an adjustment factor when 
credits are carried forward or back within the same fleet. Thus, 
retroactively requiring an adjustment factor could disadvantage certain 
manufacturers without credits, and result in windfalls for other 
manufacturers.
    To explore the impact on future model years, NHTSA sought 
additional comments in the NPRM on the feasibility of applying the fuel 
savings adjustment factor to credits carried forwards or back starting 
in MY 2021. Global Automakers submitted new comments arguing that the 
application of fuel savings adjustment factors to credits carried 
forward or back would not result in a credit windfall. They believed 
this practice would ensure that credits have a consistent value over 
time.\3273\
---------------------------------------------------------------------------

    \3273\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
---------------------------------------------------------------------------

    Comments from Global Automakers provided no further justification 
that would persuade NHTSA to consider changing its position on denying 
the application of the adjustment factor to carry-forward and carryback 
credits beginning with MY 2011. NHTSA continues to be concerned about 
the inequitable outcome retroactive adjustments would have on the 
credit market. Therefore, NHTSA is finalizing its decision to deny the 
Alliance/Global request to apply the adjustment factor to credits 
carried forward or carried back within a compliance category 
retroactively beginning as early as MY 2011.
    Congress expressly required that DOT establish a credit 
``transferring'' regulation, to allow individual manufacturers to move 
credits from one of their fleets to another (e.g., using a credit 
earned for exceeding the light truck standard for compliance with the 
domestic passenger car standard). Congress also gave DOT discretion to 
establish a credit ``trading'' regulation so that credits may be bought 
and sold between manufacturers.\3274\ Congress specified that trading 
was for earned credits ``to be sold to manufacturers whose automobiles 
fail to achieve the prescribed standards such that the total oil 
savings associated with manufacturers that exceed the prescribed 
standards are preserved.'' \3275\ NHTSA established 49 CFR part 536 
believing it was consistent with the statute for transferred credits to 
be subject to the same ``adjustment factor'' to ensure total oil 
savings are preserved.\3276\ NHTSA believed that no further application 
of the adjustment factor to other credit flexibilities would be 
appropriate at that time. NHTSA sought comments in the NPRM to explore 
the consequences associated with applying the adjustment factor to 
credits carried forward and back starting in MY 2021, but no further 
insight was gained from the comments received. Therefore, NHTSA is 
retaining its existing requirements for the adjustment factor to be 
applied to transferred and traded credits only. NHTSA will continue 
considering potential application of the adjustment factor for all 
types of credit flexibilities in the future, and may consider 
regulatory changes in subsequent rulemakings.
---------------------------------------------------------------------------

    \3274\ 49 U.S.C. 32903(f).
    \3275\ 49 U.S.C. 32903(f)(1).
    \3276\ 74 FR 14196, 14434 (Mar. 30, 2009).
---------------------------------------------------------------------------

(6) VMT Estimates for Fuel Savings Adjustment Factor
    NHTSA uses the vehicle miles traveled (VMT) estimate as part of its 
fuel savings adjustment equation to ensure that when traded or 
transferred credits are used, fuel economy credits are adjusted to 
ensure fuel oil savings is preserved.\3277\ For MYs 2017-2025, NHTSA 
finalized VMT values of 195,264 miles for passenger car credits, and 
225,865 miles for light truck credits.\3278\ These VMT estimates 
harmonized with those used in EPA's CO2 program. For MYs 
2011-2016, NHTSA estimated different VMTs by model year.
---------------------------------------------------------------------------

    \3277\ See 49 CFR 536.4(c).
    \3278\ 77 FR 63130 (Oct. 15, 2012).
---------------------------------------------------------------------------

    In the NPRM, NHTSA explained that Alliance/Global requested in 
their 2016 petition that NHTSA apply fixed VMT estimates to the fuel 
savings adjustment factor for MYs 2011-2016 similar to how NHTSA 
handled VMT values for MYs 2017-2025.\3279\ NHTSA rejected a similar 
request from the Auto Alliance in the MY 2017 and later rulemaking, 
citing lack of scope, and expressing concern about the potential loss 
of fuel savings.\3280\
---------------------------------------------------------------------------

    \3279\ Alliance/Global Petition at 5, 11.
    \3280\ Id.
---------------------------------------------------------------------------

    The Alliance/Global argued that data from MYs 2011-2016 demonstrate 
that no fuel savings would have been lost, as was NHTSA's 
concern.\3281\ Alliance/Global asserted that by not revising the MY 
2012-2016 VMT estimates, credits earned during that timeframe were 
undervalued.\3282\ Therefore, Alliance/

[[Page 25227]]

Global argued that NHTSA should retroactively revise its VMT estimates 
to ``reflect better the real-world fuel economy results.'' \3283\
---------------------------------------------------------------------------

    \3281\ Alliance/Global Petition at 11.
    \3282\ Id.
    \3283\ Alliance/Global Petition at 11.
---------------------------------------------------------------------------

    Such retroactive adjustments could have unfair adverse effects upon 
manufacturers for decisions they made based on the regulations as they 
existed at the time. As Alliance/Global acknowledged, adjusting VMT 
estimates would disproportionately affect manufacturers that have a 
credit deficit and were part of EPA's Temporary Lead-time Allowance 
Alternative Standards (TLAAS). The TLAAS program sunsets for MYs 2021 
and later. Given that some manufacturers would be disproportionately 
affected were NHTSA to adopt Alliance/Global's proposal, in the NPRM, 
NHTSA tentatively denied Alliance/Global's request to change the 
agency's VMT schedules for MYs 2011-2016 retroactively. Alliance/
Global's suggestion that a TLAAS manufacturer should be allowed to 
elect either approach does not change the fact that manufacturers in 
the TLAAS program made production decisions based on the regulations as 
understood at the time.\3284\ NHTSA sought comments on the Alliance/
Global requests in the NPRM.
---------------------------------------------------------------------------

    \3284\ See id. at 11-12, n.12.
---------------------------------------------------------------------------

    However, no further comments were received on this issue in 
response to the NPRM. Therefore, NHTSA is finalizing its decision to 
deny the Alliance/Global request to modify the VMT schedules for MYs 
2011-2016.
(7) Special Fuel Economy Calculations for Dual and Alternative Fueled 
Vehicles
    As discussed at length in prior rulemakings, EPCA, as amended by 
EISA, encouraged manufacturers to build alternative-fueled and dual- 
(or flexible-) fueled vehicles by providing special fuel economy 
calculations for ``dedicated'' (that is, 100 percent) alternative 
fueled vehicles and ``dual-fueled'' (that is, capable of running on 
either the alternative fuel or gasoline/diesel) vehicles.
    Dedicated alternative-fuel automobiles include electric, fuel cell, 
and compressed natural gas vehicles, among others. The statutory 
provisions for dedicated alternative fuel vehicles in 49 U.S.C. 
32905(a) state that the fuel economy of any dedicated automobile 
manufactured after MY 1992 shall be measured ``based on the fuel 
content of the alternative fuel used to operate the automobile. A 
gallon of liquid alternative fuel used to operate a dedicated 
automobile is deemed to contain 0.15 gallon of fuel.'' Under EPCA, for 
dedicated alternative fuel vehicles, there are no limits or phase-out 
for this special fuel economy calculation, unlike for duel-fueled 
vehicles, as discussed below.
    EPCA's statutory incentive for dual-fueled vehicles at 49 U.S.C. 
32906 and the measurement methodology for dual-fueled vehicles at 49 
U.S.C. 32905(b) and (d) expire after MY 2019; therefore, NHTSA had to 
examine the future of these provisions in the MY 2017 and later CAFE 
rulemaking. NHTSA and EPA concluded that it would be inappropriate to 
measure duel-fueled vehicles' fuel economy like that of conventional 
gasoline vehicles with no recognition of their alternative fuel 
capability, which would be contrary to the intent of EPCA/EISA. The 
agencies determined that for MY 2020 and later vehicles, the general 
statutory provisions authorizing EPA to establish testing and 
calculation procedures provide discretion to set the CAFE calculation 
procedures for those vehicles. The methodology for EPA's approach is 
outlined in the 2012 final rule for MYs 2017 and later at 77 FR 63128 
(Oct. 15, 2012). In the NPRM, NHTSA sought comments on that current 
approach.
    NHTSA received comments from the Coalition for Renewable Natural 
Gas, NGV America, the American Gas Association, the American Public Gas 
Association, CARB, Ingevity Corporation, Fuel Freedom Foundation, UCS, 
National Farmers Union, Indiana Corn Growers Association, Volkswagen, 
and a joint submission from Ariel Corp. and VNG.co.
    Fuel Freedom Foundation and the National Farmers Union asserted 
that the agencies should continue offering incentives for emerging 
technology vehicles including natural gas vehicles, internal combustion 
engine (ICE) vehicles that encourage renewable fuel use, electric and 
hydrogen fuel cell vehicles, flex-fuel vehicles (FFVs), and dedicated 
high-octane vehicles designed for compatibility with mid-level ethanol 
blends.\3285\
---------------------------------------------------------------------------

    \3285\ Fuel Freedom Foundation, Detailed Comments, NHTSA-2018-
0067-12016; National Farmers Union, Detailed Comments, NHTSA-2018-
0067-11972.
---------------------------------------------------------------------------

    Indiana Corn Growers Association and Fuel Freedom Foundation 
specified that FFVs, as well as vehicles that run on mid-level ethanol 
blends, should receive credit for the petroleum reduction value.\3286\ 
For vehicles using higher-ethanol blends, these commenters stated that 
the agencies should establish more accurate petroleum equivalency 
factors for the proportion of ethanol versus gas.\3287\ Clean Fuels 
Development Coalition requested credits for producing ``Engines 
Optimized for High-Octane'' be reinstated.\3288\ Volkswagen made the 
same request and added that a pathway to higher-octane fuel is 
important to it.\3289\
---------------------------------------------------------------------------

    \3286\ Indiana Corn Growers Association, Detailed Comments, 
NHTSA-2018-0067-12003; Fuel Freedom Foundation, Detailed Comments, 
NHTSA-2018-0067-12016.
    \3287\ Fuel Freedom Foundation, Detailed Comments, NHTSA-2018-
0067-12016.
    \3288\ Clean Fuels Development Coalition, Detailed Comments, 
NHTSA-2018-0067-12031.
    \3289\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    Ariel Corp. and VNG.co, the Coalition for Renewable Natural Gas, 
NGVAmerica, the American Gas Association, and the American Public Gas 
Association commented that the agencies should expand incentives for 
natural gas vehicles in the light-duty sector especially for pick-up 
trucks, work vans, and sport utility vehicles.\3290\ They argued that 
current incentives are not strong enough to induce manufacturers to 
produce natural gas vehicles. They further requested that the market 
penetration rates be removed for light-duty trucks.\3291\
---------------------------------------------------------------------------

    \3290\ Joint submission from Ariel Corp and VNG.co LLC, Detailed 
Comments, NHTSA-2018-0067-7573; Joint submission from the Coalition 
for Renewable Natural Gas, NVG America, the American Gas 
Association, and American Public Gas Association, Detailed Comments, 
NHTSA-2018-0067-11967.
    \3291\ See, e.g., joint submission from the Coalition for 
Renewable Natural Gas, NGVAmerica, the American Gas Association, and 
the American Public Gas Association, Detailed Comments, NHTSA-2018-
0067-11967.
---------------------------------------------------------------------------

    The Coalition for Renewable Natural Gas, NGVAmerica, the American 
Gas Association, and the American Public Gas Association argued that an 
AMFA factor of 0.15 is low and because some natural gas vehicles can 
operate at 100 percent natural gas, a higher fuel economy credit is 
justified. They further supported a permanent use of the 0.15 factor 
for dual-fuel vehicles.\3292\ Similarly, Ingevity Corporation, and 
Ariel Corp. and VNG.co argued that natural gas vehicle emissions should 
return to the 0.15 divisor.\3293\
---------------------------------------------------------------------------

    \3292\ Joint submission from the Coalition for Renewable Natural 
Gas, NGVAmerica, the American Gas Association, and the American 
Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
    \3293\ Ingevity Corporation, Detailed Comments, NHTSA-2018-0067-
8666; Joint submission from Ariel Corp. and VNG.co LLC, Detailed 
Comments, NHTSA-2018-0067-7573.
---------------------------------------------------------------------------

    Ingevity Corporation, Ariel Corp. and VNG.co, the Coalition for 
Renewable

[[Page 25228]]

Natural Gas, NGVAmerica, the American Gas Association, and the American 
Public Gas Association requested that the agencies remove the minimum 
driving range of natural gas compared to gasoline and ``drive to 
empty'' design requirements for dual-fueled natural gas vehicles and 
allow higher utility factors based on driving range only, so that dual-
fuel NGVs are treated similarly to PHEVs. They stated a belief that the 
design constraints for dual-fuel NGVshold NGVs to an unfairly higher 
standard.\3294\ As discussed above in Section IX.B, EPA is removing 
these design constraints for dual-fuel NGVs.
---------------------------------------------------------------------------

    \3294\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666; Joint 
submission from Ariel Corp. and VNG.co LLC, Detailed Comments, 
NHTSA-2018-0067-7573; Joint submission from The Coalition for 
Renewable Natural Gas, NGVAmerica, the American Gas Association, the 
American Public Gas Association, Detailed Comments, NHTSA-2018-0067-
11967.
---------------------------------------------------------------------------

    CARB argued that flexibilities for natural gas vehicles and high-
octane blend vehicles are not yet warranted.\3295\ Similarly, UCS 
argued that natural gas is a greenhouse gas and benefits from natural 
gas vehicles are undermined by their costs. UCS further commented that 
natural gas vehicle technology does not need any incentives since it 
has already been deployed and in the market.\3296\
---------------------------------------------------------------------------

    \3295\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
    \3296\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    In response to comments, NHTSA has determined that EPCA and EISA 
prescribe the incentive that is used for dedicated liquid and gaseous 
alternative fuel vehicles, and the CAFE program will continue to use 
those statutory incentives. For dedicated alternative fuel vehicles, 
the statute provides a significant incentive that only counts 15 
percent of the actual energy used.\3297\ For dual fuel vehicles, NHTSA 
has determined that, for the portion of operation that occurs on an 
alternative fuel, it is consistent to use the same incentive that is 
specified by EPCA and EISA for dedicated fuel vehicles. For example, 
for the hypothetical case of a vehicle that operates 99 percent of the 
time on an alternative fuel, it would be appropriate for that vehicle 
to receive nearly the same incentive as a dedicated alternative fuel 
vehicle that operates 100 percent of the time on alternative fuel. 
Applying the same 15 percent of energy used incentive for both 
dedicated and duel fuel vehicles remains appropriate. NHTSA therefore 
is not adopting any new incentives for any alternative fueled vehicles.
---------------------------------------------------------------------------

    \3297\ 32905(a) ``. . . A gallon of a liquid alternative fuel 
used to operate a dedicated automobile is deemed to contain .15 
gallon of fuel.'' 32905(c) ``. . . One hundred cubic feet of natural 
gas is deemed to contain .823 gallon equivalent of natural gas. The 
Secretary of Transportation shall determine the appropriate gallon 
equivalent of other gaseous fuels. A gallon equivalent of gaseous 
fuel is deemed to have a fuel content of .15 gallon of fuel.''
---------------------------------------------------------------------------

D. Compliance Issues That Affect Both the CO2 and CAFE 
Programs

    Because the real world CO2 emissions reduction benefits 
of certain technologies cannot be measured or fully measured using 2-
cycle test procedures, EPA established new compliance flexibilities 
under its CAA authority, starting in MY 2012, that allow manufacturers 
credit for emission compliance for installing these technologies. Those 
flexibilities are designed to recognize improvements in A/C systems 
with greater efficiency and other ``off-cycle'' technologies that 
reduce real world tailpipe CO2 emissions. More specifically, 
real world improvements that cannot be measured or fully measured on 2-
cycle tests are determined and used to calculate additional 
CO2 credits (in Megagrams (Mg)) for each model type that has 
the technologies. Because these tailpipe CO2 improving 
technologies also impact fuel economy, NHTSA adopted the same 
flexibilities and incentives beginning in MY 2017. EPA and NHTSA also 
established incentives for both the CO2 and CAFE programs 
that give added compliance credits and fuel consumption improvement 
values for the production of strong and mild hybrid full-size pickup 
trucks beginning in MY 2017.\3298\ EPA adjusts manufacturers' CAFE 
performance values using the emissions benefits or incentives provided 
for these technologies. EPA developed a methodology for manufacturers 
to increase their passenger car and light truck fuel economy 
performance in accordance with procedures set forth by EPA in 40 CFR 
part 600. For the NHTSA CAFE program, the CO2 reductions (in 
grams per mile) are converted to fuel consumption improved values 
(FCIVs, gallons per mile) and then the benefits are summed for all the 
model types in the manufacturer's fleets. The total FCIVs are used to 
adjust and increase manufacturers' CAFE (mpg) performance values.
---------------------------------------------------------------------------

    \3298\ See 40 CFR 86.1867-86.1868, 86.1870.
---------------------------------------------------------------------------

    It is important to note that while these flexibilities and 
incentives have similar value for compliance in the CAFE and 
CO2 programs, there are differences in how they are 
accounted for in each of the programs due to differences in the 
structure of the programs. The CAFE program accounts for A/C efficiency 
and off-cycle improvements through EPA measurement procedures that 
determine fuel consumption improvement values (FCIVs). The CAFE A/C 
efficiency and off-cycle provisions do not involve manufacturer 
credits.\3299\ There are no bankable, tradable, or transferrable 
credits earned by a manufacturer for implementing more efficient A/C 
systems or installing an off-cycle technology. In fact, the only 
credits provided for in NHTSA's CAFE program are those earned by 
overcompliance with a standard.\3300\ As discussed above, EPA adjusts 
CAFE performance values based on the FCIVs generated through the use of 
these technologies. Off-cycle technologies and A/C efficiency 
improvements represent adjustments to individual vehicle compliance 
values based on the fuel consumption improvement values of these 
technologies.
---------------------------------------------------------------------------

    \3299\ This is not to be confused with EPA's parallel program, 
which refers to the GHG's consideration of A/C improvements and off-
cycle technologies as ``credits.''
    \3300\ 49 U.S.C. 32903.
---------------------------------------------------------------------------

    Illustrative of this confusion, in the 2016 Alliance/Global 
petition, the petitioners asked NHTSA to avoid imposing unnecessary 
restrictions on the use of credits. Alliance/Global referenced language 
from an EPA report that stated compliance is assessed by measuring the 
tailpipe emissions of a manufacturer's vehicles, and then reducing 
vehicle CO2 compliance values depending on A/C efficiency 
improvements and off-cycle technologies.\3301\ This language is 
consistent with NHTSA's statement in the MY 2017 and later final rule, 
which explained how the agencies coordinate and apply off-cycle and A/C 
adjustments. ``There will be separate improvement values for each type 
of credit, calculated separately for cars and for trucks. These 
improvement values are subtracted from the manufacturer's 2-cycle-based 
fleet fuel consumption value to yield a final new fleet fuel 
consumption value, which would be inverted to determine a final fleet 
fuel CAFE value.'' \3302\
---------------------------------------------------------------------------

    \3301\ See Alliance/Global Petition at 15.
    \3302\ 77 FR 62726 (Oct. 15, 2012).
---------------------------------------------------------------------------

    In the NPRM, NHTSA proposed to deny Alliance/Global's request 
because what the petitioners refer to as ``technology credits'' are 
actually FCIVs applied to the fuel economy performance of individual 
vehicles.\3303\

[[Page 25229]]

Thus, these adjustments are not actually ``credits,'' per the usage of 
``credit'' in EPCA/EISA and are not subject to the ``carry-forward'' 
and ``carryback'' provisions in 49 U.S.C. 32903. To alleviate 
confusion, and to ensure consistency in nomenclature, NHTSA proposed to 
update language in its regulations to reflect that the use of the term 
``credits'' to refer to A/C efficiency and off-cycle technology 
adjustments should actually be termed fuel consumption improvement 
values (FCIVs). No further comments were received on this issue in 
response to the NPRM. For the final rule, NHTSA is finalizing the 
proposed changes in its regulations to remove the term ``credits'' and 
to replace it with the term ``adjustments'' for the FCIV benefit for A/
C and off-cycle technologies in the CAFE program.
---------------------------------------------------------------------------

    \3303\ The agencies also refer to A/C and off-cycle technology 
improvement values as ``credits'' sporadically throughout their 
regulations. NHTSA is amending its regulations to reflect these are 
adjustments and not actual credits that can be carried forward or 
back. For a further discussion, see above.
---------------------------------------------------------------------------

    Manufacturers seeking to use these flexibilities and incentives 
start the process each model year by submitting information to EPA and 
seeking any necessary approvals, as appropriate. The use of certain 
technologies only requires submitting information to EPA, whereas 
others require a formal request process for approval. The differences 
are explained in the following sections. The compliance information 
manufacturers must submit to EPA describes the technologies, the 
flexibilities or incentives being used, and the testing approach for 
deriving benefits. Initial information is required as a part of the EPA 
certification process, as specified by 40 CFR 86.1843-01 in advance of 
each model year. For technologies requiring approvals, EPA must confirm 
the manufacturer's testing approach, receive test results to assess the 
benefit of the technology, and then where applicable issue a Federal 
Register notice that invites public comment. EPA review and 
determination usually occurs before the end of the compliance model 
year, if manufacturers provide information to EPA on a timely basis. To 
receive the benefit under the CAFE program for technologies that 
require approvals, manufacturers must concurrently submit to NHTSA the 
same information that is sent to EPA. EPA consults with NHTSA in 
reviewing A/C efficiency and off-cycle adjustments to fuel economy 
performance values that require approval. NHTSA provides EPA its 
assessment of the suitability of a technology considering: (1) Whether 
the technology has a direct impact upon improving fuel economy 
performance; (2) whether the technology is related to crash-avoidance 
technologies, safety critical systems or systems affecting safety-
critical functions, or technologies designed for the purpose of 
reducing the frequency of vehicle crashes; (3) information from any 
assessments conducted by EPA related to the application, the 
technology, and/or related technologies; and (4) any other relevant 
factors.
    EPA and NHTSA sought comments on several aspects of the shared 
flexibilities and incentives in the NPRM. Presented in the following 
sections is a summary of the comments received and the agencies final 
decisions for the final rule.
1. Incentives for Advanced Technologies in Full-Size Pickup Trucks
    In the 2012 rulemaking for MYs 2017 and beyond, EPA and NHTSA 
created incentives to encourage implementation of hybrid electric full 
size pickup trucks for both the CO2 and CAFE programs. 
CO2 credits and CAFE FCIVs were made available for 
manufacturers that produce full-size pickup trucks with Mild HEV or 
Strong HEV technology, provided the percentage of production with the 
technology is greater than specified percentages.\3304\ In addition, 
CO2 credits and CAFE FCIVs were made available for 
manufacturers that produce full-size pickups with other technologies 
that enables full size pickup trucks to exceed performance of their 
CO2 or CAFE targets based on footprints by specified 
amounts.\3305\ These performance-based incentives created a technology-
neutral path (as opposed to the other technology-encouraging path) to 
achieve the CO2 credits and CAFE FCIVs, which would 
encourage the development and application of new technological 
approaches.
---------------------------------------------------------------------------

    \3304\ 77 FR 62651 (Oct. 15, 2012).
    \3305\ Id.
---------------------------------------------------------------------------

    EPA and NHTSA established limits on the vehicles eligible to 
qualify for these incentives; a truck must meet minimum criteria for 
bed size and towing or payload capacity, and meet minimum production 
thresholds (in terms of a percentage of a manufacturer's full-size 
pickup truck fleet) in order to qualify for the incentives. As 
designed, the strong hybrid credit is 20 grams/mile per vehicle, 
available through MY 2025, if installed on at least 10 percent of the 
manufacturer's full-size pickup truck fleet in the model year. The 
program also included an incentive for mild hybrids of 10 grams/mile 
per vehicle during MYs 2017-2021. To be eligible the manufacturer would 
have to show that the mild hybrid technology is utilized in a specified 
portion of its truck fleet beginning with at least 20 percent of a 
company's full-size pickup production in MY 2017 and ramping up to at 
least 80 percent in MY 2021.\3306\
---------------------------------------------------------------------------

    \3306\ 77 FR 62651-2 (Oct. 15, 2012).
---------------------------------------------------------------------------

    At present, no manufacturer has qualified to use the full-size 
pickup truck incentives. One vehicle manufacturer introduced a mild 
hybrid pickup truck for MY 2019 but did not meet the minimum production 
threshold. Others have announced potential collaborations, or have 
already started production on future hybrid or electric models.\3307\
---------------------------------------------------------------------------

    \3307\ Chrysler released the 2019 Dodge Ram 1500 ``eTorque'' 
(see https://www.fueleconomy.gov/feg/Find.do?action=sbs&id=40736&id=40737&id=40394&id=40397) which 
qualifies as a mild hybrid pickup truck by replacing the traditional 
alternator on the engine with a 48-volt Li-on battery-powered, belt-
driven motor generator that improves performance, efficiency, 
payload, towing capabilities and drivability. The production volume 
of these vehicles did not qualify for the full-size pickup truck 
electric/hybrid incentive for MY 2019. Other vehicle models are 
currently in research or in development for future years but it is 
uncertain whether they will reach the required sales volumes to 
qualify for incentives. For example, the hybrid and battery-electric 
versions of the F-150 pickup, see https://www.trucks.com/2019/09/18/ford-truck-engineer-explains-electric-f-150-pickup-plans (September 
18, 2019), or the new electric pickup truck manufactured by Rivian, 
https://www.trucks.com/2019/04/24/ford-plans-new-electric-truck-rivian-invests-500-million/ (April 24, 2019); or the Tesla all 
electric pickup truck (https://www.cnn.com/2019/11/08/success/tesla-pickup-reveal/index.html) (November 8, 2019).
---------------------------------------------------------------------------

    Prior to the NPRM, the agencies received input from automakers that 
these incentives should be extended and available to all light-duty 
trucks (e.g., cross-over vehicles, minivans, sport utility vehicles, 
and smaller-sized pickups) and not only full-size pickup trucks.\3308\ 
Automakers also recommended that the program's eligibility production 
thresholds should be removed because they discourage the application of 
technology since manufacturers cannot be confident of achieving the 
thresholds. Some stakeholders have also suggested an additional 
incentive for strong and mild hybrid passenger cars. In the proposal, 
the agencies sought comment on whether these incentives should be 
expanded along the lines suggested by stakeholders, on the basis that 
perhaps these incentives could lead to additional product offerings of 
strong hybrids, and technologies that offer similar emissions 
reductions, which could enable manufacturers to achieve additional 
long-term CO2 emissions reductions. In addition, the 
agencies sought comment on whether to extend either the incentive for 
hybrid full-size pickup trucks or the performance-based incentive past 
the dates that EPA specified in the 2012 final rule for MY

[[Page 25230]]

2017 and later. The agencies also sought comment on eliminating 
incentive programs, as discussed above.
---------------------------------------------------------------------------

    \3308\ 83 FR 43461 (Aug. 24, 2018).
---------------------------------------------------------------------------

    The agencies received a variety of comments on the full-size pickup 
truck incentives. Comments were received from General Motors, 
Volkswagen, Honda, BorgWarner, Fiat Chrysler, Toyota, DENSO 
International, Ford, CARB, Global Automakers, UCS, Electric Drive 
Transportation Association, the Auto Alliance, Ariel Corp. and VNG.co, 
ACEEE, the Coalition for Renewable Natural Gas, NGVAmerica, the 
American Gas Association, and the American Public Gas Association.
    The Auto Alliance, Toyota, General Motors, BorgWarner, Global 
Automakers, and Volkswagen advocated to expand the full-size pickup 
truck hybrid incentives to all hybrid vehicles.\3309\ They argued that 
prices for all hybrid-drive technologies are projected to remain high 
and consumer demand for these vehicles is still slow to increase.\3310\ 
They asserted that expanding the full-size pickup truck hybrid 
incentive to all hybrid vehicles will help encourage investments in 
hybrid technology and continue to help manufacturers address their 
compliance challenges.\3311\ Similarly, these commenters reported that 
the current market, fueled by consumer demand for SUVs and lower than 
expected gas prices, is not conducive to consumer acceptance of or 
demand for electric vehicles.\3312\ For these reasons, they stated 
their belief that it is important to support adjustments and expansion 
of the current incentives to promote hybrid technologies.
---------------------------------------------------------------------------

    \3309\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Toyota, Detailed Comments, NHTSA-2018-0067-12150; General Motors, 
Detailed Comments, NHTSA-2018-0067-11858; BorgWarner, Detailed 
Comments, NHTSA-2018-0067-11895; Global Automakers, Detailed 
Comments, NHTSA-2018-0067-12032; Volkswagen, Detailed Comments, 
NHTSA-2017-0069-0583.
    \3310\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
    \3311\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
0067-11858.
    \3312\ See, e.g., Toyota, Detailed Comments, NHTSA-2018-0067-
12150.
---------------------------------------------------------------------------

    The Auto Alliance, DENSO International, Global Automakers, Fiat 
Chrysler, and Honda also argued for alternative pathways for the 
agencies to consider allowing the full-size pickup truck hybrid 
incentives to be expanded to the light-duty truck segment, but not to 
all passenger vehicles. They argued that hybrid technology has been 
slow to be applied in the light-duty truck segment, but has been 
broadly applied to passenger cars.\3313\
---------------------------------------------------------------------------

    \3313\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
DENSO, Detailed Comments, NHTSA-2018-0067-11880; Global Automakers, 
Detailed Comments, NHTSA-2018-0067-12032; Fiat Chrysler, Detailed 
Comments, NHTSA-2018-0067-11943; Honda, Detailed Comments, NHTSA-
2018-0067-11818.
---------------------------------------------------------------------------

    Toyota, Global Automakers, and the Auto Alliance suggested the 
incentives for light-duty trucks should amount to 20 grams/mile.\3314\ 
Global Automakers added that in addition to expanding full-size pickup 
truck hybrid incentives to light trucks, the agency should consider a 
smaller incentive for hybrid electric passenger vehicles as well.\3315\ 
The Auto Alliance and Toyota suggested a 10 grams/mile credit for 
passenger cars.\3316\ Volkswagen further requested the hybrid pickup 
credit to be expanded to all hybrid cars and trucks.\3317\
---------------------------------------------------------------------------

    \3314\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Global 
Automakers, Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, 
Detailed Comments, NHTSA-2018-0067-12073.
    \3315\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
    \3316\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Toyota, Detailed Comments, NHTSA-2018-0067-12150.
    \3317\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    Toyota, the Auto Alliance, Electric Drive Transportation 
Association, Ford, DENSO International, Global Automakers, Fiat 
Chrysler, and BorgWarner commented that having minimum production 
percentages for hybrid pickup trucks discourages manufacturers from 
investing in hybrid technologies. They requested that the agencies 
consider eliminating the percentage of production requirement and 
provide incentives in proportion to the value of the technology.\3318\ 
Ford stated that the minimum production percentages unfairly penalize 
larger manufacturers who must produce more pickup trucks to claim the 
incentives than a smaller volume manufacturer.\3319\
---------------------------------------------------------------------------

    \3318\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Auto 
Alliance, Detailed Comments, NHTSA-2018-0067-12073; Electric Drive 
Transportation Association, Detailed Comments, NHTSA-2018-0067-1201; 
Ford, Detailed Comments, NHTSA-2018-0067-11928; DENSO, Detailed 
Comments, NHTSA-2018-0067-11880; Global Automakers, Detailed 
Comments, NHTSA-2018-0067-12032; Fiat Chrysler, Detailed Comments, 
NHTSA-2018-0067-11943; BorgWarner, Detailed Comments, NHTSA-2018-
0067-11895.
    \3319\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
---------------------------------------------------------------------------

    Ariel Corp. and VNG.co, the Coalition for Renewable Natural Gas, 
NGVAmerica, the American Gas Association, and the American Public Gas 
Association commented the pickup truck incentives should be expanded to 
include natural gas vehicles.\3320\ They suggested a ``Natural Gas 
Pickup'' incentive like the hybrid-electric and performance-based 
pickup credits, but no minimum production requirement.\3321\
---------------------------------------------------------------------------

    \3320\ Joint submission from Ariel Corp. and VNG.co, Detailed 
Comments, NHTSA-2018-0067-7573; Joint submission from The Coalition 
for Renewable Natural Gas, NGVAmerica, the American Gas Association, 
and the American Public Gas Association, Detailed Comments, NHTSA-
2018-0067-11967.
    \3321\ See, e.g., Joint submission from Ariel Corp. and VNG.co, 
Detailed Comments, NHTSA-2018-0067-7573.
---------------------------------------------------------------------------

    ACEEE and UCS commented that hybrid technology has been around for 
quite a while and has been applied in every vehicle class. They 
discouraged the agencies from applying more incentives to these 
vehicles.\3322\ Specifically, UCS stated that incentives for electric 
vehicles are mostly driven by state regulation, and EPA and NHTSA 
policies are rewarding manufacturers for meeting standards they were 
already required to meet.\3323\ UCS commented that hybrids are not 
innovators or game-changing vehicles--they are simply one of many 
strategies by which manufacturers can reduce emissions and should not 
receive special treatment.\3324\
---------------------------------------------------------------------------

    \3322\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122-29; UCS, 
Detailed Comments, NHTSA-2018-0067-12039.
    \3323\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3324\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    CARB commented that incentives for full-size hybrid pickup trucks 
should remain limited in their scope and that increasing or expanding 
those incentives can erode emissions benefits.\3325\ CARB further 
commented that hybrid electric vehicles (HEVs) are widely available at 
varying levels of power and performance across vehicle sizes, and CARB 
does not believe HEVs deserve special treatment in the CO2 
vehicle regulations.
---------------------------------------------------------------------------

    \3325\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
---------------------------------------------------------------------------

    After carefully considering the comments received, EPA and NHTSA 
are not adopting any new or expanded incentives for hybrid vehicles or 
full-size pickup trucks, and are removing these incentives beginning in 
MY 2022 (the incentive for mild hybrids expires after MY 2021 
regardless, so that does not change). The agencies believe any new or 
expanded incentives would likely not result in any further emissions 
benefits or fuel economy improvements since an increase in sales volume 
would not be expected. The agencies agree with CARB and ACEEE, and UCS 
that hybrids are a well-

[[Page 25231]]

established technology that has already been applied to a wide range of 
vehicles and, as such, no further incentives are warranted at this 
time. Further, the agencies believe that incentivizing manufacturers to 
implement specific technologies is inappropriate, as manufacturer fuel 
economy performance should represent actual fuel consumption. The 
agencies believe any new or expanded incentives for hybrids would 
likely not result in any further emissions benefits or fuel economy 
improvements beyond those measured during testing; to the extent that 
manufacturers choose to build full-size pickup trucks that exceed their 
targets, those will reap the benefits of target exceedance in the 
overall fleet averaging. Manufacturers did not provide sufficient 
evidence to support their position in a manner that leads the agencies 
to conclude otherwise, and there does not appear to be any likelihood 
that manufacturers will be able to take advantage of these 
flexibilities beyond MY 2021 that makes it necessary to retain them. 
Therefore, the agencies are removing these flexibilities from the 
program starting with MY 2022.
2. Flexibilities for Air Conditioning Efficiency
    A/C systems are virtually standard automotive accessories, and more 
than 95 percent of new cars and light trucks sold in the U.S. are 
equipped with mobile A/C systems. A/C system usage places a load on an 
engine, which results in additional tailpipe CO2 emissions 
and fuel consumption; the high penetration rate of A/C systems 
throughout the light-duty vehicle fleet means that efficient systems 
can significantly impact the total energy consumed and CO2 
emissions. A/C systems also have non-CO2 emissions 
associated with refrigerant leakage.\3326\ Manufacturers can improve 
the efficiency of A/C systems though redesigned and refined A/C system 
components and controls.\3327\ That said, such improvements are not 
measurable or recognized using 2-cycle test procedures, since A/C is 
turned off during 2-cycle testing. Any A/C system efficiency 
improvements that reduce load on the engine and improve fuel economy is 
therefore not measurable on those tests.
---------------------------------------------------------------------------

    \3326\ See Section V for further details. Notably, manufacturers 
cannot claim CAFE-related benefits for reducing A/C leakage or 
switching to an A/C refrigerant with a lower global warming 
potential. While these improvements reduce GHG emissions consistent 
with the purpose of the CAA, they generally do not impact fuel 
economy and, thus, are not relevant to the CAFE program.
    \3327\ The approach for recognizing potential A/C efficiency 
gains is to utilize, in most cases, existing vehicle technology/
componentry, but with improved energy efficiency of the technology 
designs and operation. For example, most of the additional A/C-
related load on an engine is because of the compressor, which pumps 
the refrigerant around the system loop. The less the compressor 
operates, the less load the compressor places on the engine 
resulting in less fuel consumption and CO2 emissions. 
Thus, optimizing compressor operation with cabin demand using more 
sophisticated sensors, controls, and control strategies is one path 
to improving the efficiency of the A/C system. For further 
discussion of A/C efficiency technologies, see Section II.D of the 
NPRM and Chapter 6 of the accompanying PRIA.
---------------------------------------------------------------------------

    The CO2 and CAFE programs include flexibilities to 
account for the real world CO2 emissions and fuel economy 
improvements associated with improved A/C systems and to include the 
improvements for compliance.\3328\ The total of A/C efficiency credits 
is calculated by summing the individual credit values for each 
efficiency improving technology used on a vehicle, as specified in the 
A/C credit menu. The total A/C efficiency credit sum for each vehicle 
is capped at 5.0 grams/mile for cars and 7.2 grams/mile for trucks. 
Additionally, the off-cycle credit program contains credit earning 
opportunities for technologies that reduce the thermal loads on a 
vehicle from environmental conditions (solar loads or parked interior 
air temperature).\3329\ These technologies are listed on a thermal 
control menu that provides a predefined improvement value for each 
technology. If a vehicle has more than one thermal load improvement 
technology, the improvement values are added together, but subject to a 
cap of 3.0 grams/mile for cars and 4.3 grams/mile for trucks.
---------------------------------------------------------------------------

    \3328\ See 40 CFR 86.1868-12.
    \3329\ See 40 CFR 86.1869-12(b).
---------------------------------------------------------------------------

    EPA requested comment on the A/C caps and on whether A/C efficiency 
technologies and off-cycle thermal control technologies should be 
combined under a single cap, since the technologies directly interact 
with each other. That is, improved thermal control results in reduced 
A/C loads for the more efficient A/C technologies. If the thermal 
credits were removed from the off-cycle menu, they would no longer be 
counted against the 10 grams/mile menu cap discussed above, 
representing a way to provide more room under the menu cap for other 
off-cycle technologies. Specifically, EPA sought comment on replacing 
the current off-cycle thermal efficiency capped value of 10 grams/mile, 
with separate caps of 8 grams/mile for cars and 11.5 grams/mile for 
trucks.
    Comments concerning the A/C caps were received from the Auto 
Alliance, DENSO, Fiat Chrysler, and Volkswagen. DENSO commented that A/
C efficiency credits earned through the off-cycle petition process 
should not count toward the A/C credit cap. If A/C credits granted 
through the off-cycle petition process are no longer counted toward the 
A/C credit cap, it stated that manufacturers would be significantly 
incentivized to develop new and innovative technologies.\3330\ Fiat 
Chrysler requested that certain A/C credits for electrical technologies 
(i.e., A/C blower motor controls that limit wasted electrical energy) 
be transferred to the off-cycle credit list.\3331\ Volkswagen further 
supported the removal of the thermal control technology credit caps and 
suggested that implementing caps at the fleet average level, rather 
than per-vehicle, could be less constraining.\3332\ DENSO pointed to an 
NREL study which found that A/C improvements were greater than 
previously thought possible. Therefore, it requested the agencies 
consider increasing the A/C credit cap.\3333\
---------------------------------------------------------------------------

    \3330\ DENSO, Detailed Comments, NHTSA-2018-0067-11880.
    \3331\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3332\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
    \3333\ DENSO, Detailed Comments, NHTSA-2018-0067-11880.
---------------------------------------------------------------------------

    Similarly, the Auto Alliance and Fiat Chrysler suggested raising 
the cap on A/C efficiency and thermal control technology by 64 percent 
and combine them under a single cap.\3334\ Additionally, they proposed 
increasing A/C efficiency and thermal control technology credits by up 
to 64 percent.\3335\ They also proposed that the agencies create new 
regulatory provisions to handle additional new A/C and thermal 
technologies.\3336\
---------------------------------------------------------------------------

    \3334\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
    \3335\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
0067-11943.
    \3336\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
0067-12073.
---------------------------------------------------------------------------

    As with increasing the credit caps, manufacturers and suppliers 
were generally supportive of higher credit caps, or no caps at all, for 
this combined technology group. However, EPA has decided not to adopt 
any changes to the caps, including combining the A/C efficiency and 
thermal controls menu, due to the uncertainty regarding the menu credit 
values. Additional uncertainty exists for these technology groups 
because there are likely synergistic effects between A/C efficiency and 
thermal technologies that would need to be further considered in 
determining appropriate credit levels if

[[Page 25232]]

the two groups of technologies are combined under a single cap. Data is 
not currently available to consider these effects. Therefore, the 
agencies are not making any changes to the flexibilities for A/C 
efficiency improvements in the CO2 or CAFE program, but may 
perform research to understand better the relationship between A/C 
efficiency and thermal technologies for consideration in future 
rulemakings.
3. Flexibilities for Off-Cycle Technologies
    ``Off-cycle'' technologies are those that reduce vehicle fuel 
consumption and CO2 emissions in the real world, but for 
which the fuel consumption reduction benefits cannot be measured or 
cannot be fully measured under the 2-cycle test procedures (city, 
highway or correspondingly FTP, HFET) used to determine compliance with 
the fleet average standards. The CAFE city and highway test cycles, 
collectively referred to as the 2-cycle laboratory compliance tests (or 
2-cycle tests), were developed in the early 1970s. The city test 
simulates city driving in the Los Angeles area at that time. The 
highway test simulates driving on secondary roads (not expressways). 
The cycles are effective in measuring improvements in most fuel economy 
improving technologies; however, they are unable to measure or 
underrepresent certain fuel economy improving technologies because of 
limitations in the test cycles. For example, off-cycle technologies 
that improve emissions and fuel economy at idle (such as ``stop start'' 
systems) and those technologies that improve fuel economy to the 
greatest extent at expressway speeds (such as active grille shutters 
which improve aerodynamics) receive less than their real-world benefits 
in the 2-cycle compliance tests.
    Starting with MY 2008, EPA began employing a ``five-cycle'' test 
methodology to measure fuel economy for the purpose of improving new 
car window stickers (labels) and giving consumers better information 
about the fuel economy they could expect under real-world driving 
conditions.\3337\ However, for CO2 and CAFE compliance, EPA 
continues to use the established ``two-cycle'' test methodology.\3338\ 
As learned through development of the ``five-cycle'' methodology and 
prior rulemakings, there are technologies that provide real-world 
CO2 emissions and fuel consumption improvements, but those 
improvements are not fully reflected on the ``two-cycle'' test. EPA 
established the off-cycle credit program to provide an appropriate 
level of CO2 credit for technologies that achieve 
CO2 reductions, but are normally not chosen as a 
CO2 control strategy because their CO2 benefits 
are not measured on the specified 2-cycle test.
---------------------------------------------------------------------------

    \3337\ https://www.epa.gov/vehicle-and-fuel-emissions-testing/dynamometer-drive-schedules.
    \3338\ The city and highway test cycles, commonly referred to 
together as the 2-cycle tests are laboratory compliance tests 
required by law for CAFE and are also used for determining 
compliance with the GHG standards.
---------------------------------------------------------------------------

    Currently, EPA has three compliance pathways. The first approach 
allows manufacturers to gain credits without having to prove the 
benefits of the technologies on a case-by-case basis. A predetermined 
list or ``menu'' of credit values for specific off-cycle technologies 
exists and became effective starting in MY 2014.\3339\ This pathway 
allows manufacturers to use credit values established by EPA for a wide 
range of off-cycle technologies, with minimal or no data submittal or 
testing requirements.\3340\ Specifically, EPA established a menu with a 
number of technologies that have real-world CO2 and fuel 
consumption benefits not measured, or not fully measured, by the two-
cycle test procedures, and those benefits were reasonably quantified by 
the agencies at that time. For each of the pre-approved technologies on 
the menu, EPA established a quantified default value that is available 
without additional testing. Manufacturers must demonstrate that they 
were in fact using the menu technology, but not required to conduct 
testing to quantify the technology's effects, unless they wish to 
receive a credit larger than the default value. The default values for 
these off-cycle credits were largely determined from research, 
analysis, and simulations, rather than from full vehicle testing, which 
would have been both cost and time prohibitive. EPA generally used 
conservative predefined estimates to avoid any potential credit 
windfall.\3341\
---------------------------------------------------------------------------

    \3339\ See 40 CFR 86.1869-12(b).
    \3340\ The Technical Support Document (TSD) for the 2012 final 
rule for MYs 2017 and beyond provides technology examples and 
guidance with respect to the potential pathways to achieve the 
desired physical impact of a specific off-cycle technology from the 
menu and provides the foundation for the analysis justifying the 
credits provided by the menu. The expectation is that manufacturers 
will use the information in the TSD to design and implement off-
cycle technologies that meet or exceed those expectations in order 
to achieve the real-world benefits of off-cycle technologies from 
the menu.
    \3341\ While many of the assumptions made for the analysis were 
conservative, others were ``central.'' For example, in some cases, 
an average vehicle was selected on which the analysis was conducted. 
In that case, a smaller vehicle may presumably deserve fewer credits 
whereas a larger vehicle may deserve more. Where the estimates are 
central, it would be inappropriate for the agencies to grant greater 
credit for larger vehicles, since this value is already balanced by 
smaller vehicles in the fleet. The agencies take these matters into 
consideration when applications are submitted for credits beyond 
those provided on the menu.
---------------------------------------------------------------------------

    For off-cycle technologies not on the pre-defined technology list, 
or obtained through petitioning, EPA created a second pathway which 
allows manufacturers to use 5-cycle testing to demonstrate and justify 
off-cycle CO2 credits.\3342\ EPA established this 
alternative for a manufacturer to demonstrate the benefits of the 
technology using 5-cycle testing. The additional emissions tests allow 
emission benefits to be demonstrated over some elements of real-world 
driving not captured by the CO2 compliance tests, including 
high speeds, rapid accelerations, and cold temperatures. Under this 
pathway, manufacturers submit test data to EPA, and EPA determines 
whether there is sufficient technical basis to approve the off-cycle 
credits. No public comment period is required for manufacturers seeking 
credits using the EPA menu or using 5-cycle testing.
---------------------------------------------------------------------------

    \3342\ See 40 CFR 86.1869-12(c). EPA proposed a correction for 
the 5-cycle pathway in a separate technical amendments rulemaking. 
See 83 FR 49344 (Oct. 1, 2019). EPA is not approving credits based 
on the 5-cycle pathway pending the finalization of the technical 
amendments rule.
---------------------------------------------------------------------------

    The third pathway allows manufacturers to seek EPA approval, 
through a notice and comment process, to use an alternative methodology 
other than the menu or 5-cycle methodology for determining the off-
cycle technology CO2 credits.\3343\ Manufacturers must 
provide supporting data on a case-by-case basis demonstrating the 
benefits of the off-cycle technology on their vehicle models. 
Manufacturers may also use the third pathway to apply for credits and 
FCIVs for menu technologies where the manufacturer is able to 
demonstrate credits and FCIVs greater than those provided by the menu.
---------------------------------------------------------------------------

    \3343\ See 40 CFR 86.1869-12(d).
---------------------------------------------------------------------------

    Due to the uncertainties associated with combining menu 
technologies and the fact that some uncertainty is introduced because 
off-cycle credits are provided based on a general assessment of off-
cycle performance, as opposed to testing on the individual vehicle 
models, EPA established caps that limit the amount of credits a 
manufacturer may generate using the EPA menu. Off-cycle technology is 
capped at 10 grams/mile per year on a combined car and truck fleet-wide 
average basis. No caps were established for technologies gaining 
credits through the petitioning or 5-cycle approval methodologies.

[[Page 25233]]

a) Consideration of Eliminating A/C and Off-Cycle Adjustments in the 
CO2 and CAFE Programs
    The agencies sought comments in the NPRM on whether to remove the 
A/C and off-cycle flexibilities from the CAFE program and adjust the 
stringency levels accordingly based upon concern that the flexibilities 
might distort the market. Several commenters provided responses 
concerning the feasibility of removing any of these flexibilities. 
Commenters included the Auto Alliance, the National Automobile Dealers 
Association, Global Automakers, the Alliance for Vehicle Efficiency, 
ACEEE, BorgWarner, Fiat Chrysler, General Motors, International Council 
on Clean Transportation, Toyota, and UCS. Other comments were received 
requesting that the agencies look into expanding the flexibilities by 
including more technologies.
    There was widespread support from commenters for retaining these 
flexibilities for A/C and off-cycle technologies in the CO2 
and CAFE programs. Commenters preferred that the agencies continue to 
include the flexibilities, believing them to enable real world fuel 
economy improvements and compliance with CO2 and CAFE 
standards with a more cost effective combination of technologies. The 
agencies agree that these programs achieve real world fuel economy 
improvements and that keeping the flexibilities may enable more cost 
effective technology combinations to achieve those real world fuel 
economy improvements. For MY 2017, manufacturers introduced a wide 
variety of low-cost technologies through the A/C and off-cycle 
flexibilities that increased the overall industry's CAFE performance by 
1.1 mpg. The agencies also acknowledge that the continued use of these 
flexibilities under the EPA program since 2012 warrants consideration 
due to automakers' and suppliers' significant investments in developing 
the technologies, which could result in stranded capital should the 
agencies discontinue them and manufacturers choose to remove the 
technologies. For these reasons, the agencies have decided to continue 
allowing manufacturers to use the existing flexibilities for A/C 
efficiency and off-cycle technologies for future model years.
b) Final Decisions in Response to Manufacturers' and Suppliers' 
Requests
    Automakers, trade associations, and auto suppliers recommended 
several changes to the current off-cycle credit program.\3344\ Prior to 
the NPRM, automakers and suppliers suggested changes to the off-cycle 
program, including:
---------------------------------------------------------------------------

    \3344\ See generally Alliance/Global Petition.
---------------------------------------------------------------------------

     Streamlining the program in ways that would give auto 
manufacturers more certainty and make it easier for manufacturers to 
earn credits;
     Expanding the current pre-defined off-cycle credit menu to 
include additional technologies and increasing credit levels where 
appropriate;
     Eliminating or increasing the credit cap on the pre-
defined list of off-cycle technologies and revising the thermal 
technology credit cap; and
     Creating a role for suppliers directly to seek approval of 
their technologies.
    EPA requested comments on several aspects of the off-cycle credits 
program and, as discussed below, both EPA and NHTSA are adopting some 
modest changes, primarily to help streamline and clarify their 
programs, and to ease the implementation burden for manufacturers and 
the government. The agencies are not adopting a significant expansion 
of the programs in this rule, as also discussed below. EPA and NHTSA 
are taking this relatively conservative approach for their off-cycle 
programs due to the uncertainty that remains in estimating off-cycle 
benefits of technologies and the need to remain cautious to help ensure 
that emissions and fuel economy benefits expected through the off-cycle 
flexibility are realized in the real-world.
(1) Program Streamlining
    EPA requested comments on changes to the off-cycle process that 
would streamline the program. Currently, under the third pathway, 
manufacturers submit an application that includes the methodology they 
used to determine the off-cycle credit value and data, which then 
undergoes a public notice and comment process prior to an EPA decision 
regarding the application. Each manufacturer separately submits an 
application to EPA that must undergo a public notice and comment 
process even if the manufacturer uses a methodology previously approved 
by EPA for another manufacturer. For example, under the current 
program, multiple manufacturers have separately submitted applications 
for high-efficiency alternators and advanced A/C compressors using 
similar methodologies and producing similar levels of credits. If 
manufacturers also seek fuel economy improvement values for the CAFE 
program, they are also required to send the submissions to NHTSA, as 
EPA consults with NHTSA in its determinations for the CAFE program. 
NHTSA's involvement is discussed in more detail in Section IX.D.3.b).
    EPA requested comment on revising the regulations to allow all auto 
manufacturers to make use of a methodology once it has been approved by 
EPA under the public process, without subsequent applications from 
other manufacturers having to undergo the same process. This would 
reduce redundancy in the current program. Manufacturers would need to 
provide EPA with at least the same level of data and detail for the 
technology and methodology as the manufacturer that went through the 
initial public notice and comment process.
    EPA received supportive comments for streamlining the approval 
process from auto manufacturers and suppliers. The Auto Alliance 
commented that it supports all actions that would shorten the time it 
takes EPA to evaluate and reach decisions on applications through the 
off-cycle alternative methodology pathway, and that manufacturers 
should be allowed to use common data from applications that have 
already been approved.\3345\ Such common data would include ambient 
conditions, general consumer behavior data, and general operating and 
performance data for the same off-cycle technologies. Global Automakers 
also commented that EPA should streamline efforts to avoid 
reduplication of applications in situations where multiple automakers 
have submitted petitions for the same technology and recommended 
blanket approval for applications using the same specific technologies 
and calculation and measurement procedures.\3346\ General Motors 
commented that when a credit for a new technology is approved for one 
manufacturer, the EPA decision document announcing that approval can 
serve as a guidance document that assigns a credit value or calculation 
methodology for the technology for all manufacturers without requiring 
duplicative testing.\3347\ MEMA commented that it would be sufficient 
to uphold the integrity of the off-cycle program to require the next 
vehicle manufacturer's application to provide at least the same level 
of data and details as the original vehicle manufacturer application 
and to validate the level of credit the next vehicle manufacturer is

[[Page 25234]]

applying for based on how the technology is applied in its fleet.\3348\
---------------------------------------------------------------------------

    \3345\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3346\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
    \3347\ General Motors, Detailed Comments, NHTSA-2018-0067-11858-
21.
    \3348\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

    ACEEE commented that any streamlining of the process by which 
automakers petition for off-cycle credits must maintain the requirement 
that a thorough methodology show real-world benefits and ensure 
adequate opportunity for public review.\3349\ International Council on 
Clean Transportation (ICCT), while not commenting on this specific 
request for comment, commented that the program should remain unchanged 
until potential changes can be further analyzed.\3350\
---------------------------------------------------------------------------

    \3349\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
    \3350\ International Council on Clean Transportation, Detailed 
Comments, NHTSA-2018-0067-11741.
---------------------------------------------------------------------------

    After considering the comments, consistent with its request for 
comment, EPA is streamlining the approval process as follows: Once a 
methodology for a specific off-cycle technology has gone through the 
public notice and comment process and is approved for one manufacturer, 
other manufacturers may follow the same methodology to collect data on 
which to base their off-cycle credits. Once a methodology is approved, 
other manufacturers may submit applications citing the approved 
methodology, but those manufacturers must provide their own necessary 
test data, modeling, and calculations of credit value specific to their 
vehicles, and any other vehicle-specific details pursuant to that 
methodology, to assess an appropriate credit value. This is similar to 
what occurred, for example, with the advanced A/C compressor, where one 
manufacturer applied for credits with data collected through bench 
testing and vehicle testing and subsequent manufacturers applied for 
credits following the same methodology, but by submitting test data 
specific to their vehicle models. However, those subsequent 
applications previously required a public notice and comment process. 
For future applications, as long as the testing is conducted using the 
previously-approved methodology, EPA will evaluate the credit 
application and issue a decision with no additional notice and comment, 
since the first application that established the methodology was 
subject to notice and comment.
    EPA is not providing blanket approval for a specific credit value, 
nor amending the requirement that manufacturers collect necessary data 
or perform modeling or other analyses on their specific vehicle models 
as the basis for the credit. However, once a methodology has been fully 
vetted and approved through the public process, EPA believes additional 
public review of the identical methodology is unnecessarily 
duplicative. In EPA's experience thus far (for example with high-
efficiency alternators and advanced A/C compressors for which EPA has 
received applications from several manufacturers based on the same 
methodology), additional public review has yielded no additional 
substantive public comments. EPA believes this change in the program 
will help reduce the time necessary for review of applications. EPA 
will maintain the option to seek additional public comment in cases 
where the agency believes a new application deviates from a previously 
approved methodology or raises new issues on which the agency believes 
it is prudent to seek comment.
    EPA also requested comment on revising the regulations to allow EPA 
to, in effect, add technologies to the pre-approved credit menu without 
going through a subsequent rulemaking. For example, if one or more 
manufacturers submit applications with sufficient supporting data for 
the same or similar technology, the data from that application(s) could 
potentially be used by EPA as the basis for adding technologies to the 
menu. EPA requested comment on revising the regulations to allow EPA to 
establish through a decision document a credit value, or scalable value 
as appropriate, and technology definitions or other criteria to be used 
for determining whether a technology qualifies for the new menu credit. 
As envisioned in the NPRM, this streamlined process of adding a 
technology to the menu would involve an opportunity for public review 
but not a formal rulemaking to revise the regulations, allowing EPA to 
add technologies to the menu in a timely manner, where EPA believes 
that sufficient data exist to estimate an appropriate credit level for 
that technology across the fleet.
    EPA received supportive comments regarding this request for 
comments from auto manufacturers and suppliers who believe that the 
change would help streamline the program. EPA also received comments 
from environmental NGOs suggesting that the program should not be 
changed at this time. After consideration of these comments, the 
agencies are not revising the regulations to allow technologies to be 
added to the menu without a rulemaking because EPA believes that menu-
based off-cycle credits should be based on a robust demonstration of 
the technology, consistent with the regulations. The agencies will 
retain the option to add technologies to the menu through a rulemaking, 
similar to the approach being taken for high-efficiency alternators and 
advanced A/C compressors as discussed below, where sufficient data has 
been collected from multiple manufacturers and vehicle models on which 
to base a menu credit. The menu credits are meant to be conservative. 
The agencies are concerned that basing a menu credit on data from only 
one or a few manufacturers does not guarantee a robust and accurate 
credit level representing vehicles across the fleet. At this time, the 
agencies continue to believe a rulemaking process with full opportunity 
for public comment remains the best approach for adding technologies to 
the menu. A rulemaking ensures that all stakeholders including 
automakers have an opportunity to provide data to support an 
appropriate and conservative credit level for the fleet. This approach 
also provides an incentive for manufacturers to, in the meantime, 
continue to perform testing and provide actual data that could 
eventually be used to inform a rulemaking process to add a technology 
to the menu. The agencies want to preserve that element of the program 
to maintain the integrity of off-cycle credits representing real-world 
reductions.
(2) A/C and Off-Cycle Application Process
    The agencies received several comments, in addition to those 
received in the petitions from the Auto Alliance and Global Automakers, 
discussed below, on the application process for approving additional A/
C and off-cycle credits. Commenters included the Global Automakers, the 
Auto Alliance, Volkswagen, Edison Electric Institute, Ford, Fiat 
Chrysler, NCAT, Toyota, General Motors, and DENSO International.
    Fiat Chrysler, Ford, Volkswagen, DENSO International, Global 
Automakers, and the Auto Alliance requested that the agencies respond 
more quickly to applications for A/C and off-cycle technologies.\3351\ 
They

[[Page 25235]]

prefer that petitions be addressed before the close of a model year so 
manufacturers can have a better idea of what credits they will earn.
---------------------------------------------------------------------------

    \3351\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943-
50; Ford, Detailed Comments, NHTSA-2018-0067-11928-15; Volkswagen, 
Detailed Comments, NHTSA-2017-0069-0583-13; DENSO, Detailed 
Comments, NHTSA-2018-0067-11880-5; Global Automakers, Detailed 
Comments, NHTSA-2018-0067-12032-50; Auto Alliance, Detailed 
Comments, NHTSA-2018-0067-12073-120.
---------------------------------------------------------------------------

    The agencies agree that responding to petitions before the end of a 
model year is beneficial to manufacturers and the government. 
Manufacturers would have a better idea of the approved credits, and the 
government could carry-out its compliance processes more efficiently. 
EPA structured the A/C and off-cycle programs to make it possible to 
complete the processes by the end of the model year so manufacturers 
could submit their final reports within the required deadline, 90 days 
after the calendar year. However, delays currently exist due to the 
timing needed to review and approve technologies for the first time and 
issue Federal Register notices seeking public comments, where 
applicable. The agencies anticipate these problems will resolve 
themselves as the off-cycle program reaches maturity and EPA initiates 
the new streamlining approaches adopted in this final rule, discussed 
in the previous section.
    The agencies are also aware that delays exist because manufacturers 
frequently submit late applications, new applications, and ask for 
retroactive credits or FCIVs for off-cycle technologies equipped on 
previously-manufactured vehicles after the model year has ended. As 
required under both the CO2 and CAFE programs, manufacturers 
are to submit applications for off-cycle credits and FCIVs before the 
beginning of each compliance model year, to enable the agencies to make 
better informed final decisions before the model year ends.
    To expedite the process of approvals, the agencies will enforce 
existing EPA and NHTSA regulations requiring manufacturers to notify 
and report information on the technologies before the beginning of the 
model year. Presently, manufacturers must notify EPA in their pre-model 
year reports, and in their applications for certification, of their 
intention to generate any A/C and off-cycle credits before the model 
year, regardless of the methodology for generating credits.\3352\ 
Manufacturers choosing to generate credits using the alternative EPA-
approval methodology are required to submit a detailed analytical plan 
to EPA prior to a model year in which a manufacturer intends to seek 
these credits. The manufacturer may seek EPA input on the proposed 
methodology prior to conducting testing or analytical work, and EPA 
will provide input on the manufacturer's analytical plan. The 
alternative demonstration program must be approved in advance by the 
Administrator. NHTSA has similar provisions for its projections reports 
in which detailed information on the technologies must be included in 
those submissions during the month of December before the model 
year.\3353\ NHTSA's provisions also require manufacturers to submit 
information to NHTSA at the same time as to EPA. Consequently, the 
eligibility of a manufacturer to gain off-cycle CO2 credits 
or CAFE adjustments for a given compliance model year requires 
appropriate submissions to the agencies. The agencies intend to enforce 
these provisions starting with the 2020 compliance model year. 
Manufacturers may resubmit MY 2020 information until May 1, 2020. After 
that time, the agencies will deny any manufacturers' late submissions 
requesting retroactive credits. However, manufacturers who properly 
submit information ahead of time will be allowed to make corrections to 
resolve inadvertent errors during or after the model year. The agencies 
believe that enforcing the existing submission requirements will be the 
most efficient approach to expedite approvals until new regulatory 
deadlines or additional requirements can be adopted.
---------------------------------------------------------------------------

    \3352\ See 40 CFR 86.1869(a) and 40 CFR 1843-01.
    \3353\ See 49 CFR part 537.7(c)(7) and 49 CFR part 531.6 and 
533.6.
---------------------------------------------------------------------------

    Fiat Chrysler, Volkswagen, Global Automakers, and the Auto Alliance 
further suggested the EPA issue a Federal Register notice for submitted 
off-cycle applications within 30 days and issue a final decision within 
90 days.\3354\
---------------------------------------------------------------------------

    \3354\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; 
Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; Global 
Automakers, Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, 
Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    As mentioned, EPA is addressing the issues raised by commenters by 
streamlining its required regulatory processes to eliminate the need to 
submit multiple Federal Register notices concerning requests from 
different manufacturers for the same technology. Under this streamlined 
process, after a technology is approved for the initial 
manufacturer(s), EPA will approve any subsequent manufacturer requests 
for the same technology upon receipt of data submissions validating the 
benefit specific to their model types.
    General Motors, Toyota, NCAT, Fiat Chrysler, Ford, Volkswagen, 
DENSO, Edison Electric Institute, Global Automakers, and the Auto 
Alliance further suggested that technologies approved for multiple 
manufacturers, to the extent additional automakers will have the same 
requests, be added to the menu to encourage additional implementation 
of the technology. Doing so would reduce duplicative efforts for the 
agencies, as well as manufacturers.\3355\
---------------------------------------------------------------------------

    \3355\ General Motors, Detailed Comments, NHTSA-2018-0067-11858; 
Toyota, Detailed Comments, NHTSA-2018-0067-12150; NCAT, Detailed 
Comments, NHTSA-2018-0067-11969; Fiat Chrysler, Detailed Comments, 
NHTSA-2018-0067-11943; Ford, Detailed Comments, NHTSA-2018-0067-
11928; Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; DENSO, 
Detailed Comments, NHTSA-2018-0067-11880; Edison Electric Institute, 
Detailed Comments, NHTSA-2018-0067-11918; Global Automakers, 
Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed 
Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    As mentioned previously, the agencies have decided to allow only 
new technologies to be added to the menu through the regular rulemaking 
processes including the opportunity for notice and public comment.
    General Motors, DENSO, Global Automakers, and the Auto Alliance 
further suggested that suppliers should be allowed to request a ``grams 
per mile'' value for their off-cycle technologies. They asserted that 
this will provide certainty to manufacturers before they buy that 
technology.\3356\ Toyota and the Auto Alliance suggested that the 
agencies could improve efficiency and reduce burdens by creating a 
``toolbox,'' methodologies that manufacturers can apply to the analysis 
of off-cycle credit opportunities.\3357\ They stated it would 
additionally help manufacturers if the agency would issue guidance 
letters and decision documents for off-cycle credit approvals.\3358\
---------------------------------------------------------------------------

    \3356\ General Motors, Detailed Comments, NHTSA-2018-0067-11858; 
DENSO, Detailed Comments, NHTSA-2018-0067-11880; Global Automakers, 
Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed 
Comments, NHTSA-2018-0067-12073.
    \3357\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Auto 
Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3358\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    The agencies believe that developing a ``toolbox'' may not be 
possible due to the development of new and emerging technologies, and 
manufacturers' different approaches for evaluating the benefits of the 
technologies. The agencies may consider additional guidance, if 
feasible, as the programs further matures in the approval process of 
technologies and if the agencies can identify consistent methodologies 
that may help manufacturers analyze off-cycle technologies.

[[Page 25236]]

    NCAT and General Motors requested more transparency in the A/C and 
off-cycle approval process. They suggested that the agencies could 
provide reports including off-cycle credits approved by vehicle make 
and model and provide further clarification of data requirements that 
influenced the decision process.\3359\
---------------------------------------------------------------------------

    \3359\ NCAT, Detailed Comments, NHTSA-2018-0067-11969; General 
Motors, Detailed Comments, NHTSA-2018-0067-11858.
---------------------------------------------------------------------------

    EPA and NHTSA have separate approaches for sharing information on 
these flexibilities, to provide public transparency. EPA already 
provides detailed information on manufacturers generation of A/C and 
off-cycle credits for each model year in its end of the year compliance 
report, including the magnitude of credits by manufacturer and by 
credit type, the credits generated by technology type, and the 
penetration of off-cycle technologies in each manufacturer's 
fleet.\3360\ NHTSA plans to share similar information on its PIC and to 
provide projected data on the market penetration rates of the 
technologies as soon as it starts receiving information through its new 
reporting templates for the 2023 compliance model year.
---------------------------------------------------------------------------

    \3360\ ``The 2018 EPA Automotive Trends Report: Greenhouse Gas 
Emissions, Fuel Economy, and Technology since 1975,'' EPA-420-R-19-
002. March 2019; Figures 5.8 through 5.12, and Tables 5.3 and 5.4.
---------------------------------------------------------------------------

(3) High Efficiency Alternators and Advanced Air Conditioning (A/C) 
Compressors
    EPA sought comments on modifying the off-cycle menu to add certain 
technologies for which EPA has collected sufficient data to set an 
appropriate credit level. More specifically, EPA received data from 
multiple manufacturers on high-efficiency alternators and advanced air 
conditioning (A/C) compressors that could serve as the basis for new 
menu credits for these technologies.\3361\ EPA requested comments on 
adding these two technologies to the menu including comments on credit 
level and appropriate definitions. EPA also requested comments on other 
off-cycle technologies that EPA could consider adding to the menu 
including supporting data that could serve as the basis for the credit.
---------------------------------------------------------------------------

    \3361\ https://www.epa.gov/vehicle-and-engine-certification/compliance-information-light-duty-greenhouse-gas-ghg-standards.
---------------------------------------------------------------------------

    EPA received only supportive comments on its specific request for 
comments regarding adding high efficiency alternators and advanced A/C 
compressors to the menu. Toyota, General Motors, BorgWarner, Fiat 
Chrysler, the Auto Alliance, Global Automakers, MECA, DENSO, SAFE, and 
Volkswagen submitted responses on the off-cycle menu. General Motors, 
Volkswagen, Fiat Chrysler, Global Automakers, and the Auto Alliance all 
supported adding high-efficiency alternators and advanced A/C 
compressors to the menu.\3362\ They commented that these technologies 
have already been approved for off-cycle credits through the petition 
process multiple times. They contend that it would be less burdensome 
if the technologies would be added to the pre-approved off-cycle credit 
list. That said, they were concerned about being constrained by the 
off-cycle caps.\3363\
---------------------------------------------------------------------------

    \3362\ General Motors, Detailed Comments, NHTSA-2018-0067-11858; 
Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; Fiat Chrysler, 
Detailed Comments, NHTSA-2018-0067-11943; Global Automakers, 
Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed 
Comments, NHTSA-2018-0067-12073.
    \3363\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
0067-11858.
---------------------------------------------------------------------------

    The agencies believe that adding high-efficiency alternators and 
advanced A/C compressors to the menu is a reasonable step to help 
streamline the program by allowing manufacturers to select the menu 
credit rather than continuing to seek credits through the public 
approval process. Therefore, EPA is revising the regulations to add 
these two technologies to the menus. The high-efficiency alternator is 
being added to the off-cycle credits menu, and the advanced A/C 
compressor with a variable crankcase valve is being added to the menu 
for A/C efficiency credits. The credit levels are based on data 
previously submitted by multiple manufacturers through the off-cycle 
credits application process, and discussed in the NPRM. The high 
efficiency alternator credit is scalable with efficiency, providing an 
increasing credit value of 0.16 grams/mile CO2 per percent 
improvement as the efficiency of the alternator increases above a 
baseline level of 67 percent efficiency. The advanced A/C compressor 
credit value is 1.1 grams/mile for both cars and light trucks.\3364\
---------------------------------------------------------------------------

    \3364\ For additional details regarding the derivation of these 
credits see EPA's Memorandum to Docket EPA-HQ-OAR-2018-0283 
(``Potential Off-cycle Menu Credit Levels and Definitions for High 
Efficiency Alternators and Advanced Air Conditioning Compressors'').
---------------------------------------------------------------------------

    EPA also received comments from the Auto Alliance, Fiat Chrysler, 
General Motors, Mitsubishi, Gentherm, ITB, and MEMA on a variety of 
individual technologies that they suggest adding to the menu.\3365\ 
These commenters provided little data to support their recommended 
credit levels. The Auto Alliance and Alliance for Vehicle Efficiency 
further asserted that flexibility mechanisms are increasingly important 
and there is a need to develop unconventional and non-traditional fuel 
economy technologies to meet standards.\3366\ They requested additional 
pre-defined and pre-approved technologies to be included in this 
regulation.\3367\
---------------------------------------------------------------------------

    \3365\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073-
48; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; General 
Motors, Detailed Comments, NHTSA-2018-0067-11858; Mitsubishi, 
Detailed Comments, NHTSA-2018-0067-12056; MEMA, Detailed Comments, 
MEMA, EPA-HQ-OAR-2018-0283-5692 (See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20 
Appendices%20Oct%2026%202018.pdf); ITB, Detailed Comments, EPA-HQ-
OAR-2018-0283-5469; Gentherm, Detailed Comments, EPA-HQ-OAR-2018-
0283-5058.
    \3366\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Alliance for Vehicle Efficiency, Detailed Comments, NHTSA-2018-0067-
11696.
    \3367\ NHTSA-2018-0067-12073-48.
---------------------------------------------------------------------------

    The agencies have reviewed manufacturers' requests for adding 
additional technologies to the picklist and concluded that there is 
insufficient data in the record at this time on which to base an 
appropriate menu credit value for the technologies. Therefore, none of 
these technologies are being added to the menu at this time. Given the 
limited data and uncertainty, EPA also does not believe it would be 
appropriate to add any of the technologies to the menu without an 
opportunity for public review and comment. Although the agencies are 
not adding these technologies to the menu at this time, manufacturers 
may seek off-cycle credits for these technologies through the other 
program pathways.
(4) Stop-Start Technology
    In 2014, EPA approved additional credits for the Mercedes-Benz's 
stop-start system through the off-cycle credit process based on data 
submitted by Mercedes-Benz on fleet idle time and its system's real-
world effectiveness (i.e., how much of the time the system turns off 
the engine when the vehicle is stopped).\3368\ Prior to proposal, 
multiple auto manufacturers requested that EPA revise the table menu 
value for stop-start technology based solely on one input value EPA 
considered, idle time, in the context of the Mercedes-Benz stop-start 
system. No manufacturers provided additional data on any of the other 
factors evaluated during consideration of a conservative credit value 
for stop-start systems. Stop-start systems vary significantly in 
hardware,

[[Page 25237]]

design, and calibration, leading to wide variations in the amount of 
idle time during which the engine is actually turned off in real-world 
driving. EPA has learned that some stop-start systems may be less 
effective in the real-world than the agency estimated in its 2012 
rulemaking analysis, for example, due to systems having a disable 
switch available to the driver, or because stop-start systems can be 
disabled under certain temperature conditions or auxiliary loads, which 
would offset the benefits of the higher idle time estimates. EPA 
requested additional data from manufacturers, suppliers, and other 
stakeholders regarding a comprehensive update to the stop-start off-
cycle credit table value. EPA did not receive any additional real-world 
system effectiveness data from commenters on which to base an adjusted 
credit level. MEMA commented that EPA should base an increase in the 
credit on the agencies' updated estimated effectiveness of stop-start 
technology in the Draft Technical Assessment Report (TAR), which shows 
a 67 percent increase in effectiveness.\3369 3370\ However, EPA notes 
that this estimate is for system effectiveness over the 2-cycle test 
procedures and, therefore, is not an appropriate basis to adjust the 
off-cycle credits. The agencies are not adjusting the menu credits for 
stop-start systems at this time. Manufacturers may apply for additional 
credits if they are able to collect data demonstrating a system 
effectiveness that would serve as the basis for those credits.
---------------------------------------------------------------------------

    \3368\ ``EPA Decision Document: Mercedes-Benz Off-cycle Credits 
for MY 2012-2016,'' EPA-420-R-14-025 (Sept. 2014).
    \3369\ Draft Technical Assessment Report: Midterm Evaluation of 
Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate 
Average Fuel Economy Standards for Model Years 2022-2025, EPA-420-D-
16-900 (July 2016).
    \3370\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

(5) Menu Credit Cap
    The off-cycle menu currently includes a fleetwide cap on credits of 
10 grams/mile to address the uncertainty surrounding the data and 
analysis used as the basis of the menu credits.\3371\ Prior to 
proposal, some stakeholders expressed concern that the current cap may 
constrain manufacturers' future ability to fully utilize the menu 
especially if the menu is expanded to include additional technologies, 
as described above. For example, Global Automakers suggested raising 
the cap from 10 grams/mile to 15 grams/mile.\3372\ EPA requested 
comments on increasing the current cap, for example, from the current 
10 grams/mile to 15 grams/mile to accommodate increased use of the 
menu. EPA also requested comment on a concept that would replace the 
current menu cap with an individual manufacturer cap that would scale 
with the manufacturer's average fleetwide target levels. The cap would 
be based on a percentage of the manufacturer's fleetwide 2-cycle 
emissions performance, for example at five to ten percent of 
CO2 of a manufacturer's emissions fleet-wide target. With a 
cap of five percent for a manufacturer with a 2-cycle fleetwide average 
CO2 level of 200 grams/mile, for example, the cap would be 
10 grams/mile.
---------------------------------------------------------------------------

    \3371\ 40 CFR 86.1869-12(b)(2).
    \3372\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
---------------------------------------------------------------------------

    There was widespread support from automakers and suppliers for 
removing the cap entirely or raising the cap from 10 grams/mile to 15-
20 grams/mile. Toyota, General Motors, BorgWarner, Fiat Chrysler, the 
Auto Alliance, Global Automakers, MECA, DENSO, SAFE, and Volkswagen 
submitted responses on the off-cycle cap to EPA.\3373\ They argued that 
the 2-cycle test does not always account for all the benefits a 
technology provides.\3374\ General Motors, Fiat Chrysler, the Auto 
Alliance, Global Automakers, and Volkswagen agreed that EPA should 
remove the 10 grams/mile cap and, if they must keep the cap, increasing 
it to 15 grams/mile.\3375\
---------------------------------------------------------------------------

    \3373\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; General 
Motors, Detailed Comments, NHTSA-2018-0067-11858; BorgWarner, 
Detailed Comments, NHTSA-2018-0067-11895; Fiat Chrysler, Detailed 
Comments, NHTSA-2018-0067-11943; Auto Alliance, Detailed Comments, 
NHTSA-2018-0067-12073; Global Automakers, Detailed Comments, NHTSA-
2018-0067-12032; MECA, Detailed Comments, NHTSA-2018-0067-11994; 
DENSO, Detailed Comments, NHTSA-2018-0067-11880; SAFE, Detailed 
Comments, NHTSA-2018-0067-11981; Volkswagen, Detailed Comments, 
NHTSA-2017-0069-0583.
    \3374\ See, e.g., DENSO, Detailed Comments, NHTSA-2018-0067-
11880.
    \3375\ General Motors, Detailed Comments, NHTSA-2018-0067-11858; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Auto 
Alliance, Detailed Comments, NHTSA-2018-0067-12073; Global 
Automakers, Detailed Comments, NHTSA-2018-0067-12032; Volkswagen, 
Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    Global Automakers commented that, as more technology receives off-
cycle credit values, the cap will restrict innovation and therefore EPA 
should lift the cap now in anticipation of increased use of 
technologies.\3376\ General Motors similarly commented that the cap was 
an arbitrary limit without any technical justification and that, if the 
agency was to add emission reduction technologies to the menu these 
devices could not be effectively incentivized if the 10 grams/mile cap 
remains in place, since there would be no room under the cap.\3377\ 
General Motors suggested that as the program continues, manufacturers 
will continue to find new technologies and will be limited by the cap. 
They stated that the cap will stifle additional investments for 
technologies. MEMA commented that if EPA expands the off-cycle 
technologies menu and continually adds off-cycle technologies to the 
menu, it is critical that EPA increase or eliminate the cap on the 
credits gained from the off-cycle menu.\3378\
---------------------------------------------------------------------------

    \3376\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
12032.
    \3377\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
    \3378\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

    The Auto Alliance argued that putting caps on emerging new 
technologies will hinder further vehicle investments and improvements. 
The planning cycle is implemented years out and without a guarantee 
they will see benefits, the Auto Alliance stated that manufacturers 
lack incentivization to work toward large technological advances.\3379\ 
The Auto Alliance and Alliance for Vehicle Efficiency further asserted 
that flexibility mechanisms are increasingly important and there is a 
need to develop unconventional and non-traditional fuel economy 
technologies.\3380\
---------------------------------------------------------------------------

    \3379\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3380\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Alliance for Vehicle Efficiency, Detailed Comments, NHTSA-2018-0067-
11696.
---------------------------------------------------------------------------

    ACEEE commented that the off-cycle credit menu cap should not be 
increased or modified without the agency first defining any other 
changes it might consider making to the off-cycle credit program and 
this should be done through a separate NPRM and public review 
process.\3381\ ICCT commented that if the agencies allow more use of 
off-cycle credits without clear validation of their real-world 
benefits, the regulations cannot serve their intended objectives to 
reduce CO2 and fuel use.\3382\
---------------------------------------------------------------------------

    \3381\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
    \3382\ ICCT, Detailed Comments, NHTSA-2018-0067-11741-43.
---------------------------------------------------------------------------

    EPA also received a few comments warning about the risks of 
removing the caps and over incentivizing the CAFE and CO2 programs. 
ACEEE pointed out that while expanding and updating the flexibilities 
that incentivize innovation

[[Page 25238]]

and research is a great method to increase fuel efficiency, it is 
important to put a time limit on those incentives and carefully design 
them so manufacturers do not take advantage. ACEEE argued that, if 
these flexibilities are not implemented thoughtfully, they can end up 
reducing the program benefits. UCS commented that, given the potential 
interaction from multiple incentives, it is important to consider the 
combined impacts of flexibilities on the overall stringency of the 
regulation. UCS stated that given the potential for widespread harm, 
credits within the program should be severely limited, and the 
agencies' assessment of the impacts of such incentives should be 
extremely conservative in order to promote increased environmental 
benefits of the fuel economy and carbon dioxide emissions 
standards.\3383\
---------------------------------------------------------------------------

    \3383\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
---------------------------------------------------------------------------

    The agencies are not increasing the 10 grams/mile menu credit cap 
at this time. EPA established the 10 grams/mile credit cap to address 
the uncertainty surrounding the data and analysis used as the basis of 
the menu credits, and agrees with ACEEE, ICCT, and UCS that sufficient 
uncertainty remains such that increasing the current cap is not 
justified. As noted in the 2012 final rule, EPA included the fleet-wide 
cap because the default credit values were based on limited data, and 
also because the agencies recognized that some uncertainty is 
introduced when credits are provided based on a general assessment of 
off-cycle performance as opposed to testing on the individual vehicle 
models.\3384\ That uncertainty has not significantly diminished since 
the 2012 final rule. Also, over the course of implementing the program, 
EPA has encountered issues with the regulatory definitions currently in 
place for some technologies. The regulations specify that manufacturers 
may claim credits for technologies that meet the regulatory 
definitions. However, there have been instances where manufacturers 
have claimed credits for a technological approach that they have argued 
meets the regulatory definition, but EPA found that the technology was 
not implemented consistent with the technological approach envisioned 
when the off-cycle program was established. This has raised questions 
of whether the credits for the technological approach in question truly 
represent real-world reductions, and whether the credits should 
ultimately be allowed. These types of issues have resulted in 
uncertainty, which can lead to delays in credit calculations, 
competitive inequities, as well as increased burden on the agency to 
review and resolve issues. The caps continue to serve as an important 
measure against the loss of emissions reductions and fuel savings given 
the uncertainty in the credit values as the program is implemented. 
Since the agencies are not expanding the menu beyond the two 
technologies discussed above, the agencies believe there remains enough 
room under the cap such that the menu may continue to serve its purpose 
as a source of off-cycle credits. Although a few manufacturers 
approached the cap limit in MY 2018, the fleet average menu credit was 
4.7 grams/mile, less than half the cap value.\3385\ If the agencies 
undertake a rulemaking in the future to modify the menu or regulatory 
definitions, the agencies may re-evaluate the cap levels at that time. 
The agencies note that the cap only applies to credits based on the 
menu. Under the current program, manufacturers may apply for credits 
beyond the cap through other available pathways based on a 
demonstration of off-cycle technology emission reduction data for their 
fleets.
---------------------------------------------------------------------------

    \3384\ 77 FR 62834 (Oct. 15, 2012).
    \3385\ The 2018 EPA Automotive Trends Report, Greenhouse Gas 
Emissions, Fuel Economy, and Technology since 1975, EPA-420-R-19-002 
(Mar. 2019).
---------------------------------------------------------------------------

    As noted above, the agencies have decided to continue the option to 
add technologies to the menu only through the rulemaking process and, 
for this final rule, have decide to add two new menu items; one for 
high-efficiency alternators and another for advanced A/C compressors. 
The agencies stated that they will only add technologies when 
sufficient data has been collected from multiple manufacturers and 
vehicle models on which to base a menu credit. Accordingly, the 
agencies believe this approach ensures that conservative, robust and 
accurate credit levels are being added representing vehicles ``on 
average'' across the fleet.
    Finally, NHTSA has been studying how the combination of 
flexibilities and incentives may adversely affect the stringency of the 
CAFE regulations. NHTSA is aware of an instance in which combining 
incentives for alternative fueled vehicles and adjustments for A/C and 
off-cycle technologies allowed one manufacturer to increase in CAFE 
fleet performance to a combined average of 516.8 mpg for MY 2017, a 
curious result. NHTSA iscontinuing to evaluate the issue of combining 
incentives and flexibilities and may address this issue further in the 
future.
(6) Eligibility
    Though, in the NPRM, EPA did not explicitly request comment on the 
eligibility criteria for determining what technologies are eligible for 
off-cycle credits, EPA received comments on this topic. UCS commented 
that regulations should be clarified so that the program does not 
result in unwarranted credits for baseline technologies, noting that in 
the 2012 final rule EPA stated that technologies integral or inherent 
to the basic vehicle design were not eligible for credits and 
specifically excluded technologies identified by the agency as 
technologies a manufacturer may use to meet the two-cycle 
CO2 standards.\3386\ ACEEE commented that off-cycle credits 
should be limited to new and innovative technologies and, that to be 
eligible for credit, a technology must reduce emissions from the 
vehicle receiving the credit (as opposed to other vehicles on the road, 
for example, through system effects of technologies designed for crash 
avoidance or improving traffic flow).\3387\ The Auto Alliance also 
commented in the area of eligibility, suggesting regulatory changes 
that would allow off-cycle credits for any technology where the 
manufacturer could demonstrate an off-cycle emissions benefit.\3388\ 
The Auto Alliance commented that the program is intended to provide 
credit for technologies that provide more fuel economy and 
CO2 emissions reduction benefit in the real-world than is 
realized in FTP and HFET on-cycle testing and that a baseline 
technology should be eligible for such credits.
---------------------------------------------------------------------------

    \3386\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
    \3387\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
    \3388\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Given the various public comments on eligibility of technologies 
for off-cycle credits, the agencies are clarifying the regulations 
regarding technology eligibility, consistent with the intent and EPA's 
interpretation of the 2012 rule, as expressed in the preamble to the 
proposed and final rules. The agencies believe that clarifying the 
regulations will reduce confusion among manufacturers as to what 
technologies are eligible and reduce the overall program burden 
associated with EPA staff giving continued guidance to manufacturers 
regarding eligibility, as detailed in the 2012 rule preamble. 
Eligibility was thoroughly addressed in the 2012 final rule preamble, 
but the regulations were not as clear, which has led to confusion on 
the part of some manufacturers and delays in reviewing

[[Page 25239]]

credit applications.\3389\ The agencies are not establishing a new 
policy regarding eligibility, only amending the language reflecting the 
existing policy in the regulations for sake of clarity.
---------------------------------------------------------------------------

    \3389\ 77 FR 62726-36, 62835-37.
---------------------------------------------------------------------------

    As noted in the 2012 final rule preamble, the goal of the off-cycle 
credits program is to provide ``an incentive for the development and 
use of additional technologies to achieve real-world reductions in 
CO2 emissions.'' \3390\ EPA further stated that the intent 
of the program is to ``provide an incentive for CO2 and fuel 
consumption reducing off-cycle technologies that would otherwise not be 
developed because they do not offer a significant 2-cycle benefit.'' 
\3391\ The regulation at 40 CFR 86.1869-12(a) provides that 
manufacturers may generate credits for CO2 reducing 
technologies ``where the CO2 reduction benefit for the 
technology is not adequately captured on the Federal Test Procedure 
and/or Highway Fuel Economy Test.'' The regulation continues: ``[t]hese 
technologies must have a measurable, demonstrable, and verifiable real-
world CO2 reduction that occurs outside the conditions of 
the Federal Test Procedure and the Highway Fuel Economy Test.''
---------------------------------------------------------------------------

    \3390\ 77 FR 62833.
    \3391\ 77 FR 62836.
---------------------------------------------------------------------------

    Off-cycle credits are available for technologies that are not 
utilized when performing FTP and HFET tests because their operation is 
linked to a condition not found during the 2-cycle testing. For 
example, heating and cooling systems are not operated during the 2-
cycle test, and therefore, efficiency improvements to these systems are 
not captured at all on the 2-cycle tests. As the 2012 rule's language 
indicates, off-cycle credits are not necessarily limited to 
technologies listed on the menu or off-cycle technologies with no 
measurable benefit on the FTP and/or HFET. Off-cycle credits may be 
available for some technologies whose performance is measurable to some 
extent on the FTP and/or HFET but which perform measurably better off-
cycle. Active aerodynamic and stop-start technologies (menu item) are 
examples. However, there are limits on what the agencies would consider 
to be an off-cycle technology eligible for credits, as discussed below.
    Just as the regulations and preamble to the 2012 final rule listed 
technologies that the agencies considered to be off-cycle technologies, 
the preamble also discussed technologies that the agency would not 
consider off-cycle technologies--i.e., technologies the agencies 
consider to be ``adequately captured'' by the FTP and therefore not 
eligible for off-cycle credits. The preamble specifically noted that 
engine, transmission, mass reduction, passive aerodynamic design, and 
base tire technologies are not considered to be off-cycle technologies 
eligible for credits.\3392\ These are technologies that are considered 
to be ``integral or inherent to basic vehicle design.'' \3393\ In 
response to comments in the final rule, the agencies further clarified 
that advanced combustion concepts, such as camless engines, variable 
compression ratio engines, micro air/hydraulic launch assist devices, 
would not be considered to be eligible for credits.\3394\ This 
limitation to eligibility further extends to other engine designs, 
transmission designs, and electrification systems not specifically 
contemplated in the rulemaking, such as Atkinson combustion engines, 
and 9 and 10 speed transmissions, as well as to other hybrid systems 
such as 48 Volt technologies. Further, the 2012 final rule preamble 
stated that technologies included in the agencies' assessment for 
purposes of developing the standard would not be allowed to generate 
off-cycle credits and cites the technologies described in Chapter 3 of 
the 2012 final rule TSD.\3395\ Finally, off-cycle credits are not 
available for technologies required to be used by Federal Law or for 
crash avoidance systems, safety critical systems, or technologies that 
may reduce the frequency of vehicle crashes.\3396\
---------------------------------------------------------------------------

    \3392\ 77 FR 62732, 62836.
    \3393\ 77 FR 62732, 62836/1; 81 FR 73499.
    \3394\ 77 FR 62732.
    \3395\ 77 FR 62836.
    \3396\ 40 CFR 86.1869-12(a); 77 FR 62836.
---------------------------------------------------------------------------

    The preamble to the 2012 final rule provides the rationale for what 
the agency considers an off-cycle technology and, therefore, eligible 
for credits. Technologies that are integral or inherent to the vehicle 
are, by necessity, well represented on the 2-cycle test.\3397\ Examples 
provided in the preamble are engine, transmission, mass reduction, 
passive aerodynamic design, and base tire technologies. The control 
logic for these powertrain components, like the components themselves 
(i.e. engine and transmission), are constantly active, fully 
functioning, and operating over the entirety of the FTP and HFET. 
Similarly, an automatic transmission, regardless of whether it has 6-
speeds or 8-speeds, would still be constantly active, fully functioning 
and operating over the entirety of the FTP and HFET.\3398\ This would 
also be true for base engine technologies, advanced combustion 
concepts, engine components (pistons, valves, camshafts, crankshafts, 
oil pumps, etc.), and driveline components (individual components of 
the transmission, axle, and differential).\3399\
---------------------------------------------------------------------------

    \3397\ 77 FR 62732, 62836.
    \3398\ 76 FR 75024 (Dec. 1, 2011).
    \3399\ 77 FR 62732/2.
---------------------------------------------------------------------------

    Further, even if these technologies have greater benefits on 
supplemental test cycles, EPA has explained that it would be difficult 
to devise accurate A/B testing (i.e., with and without the technology) 
for these technologies.\3400\ The 2012 preamble states that ``EPA is 
limiting the off-cycle program to technologies that can be identified 
as add-on technologies conducive to A/B testing,'' partly because it 
would be very difficult accurately to parse out the off-cycle benefits 
for some integral technologies.\3401\ Because the technology is 
integral to the vehicle, there would not be an appropriate baseline 
(i.e., without the technology) vehicle to use for comparison. Vehicles 
are not built without tires, engines, passive aerodynamics or 
transmissions.
---------------------------------------------------------------------------

    \3400\ 76 FR 75024.
    \3401\ 77 FR 62836.
---------------------------------------------------------------------------

    Also, because these technologies are inherent to the vehicle 
design, their performance is already reflected in the stringency of the 
standard and giving credits for these inherent technologies would be a 
type of double-counting windfall.\3402\ ``[S]ince these methods are 
integral to basic vehicle design, there are fundamental issues as to 
whether they would ever warrant off-cycle credits. Being integral, 
there is no need to provide an incentive for their use, and (more 
importantly), these technologies would be incorporated regardless. 
Granting credits would be a windfall.'' \3403\ As such, EPA has laid 
out a clear basis that technological improvements to integral and 
inherent components are considered to be adequately captured on the FTP 
and HFET test.
---------------------------------------------------------------------------

    \3402\ 77 FR 62732.
    \3403\ See also 76 FR 75024.
---------------------------------------------------------------------------

    EPA is clarifying the regulations in a manner that is consistent 
with the intent and our interpretation of the 2012 rule, as expressed 
in the preambles to the proposed and final rules. The regulations are 
revised to specify that technologies used primarily to meet the 2-cycle 
standards are not eligible for off-cycle credits and that only 
technologies primarily installed for reducing off-cycle emissions would 
be eligible. The revised regulations specify that the technologies must 
not be integral or inherent to the basic vehicle design, such as, for 
example, engine,

[[Page 25240]]

transmission, mass reduction, passive aerodynamic design, and tire 
technologies. Exceptions to these general provisions include 
technologies already specified on the menu, including engine idle stop-
start, active aerodynamic improvements, and high-efficiency 
alternators. These technologies may provide some benefit on the 2-cycle 
test, but EPA determined in the 2012 rule that they are eligible for 
off-cycle credits because they are technologies that could be added to 
vehicles to provide discernable off-cycle reductions.
    Regulatory text at 40 CFR 86.1869-12(a) states: ``Manufacturers may 
generate credits for CO2 reducing technologies where the 
CO2 reduction benefit of the technology is not adequately 
captured on the Federal Test Procedure and/or the Highway Fuel Economy 
Test,'' to which EPA is adding, ``such that the technology would not be 
otherwise installed for purposes of reducing emissions (directly or 
indirectly) over those test cycles (i.e., on-cycle) for compliance with 
the [CO2] standards.'' EPA is also adding text to this 
paragraph of the regulations specifying: ``The technologies must not be 
integral or inherent to the basic vehicle design, such as engine, 
transmission, mass reduction, passive aerodynamic design, and tire 
technologies. Technologies installed for non-off-cycle emissions 
related reasons are also not eligible as they would be considered part 
of the baseline vehicle design. The technology must not be inherent to 
the design of occupant comfort and entertainment features except for 
technologies related to reducing passenger A/C demand and improving A/C 
system efficiency. Notwithstanding the provisions of this paragraph 
(a), off-cycle menu technologies included in paragraph (b) of this 
section remain eligible for credits.''
    The agencies believe the above regulatory changes will help reduce 
confusion over what technologies are eligible for off-cycle credits, 
refocusing the program on technologies that manufacturers would install 
on vehicles for purposes of reducing off-cycle emissions rather than 
obtaining additional credits for technologies installed primarily for 
2-cycle emissions reduction or for other reasons not related to 
emissions. This approach is consistent with the intent of the program 
as stated in the 2012 final rule to provide an incentive to develop and 
employ off-cycle technologies not adequately captured on the 2-cycle 
test procedure.
    Of the technologies recommended by manufacturers to be added to the 
menu, cooled EGR is an example of a technology that would not be 
eligible because it is an integral 2-cycle technology that EPA noted in 
its technology assessment in the MY 2012 rule. Cooled EGR is often an 
integral component of turbo charged gasoline direct injection engines 
which is a primary CO2 reduction strategy used by 
manufacturers to reduce 2-cycle emissions. The technologies are 
calibrated to act as a system such that is not possible to separate 
them in a way that would allow for a clear indication of the off-cycle 
benefit of cooled EGR as a stand-alone technology.
    EPA also received comments from the Auto Alliance regarding several 
technologies they believe should qualify as active warm-up off-cycle 
technologies. The Auto Alliance commented that systems that use waste 
heat from the exhaust gas stream should receive additional credits 
beyond the menu credits currently established for active engine and 
transmission warm-up.\3404\ However, when EPA established the menu 
credits for active transmission and engine warm-up in the 2012 rule, 
EPA envisioned waste heat from the exhaust as the primary source of 
heat to quickly bring the system to operating temperature as the basis 
for the warm-up technology credits.\3405\ Therefore, EPA does not 
believe additional credits, as suggested by the Auto Alliance, are 
warranted. EPA further notes that the definitions for active engine and 
transmission warm-up specify that ``waste heat'' be used in active 
warm-up technologies in order to qualify for the credits.\3406\ If a 
system first directs heat to warm the engine oil or warm the interior 
cabin, and only then to the engine or transmission, thereby delaying 
active warm-up, EPA would not view that heat as waste heat since it is 
serving other purposes during initial vehicle warm-up. EPA would also 
not consider this approach to be warming up the engine or transmission 
``quickly'' due to the potentially significant delay in warm-up 
activation. In developing the active warm-up credits, EPA focused on 
systems using heat from the exhaust as a primary source of waste heat 
because that heat would be available quickly and also be exhausted by 
the vehicle and otherwise unused.
---------------------------------------------------------------------------

    \3404\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3405\ See Joint Technical Support Document: Final Rulemaking 
for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards 
and Corporate Average Fuel Economy Standards, EPA-420-R-12-901, 
August 2012, p. 5-96--5-100.
    \3406\ 40 CFR 86.1869-12(b)(4)(v) and (vi).
---------------------------------------------------------------------------

    EPA allowed for the possible use of other sources of heat such as 
coolant as the basis for credits as long as those methods would 
``provide similar performance'' as extracting the heat directly from 
the exhaust system.\3407\ However, EPA may require manufacturers to 
demonstrate that the system is based on ``waste heat'' or heat that is 
not being preferentially used by the engine or other systems to warm-up 
other areas like engine oil or the interior cabin. Systems using waste 
heat from the coolant do not qualify for credits if their operation 
depends on, and is delayed by, engine oil temperature or interior cabin 
temperature. As the engine and transmission components are warming up, 
the engine coolant and transmission oil do not have any `waste' heat 
available for warming up anything else on the vehicle. During engine 
and transmission warm-up, the only waste heat source in a vehicle with 
an internal combustion engine is the engine exhaust as the transmission 
and coolant have not reached warmed-up operating temperature and 
therefore do not have any heat to share. Conserving heat in a 
transmission is not a rapid transmission warm-up using waste heat. 
Unless the component with lubricating oil and coolant is operating at 
its fully warmed-up design temperature, by EPA's definition, that 
component does not have any waste heat available for transfer from the 
lubricating oil or coolant to any other device until it has reached its 
fully warmed-up operating temperature (i.e. the temperature when the 
cooling system is enabled). A qualifying system may involve a second 
cooling loop that operates independent of the primary coolant system 
and is not dependent on or otherwise delayed by, for example, cabin 
temperature. Evaluating whether such systems qualify for menu credits 
often requires additional information regarding system design to 
understand better how the system uses waste heat. Given the complexity 
of these systems and the need to sometimes consider the details of how 
a system operates, EPA is not making any changes to the menu regarding 
warm-up technologies.
---------------------------------------------------------------------------

    \3407\ See Joint Technical Support Document: Final Rulemaking 
for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards 
and Corporate Average Fuel Economy Standards, p. 5-99, EPA-420-R-12-
901, August 2012.
---------------------------------------------------------------------------

    The Auto Alliance further commented that active transmission bypass 
valves should qualify for active transmission warm-up credits.\3408\ 
The Auto Alliance

[[Page 25241]]

commented that traditional transmission oil coolers are always active 
and sized for extreme or worst-case hot ambient conditions. The coolers 
will, in colder ambient conditions, keep the transmission temperatures 
well outside of their most efficient operating range. The bypass valve 
circumvents the cooler when the transmission is relatively cold 
preserving the transmission heat, so the transmission warms more 
quickly. EPA disagrees that this type of approach should be eligible 
for active transmission warm-up because it does not use waste heat to 
add heat to the transmission. Instead, it prevents useful heat already 
present in the transmission from being unnecessarily removed. Also, EPA 
does not view this type of bypass valve as an off-cycle technology but 
rather as part of a good engineering design of a transmission cooler 
system. Many vehicles already are designed with transmission cooler 
bypass valves. EPA does not believe existing coolers qualify as warm-up 
technologies simply because they are disabled under cold conditions. 
This approach does not represent the addition of a new off-cycle warm-
up technology but the disabling of an existing cooling technology.
---------------------------------------------------------------------------

    \3408\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    Although the agencies did not consider changes to the program to 
allow credits for safety-related technologies and autonomous vehicle 
technologies in the proposal, comments were received both in favor of 
and not in favor of allowing such credits.\3409\ The agencies note that 
the rationale for not allowing off-cycle credits for safety-related or 
crash avoidance technologies has not changed since the 2012 rule and, 
therefore, in the proposed rule the agencies did not consider making 
any changes to allow off-cycle credits for safety-related 
technologies.\3410\ The agencies continue to believe that there is a 
very significant distinction between technologies providing direct and 
reliably quantifiable improvements to fuel economy and CO2 
emission reductions, and technologies which provide those improvements 
by indirect means, where the improvement is not reliably quantifiable, 
and may be speculative (or in many instances, non-existent), or may 
provide benefit to other vehicles on the road more than for themselves. 
The agencies also continue to believe that the advancement of crash-
related and crash avoidance systems specifically is best left to 
NHTSA's exercise of its vehicle safety authority.
---------------------------------------------------------------------------

    \3409\ See, e.g., SAFE, Detailed Comments, NHTSA-2018-0067-
11981; AAA, Detailed Comments, NHTSA-2018-0067-11979.
    \3410\ 77 FR 62733.
---------------------------------------------------------------------------

    Auto manufacturers and suppliers also commented that EPA should 
adopt ``eco-innovation'' credits approved in the European Union (EU) 
vehicle CO2 reduction program as part of the off-cycle 
credits program.\3411\ No data was provided as to why the credits would 
be appropriate for the U.S. vehicle fleet. EPA did not consider or 
request comment on the EU credits program and does not believe the 
credit levels would necessarily be appropriate for the U.S. fleet given 
the very different vehicle use and driving patterns between Europe and 
the U.S. Thus, there is no assurance that the credits would be based on 
real-world emissions reductions.
---------------------------------------------------------------------------

    \3411\ See, e.g., Mitsubishi, Detailed Comments, NHTSA-2018-
0067-12056.
---------------------------------------------------------------------------

    EPA received comments from the Auto Alliance and Global Automakers 
that EPA should automatically award credits if the agency does not take 
final action within 90 days of receiving a request for credits.\3412\ 
Regarding these comments, EPA does not believe such a provision is in 
keeping with maintaining the integrity of the off-cycle credits 
program. As discussed above, EPA often requires time to sort through 
complex issues to determine if the technologies meet the regulatory 
requirements for receiving credits and whether the credits have been 
quantified appropriately. In some instances, EPA has received public 
comments and manufacturer rebuttals to those comments that takes 
additional time to consider before making a final decision. EPA's goal 
continues to be to evaluate applications for credits in as timely a 
manner as is possible given the issues that must be addressed and 
within the resources available. While EPA's need carefully to consider 
applications may slow down the approval process or result in credits 
not being approved, it remains paramount to ensure credits are not 
provided to technologies that do not provide actual off-cycle benefits, 
and thereby do not meet the regulations. In the past, longer time 
frames for EPA review have not caused manufacturers to lose credits 
where credits are determined by EPA to be warranted under the 
regulations. EPA believes that the changes EPA is making to the program 
will help streamline the program and reduce confusion, thus helping to 
reduce the time necessary to evaluate applications and provide final 
decisions to manufacturers.
---------------------------------------------------------------------------

    \3412\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Global Automakers, Detailed Comments, NHTSA-2018-0067-12032.
---------------------------------------------------------------------------

(7) Supplier Role in the Off-Cycle Credits Program
    Prior to proposal, EPA heard from many suppliers and their trade 
associations about an interest in allowing suppliers to have a formal, 
regulatorily defined role in the off-cycle credits program.\3413\ EPA 
requested comment on providing a pathway for suppliers, along with at 
least one auto manufacturer partner, to submit off-cycle applications 
for EPA approval. As described in the proposal, under such an approach, 
an application submitted by a supplier and vehicle manufacturer would 
establish a credit and/or methodology for demonstrating credits that 
all auto manufacturers could then use in their subsequent applications. 
EPA requested comment on requiring that the supplier be partnered in a 
substantive way with one or more auto manufacturers to ensure that 
there is a practical interest in the technology prior to EPA investing 
resources in the approval process. The supplier application would be 
subject to public review and comment prior to an EPA decision. However, 
once approved, subsequent auto manufacturer applications requesting 
credits based on the supplier methodology would not be subject to 
public review. Under this concept, the credits would be available 
provisionally for a limited period of time, allowing manufacturers to 
implement the technology and collect data on their vehicles in order to 
support a continuation of credits for the technology in the longer 
term. Also, as envisioned by EPA in its request for comment, the 
provisional credits could be included under the menu credit cap since 
they would be based on a general analysis of the technology rather than 
manufacturer-specific data.
---------------------------------------------------------------------------

    \3413\ 83 FR 43461.
---------------------------------------------------------------------------

    Auto manufacturers' and suppliers' comments were generally 
supportive of an expanded role for suppliers in the off-cycle credit 
program. The Auto Alliance supported allowing a supplier to lead the 
application process but did not support the provisional credit concept 
since the follow-up testing conducted by manufacturers may not support 
the level of credits initially claimed by the supplier, resulting in a 
lower than anticipated credit.\3414\ Instead, the Auto Alliance 
suggested a separate cap for supplier-based credits and noted that 
manufacturers could submit their own data if they wanted to pursue 
credits levels that exceeded the cap. General Motors similarly 
disagreed with the provisional credits that might

[[Page 25242]]

be rescinded if subsequent testing does not fully validate the value of 
the technology.\3415\ MEMA supported the request for comments regarding 
a supplier-led process but did not support requiring that suppliers 
have an auto manufacturer partner.\3416\ MEMA commented that there 
would be no incentive for a supplier to go through the product/
technology development process, collect the necessary data, and 
undertake the full application process for a product/technology that 
would not generate manufacturer interest.
---------------------------------------------------------------------------

    \3414\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3415\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
    \3416\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
---------------------------------------------------------------------------

    At this time, EPA believes additional discussions with interested 
parties and an opportunity for public comment, both of which are beyond 
the scope of this rulemaking, are needed. EPA continues to believe such 
an approach could encourage the further development of off-cycle 
technologies, but must be done in a reasonable way that ensures the 
credits are based on real-world emissions reductions.
    Under the approach suggested by the Auto Alliance, manufacturers 
could claim supplier-based credits indefinitely and EPA might never 
receive any manufacturer data substantiating the credits unless that 
data supported a credit that exceeded the level established through the 
supplier process. EPA is concerned such a one-way ratchet approach 
could result in the loss of emissions benefits and undermine the 
integrity of the off-cycle credit program. EPA also remains concerned 
about the potential for a significantly increased volume of credit 
applications, including the potential for applications for proposed 
technologies that manufacturers might in reality have no interest in 
adopting. EPA understands MEMA's perspective on the issue of requiring 
a manufacturer partner, but a supplier-only process would potentially 
open the door to many requests such that the agency would need to 
expend considerable additional resources. EPA notes that nothing in the 
current regulations prevents collaboration between manufacturers and 
suppliers. Suppliers can initiate this process; manufacturer 
participation will be necessary to complete an application. EPA will 
provide additional clarity about this process through a subsequent 
technical amendments rulemaking.
(8) Other Considerations
    Avista Oil commented that EPA should provide an opportunity for 
credits based on the use of recycled engine oil. Avista Oil commented 
that there are CO2 emissions reductions associated with the 
use of recycled used engine oil and that vehicle manufacturers should 
be awarded credits for the use of recycled oil. Avista Oil's comment is 
not within the scope of the rulemaking. The off-cycle credits program 
focuses on providing credits for technologies that, when applied to the 
vehicle, the result is lower quantifiable real-world emissions from the 
vehicle. According to Avista Oil's comment, their recycled oil 
technology benefits are associated with the recycling process rather 
than lowering vehicle emissions on the road. Therefore, EPA would not 
view the technology as eligible for off-cycle credits, and EPA did not 
propose any other credit specific to the use of recycled engine oil.
    Several commenters recommended that EPA raise the credit caps and 
credit values for thermal controls based on recent work by the National 
Renewable Energy Lab (NREL). Commenters suggested that credit values 
should be raised by 64 percent. In response, as discussed in the 
preamble, EPA is retaining the current menu credit caps and menu credit 
values due to uncertainties involved with the emissions projections and 
estimated credit values. Manufacturers may generate additional credits 
through the off-cycle credits program using the other two pathways by 
providing individual vehicle data. EPA recognizes additional modeling 
analysis has been performed by NREL that indicates the potential 
benefit of all thermal technologies including glazing. EPA designed the 
thermal control program and related caps based on previous NREL work 
and applied the thermal caps at the current levels to account for the 
wide range of uncertainties--including the uncertainty of the benefit 
from the combination of thermal technologies and the uncertainty 
highlighted by the different credit levels across the NREL studies. EPA 
believes the separate current thermal menu program cap and AC 
efficiency program cap continue to be reasonable for application across 
the fleet given these uncertainties.
    Enhanced Protective Glass Automotive Association (EPGAA) and Vitro 
commented that the regulations established by the 2012 rule included an 
oversight in defining the baseline Tts (the metric used to evaluate 
thermal reflectivity of glass). EPGAA commented that there was an 
omission in the case of trucks, where the regulations do allow the use 
of privacy glass in locations other than the windshield and the front 
doors. The commenter discussed that the reference baseline glass for 
trucks, SUVs, and CUVs should have already included privacy glass for 
some of the rearward windows. In response, EPA recognized when the 
thermal credit program was finalized in 2012 that some of the vehicles 
within the reference fleet upon which the credits were based were 
already composed of vehicles with this type of thermal reflective 
glass. However, the agency found it difficult to estimate what portion 
of the fleet contained privacy glass and what the Tts rating was for 
privacy glass across the fleet. Because of this lack of specificity in 
the fleet composition and glass ratings, the agencies determined that 
the most appropriate approach was to allow credit for any glass meeting 
the finalized Tts requirements, and the total thermal cap was designed 
to account for this and other uncertainties.
    Ford and others commented that thermal control technology credit 
caps should be implemented on a fleet average basis rather than on a 
``per VIN'' basis. These commenters argued that the per VIN basis 
creates a reporting burden that is misaligned with the current 
reporting structure and creates program complexity and unnecessary 
workload. In response, EPA continues to believe that applying the 
thermal control credit cap on a per vehicle (per VIN) basis is 
appropriate due to the synergistic effects among these technologies. 
The CO2 reduction potential of applying thermal control technologies is 
limited within any given vehicle. The program has been implemented in 
this manner since MY2014, and manufacturers have in fact reported the 
necessary information to generate thermal control credits.
    Gentherm, GM, MEMA, and The ITB Group commented that cooled seats 
should be added to the menu based on the approved GM off-cycle credits 
application and NREL study. EPA and NHTSA are not adding cooled seat 
technology to the menu because the agencies have received data from 
only a single manufacturer. By contrast, for the technologies EPA and 
NHTSA are adding to the menu in this final rule, the agencies have 
assessed data from multiple manufacturers. EPA notes however that the 
streamlining provisions being finalized in this action should 
facilitate other manufacturers in being able to apply for off-cycle 
credits by using GM's methodology.
    Finally, on October 1, 2018, EPA proposed a technical correction 
separate from the SAFE Vehicles rulemaking for

[[Page 25243]]

the off-cycle credits pathway based on 5-cycle testing (83 FR 49344). 
This proposal would correct an error in the regulations established as 
part of the 2012 final rule. Some commenters expressed their support 
for the correction as part of their SAFE Vehicles rule comments. EPA 
notes that this correction continues to be part of a separate 
rulemaking and is not being addressed in the SAFE Vehicles final rule.
c) Final Decisions on the 2016 Alliance/Global Petition
(1) Retroactive A/C and Off-Cycle CAFE Adjustments
    In 2016, the Alliance and Global submitted a petition for 
rulemaking, which included requests that: (1) NHTSA allow retroactive 
credits for A/C and off-cycle incentives for MYs 2012 to 2016; and (2) 
NHTSA and EPA revisit the average A/C efficiency benefit calculated by 
EPA applicable to MYs 2012 through 2016. The Alliance/Global argued 
that A/C efficiency improvements were not properly acknowledged in the 
CAFE program, and that manufacturers had exceeded the A/C efficiency 
improvements estimated by the agencies. The petitioners requested that 
EPA also amend its regulations such that manufacturers would be 
entitled to additional A/C efficiency improvement benefits 
retroactively. The petitioners also argued that NHTSA incorrectly 
stated the agency had taken off-cycle adjustments into consideration 
when setting standards for MYs 2017 through 2025, but not for MYs 2010-
2016. The Alliance/Global further contended that because neither NHTSA 
nor EPA considered off-cycle adjustments in formulating the stringency 
of the MY 2012-2016 standards, NHTSA should retroactively grant 
manufacturers off-cycle adjustments for those model years as EPA did. 
Doing so, they said, would maintain consistency between the agencies' 
programs.
    Of the two agencies, EPA was the first to establish an off-cycle 
technology program. For MYs 2012 through 2016, EPA allowed 
manufacturers to request off-cycle credits for ``technologies that 
achieve [CO2] reductions that are not reflected on current 
test procedures . . .'' \3417\ In the subsequent MY 2017 and later 
rulemaking, NHTSA joined EPA and included an off-cycle program for CAFE 
compliance. The Alliance/Global petition cited a statement in the MYs 
2012-2016 final rule as affirmation that NHTSA took off-cycle 
adjustments into account in formulating the MYs 2012-2016 stringencies, 
and therefore should allow manufacturers to earn off-cycle benefits in 
model years that have already passed.
---------------------------------------------------------------------------

    \3417\ 75 FR 25341, 25344 (May 7, 2010). EPA had also provided 
an option for manufacturers to claim ``early'' off-cycle credits in 
the 2009-2011 time frame.
---------------------------------------------------------------------------

    In the NPRM, NHTSA tentatively decided to retain the structure of 
the existing A/C efficiency program and not extend it to MYs 2010 
through 2016. For the rulemaking for MYs 2012 through 2016, NHTSA 
determined it was unable to consider improvements manufacturers made to 
passenger car A/C efficiency in calculating CAFE 
compliance.3418 3419 However, EPA did consider passenger car 
improvements to A/C efficiency for that timeframe. To allow 
manufacturers to build one fleet that complied with both EPA and NHTSA 
standards, the CAFE and CO2 standards were offset to account 
for the differences borne out of A/C efficiency improvements. 
Specifically, the agencies converted EPA's grams/mile standards to 
NHTSA mpg (CAFE) standards. EPA then estimated the average amount of 
improvement manufacturers were expected to earn via improved A/C 
efficiency. From there, NHTSA took EPA's converted mpg standard and 
subtracted the average improvement attributable to improvement in A/C 
efficiency. NHTSA set its standard at this level to allow manufacturers 
to comply with both standards with similar levels of technology.\3420\
---------------------------------------------------------------------------

    \3418\ At that time, NHTSA stated ``[m]odernizing the passenger 
car test procedures, or even providing similar credits, would not be 
possible under EPCA as currently written.'' 75 FR 25557 (May 7, 
2010).
    \3419\ 74 FR 49700 (Sept. 28, 2009).
    \3420\ Id.
---------------------------------------------------------------------------

    Likewise, EPA tentatively decided in the NPRM not to modify its 
regulations to change the way to account for A/C efficiency 
improvements. EPA believed this was appropriate as manufacturers 
decided what fuel economy-improving technologies to apply to vehicles 
based on the standards as finalized in 2010.\3421\ This included 
deciding whether to apply traditional tailpipe technologies, A/C 
efficiency improvements, or both. Granting A/C efficiency adjustments 
to manufacturers retroactively could result in arbitrarily varying 
levels of adjustments granted to manufacturers, similar to the 
Alliance/Global request regarding retroactive off-cycle adjustments. 
Thus, the existing A/C efficiency improvement structure for MYs 2010 
through 2016 would remain unchanged.
---------------------------------------------------------------------------

    \3421\ In the MY 2017 and later rulemaking, NHTSA reaffirmed its 
position it would not extend A/C efficiency improvement benefits to 
earlier model years. 77 FR 62720 (Oct. 15, 2012).
---------------------------------------------------------------------------

    NHTSA also tentatively decided manufacturers should not be granted 
retroactive off-cycle adjustments for MYs 2010 through 2016, and 
presented a number of clarifications to justify the denial. In 
particular, Alliance/Global pointed to a general statement where NHTSA, 
while discussing consideration of ``the effect of other motor vehicle 
standards of the Government on fuel economy,'' stated that that 
rulemaking resulted in consistent standards across the program.\3422\ 
The Alliance/Global petition took this statement as a blanket assertion 
that NHTSA's consideration of all ``relevant technologies'' included 
off-cycle technologies. To the contrary, as quoted above, NHTSA 
explicitly stated it had not considered these off-cycle 
technologies.\3423\
---------------------------------------------------------------------------

    \3422\ Id.
    \3423\ Likewise, EPA stated it had not considered off-cycle 
technologies in finalizing the MYs 2012-2016 rule. ``Because these 
technologies are not nearly so well developed and understood, EPA is 
not prepared to consider them in assessing the stringency of the 
CO2 standards.'' Id. at 25438.
---------------------------------------------------------------------------

    The fact that NHTSA had not taken off-cycle adjustments into 
consideration in setting its MYs 2012-2016 standards makes granting the 
Alliance/Global request inappropriate. Doing so could result in a 
question as to whether the MY 2012-2016 standards were maximum feasible 
under 49 U.S.C. 32902(b)(2)(B). If NHTSA had considered industry's 
ability to earn off-cycle adjustments--an incentive that allows 
manufacturers to utilize technologies other than those that were being 
modeled as part of NHTSA's analysis--the agency might have concluded 
more stringent standards were maximum feasible. Additionally, granting 
off-cycle adjustments to manufacturers retroactively raises questions 
of equity. NHTSA issued its MYs 2012-2016 standards without an off-
cycle program, and manufacturers had no reason to anticipate that NHTSA 
would allow the use off-cycle technologies to meet fuel economy 
standards. Therefore, manufacturers made fuel economy compliance 
decisions with the expectation that they would have to meet fuel 
economy standards using on-cycle technologies. Generating off-cycle 
adjustments retroactively would arbitrarily reward some (and 
potentially disadvantage other) manufacturers for compliance decisions 
they made without the knowledge such technologies would be eligible for 
NHTSA's off-cycle program. Thus, NHTSA tentatively decided to deny 
Alliance/Global's request for retroactive off-cycle adjustments.

[[Page 25244]]

    It is worth noting that in the MYs 2017 and later rulemaking, NHTSA 
and EPA did include off-cycle technologies in establishing the 
stringency of the standards. As Alliance/Global noted, NHTSA and EPA 
limited their consideration to stop-start and active aerodynamic 
features because of limited technical information on these 
technologies.\3424\ At that time, the agencies stated they ``have 
virtually no data on the cost, development time necessary, 
manufacturability, etc. [sic] of these technologies. The agencies thus 
cannot project that some of these technologies are feasible within the 
2017-2025 timeframe.'' \3425\
---------------------------------------------------------------------------

    \3424\ Alliance/Global Petition at 7.
    \3425\ Draft Joint Technical Support Document: Rulemaking for 
2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and 
Corporate Average Fuel Economy Standards (November 2011), p. 5-57.
---------------------------------------------------------------------------

    As described above, NHTSA first allowed manufacturers to generate 
off-cycle technology fuel consumption improvement values equivalent to 
CO2 off-cycle credits in MY 2017.\3426\ In finalizing the 
rule covering MYs 2017 and later, NHTSA declined to retroactively 
extend its off-cycle program to apply to model years 2012 through 
2016,\3427\ explaining ``NHTSA did not take [off-cycle credits] into 
account when adopting the CAFE standards for those model years. As 
such, extending the credit program to the CAFE program for those model 
years would not be appropriate.'' \3428\
---------------------------------------------------------------------------

    \3426\ 77 FR 62840 (Oct. 15, 2012).
    \3427\ See id.; EPA decided to extend provisions from its MY 
2017 and later off-cycle program to the 2012-2016 model years.
    \3428\ Id.
---------------------------------------------------------------------------

    In the NPRM, NHTSA and EPA sought any further comments on the 
tentative denials of the retroactive requests in the Alliance/Global. 
The Auto Alliance and Fiat Chrysler provided additional comments on the 
tentative denial of the petition requests from the Alliance/Global. The 
commenters cited that the widening gap between the regulatory standards 
and actual industry-wide new vehicle average fuel economy that has 
become evident since 2016, despite the growing use of improvement 
``credits'' from various flexibility mechanisms, such as off-cycle 
technology credits, mobile air conditioner efficiency credits, mobile 
air conditioner refrigerant leak reduction credits and credits from 
electrified vehicles.\3429\ The commenters believe that applying 
retroactive credits for the new flexibilities for MYs 2012 to 2016 can 
address the current compliance deficiencies.
---------------------------------------------------------------------------

    \3429\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; 
Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
---------------------------------------------------------------------------

    Upon consideration of the issue, NHTSA is finalizing its decision 
to deny any retroactive off-cycle adjustments in the CAFE program for 
MYs 2012-2016. As mentioned in the NPRM, NHTSA is concerned about the 
negative impact of allowing retroactive credits, which could undermine 
the stringency of the MYs 2012-2016 standards. EPA is finalizing its 
decision not to modify its regulations to change the benefits for A/C 
efficiency improvements. As mentioned by EPA, the current approach 
creates uniformity and objectivity in determining A/C efficiency 
benefits. Consequently, because EPA is maintaining the current A/C 
determination methodology and NHTSA already considered those A/C 
adjustments in its MYs 2012-2016 CAFE standards, NHTSA is also 
finalizing its decisions in this rule to deny any retroactive A/C 
adjustments in the CAFE program for MYs 2012-2016.
(2) Petition Requests on A/C Efficiency and Off-Cycle Program 
Administration
    As discussed above, NHTSA and EPA jointly administer the off-cycle 
program. The 2016 Alliance/Global petition requested that EPA and NHTSA 
make various adjustments to the off-cycle program; specifically, the 
petitioners requested that the agencies should:
     re-affirm that technologies meeting the stated definitions 
are entitled to the off-cycle credit at the values stated in the 
regulation;
     re-acknowledge that technologies shown to generate more 
emissions reductions than the pre-approved amount are entitled to 
additional credit;
     confirm that technologies not in the null vehicle set but 
which are demonstrated to provide emissions reductions benefits 
constitute off-cycle credits; and
     modify the off-cycle program to account for unanticipated 
delays in the approval process by providing that applications based on 
the 5-cycle methodology are to be deemed approved if not acted upon by 
the agencies within a specified timeframe (for instance 90 days), 
subject to any subsequent review of accuracy and good faith.\3430\
---------------------------------------------------------------------------

    \3430\ Alliance/Global Petition at 20.
---------------------------------------------------------------------------

    With respect to Alliance/Global's request regarding off-cycle 
technologies that demonstrate emissions reductions greater than what is 
allowable from the menu, this final rule retains that capability. As 
was the case for MYs 2017-2021, a manufacturer may still apply for 
FCIVs and CO2 credits beyond the values listed on the menu, 
provided the manufacturer demonstrates the CO2 and fuel 
economy improvement.\3431\ This includes the two-alternative processes 
for demonstrating CO2 reductions and fuel economy 
improvement for gaining benefits using either the 5-cycle or 
alternative approval methodologies.\3432\
---------------------------------------------------------------------------

    \3431\ 77 FR 62837 (Oct. 15, 2012).
    \3432\ 40 CFR 86.1869-12.
---------------------------------------------------------------------------

    The agencies have considered Alliance/Global's requests to 
streamline aspects of the A/C efficiency and off-cycle programs in 
response to the issues outlined above. Among other things, Alliance/
Global requested that the agencies consider providing for a default 
acceptance of petitions for off-cycle credits after a specified period 
of time, provided that all required information has been provided, to 
accelerate the processing of off-cycle credit requests. While the 
agencies agree with the merits of A/C efficiency and off-cycle 
programmatic improvements, there are significant concerns with the 
concept of approving petition requests by default because such requests 
may not address program issues like uncertainty in quantifying program 
benefits, or general program administration.
    Based on its consideration of the issues raised by the Alliance/
Global, EPA has adopted in this final rule new processes for 
streamlining the compliance mechanisms for approving off-cycle and 
applications as discussed in the preceding section.
(3) Other EPA Responses to Alliance Requests
    One issue raised in the Alliance/Global Automakers June 2016 
petition (item 6 titled ``Refrain from Imposing Unnecessary 
Restrictions on the Use of Credits'') for EPA's consideration concerns 
how credits are managed within the CO2 program. The Alliance 
and Global Automakers suggested that EPA allow more flexibility in 
using credits generated under the various credit programs such as air 
conditioning or off-cycle credits by allowing them to be carried 
forward or back independently. Under this approach, a manufacturer 
would be allowed, for example, to carry their air conditioning credits 
back to cover a previous deficit while running a deficit in a current 
model year. The Alliance referred to this petition request in their 
comments, noting they believe the request ``remains pertinent in the 
context of this rulemaking.''
    In response, EPA did not raise this issue or any related 
programmatic changes in the proposal and therefore

[[Page 25245]]

these comments are not within the scope of the rulemaking. EPA notes 
the GHG and CAFE programs are harmonized on the aggregation of credits.
    The automakers' petition also requested that EPA correct the 
multiplier equation in the regulations so that manufacturers may 
generate the intended number of credits (item 8, ``Correct the 
Multiplier for BEVs, PHEVs, FCVs, and CNGs''). This request concerns an 
error in the regulations established in the 2012 Final Rule that 
results in manufacturers generating fewer than intended for MY 2017-
2021 vehicles in some cases. In October 2018, in response to this 
petition request, EPA issued a proposed rule separate from the SAFE 
Vehicles NPRM to correct the error in the previously established 
regulations. EPA will continue to address this issue and related 
comments in that separate rulemaking. CAFE does not include multiplier 
credits and therefore this is not a harmonization issue.
4. Specialty Vehicles With Low Mileage (SVLM)
    In response to the NPRM, Volkswagen submitted comments seeking to 
adopt a new flexibility for specialty vehicles with low mileage 
(SVLM).\3433\ The flexibility would apply to specialty vehicles 
produced at low volumes and produced for infrequent use. They argued 
these specialty vehicles do not approach the vehicle miles traveled of 
typical vehicles. They requested that NHTSA and EPA allow the SVLM 
flexibility for vehicles that demonstrate limited predicted driving 
use. The flexibility would allot each manufacturer a limited annual 
production of 5,000 SVLM vehicles. It was also proposed that, within 
this limited product volume, each SVLM would retain its footprint 
derived performance target (per model type), but would utilize a 
modified VMT for determining any credits or debits associated with the 
performance of these vehicles within the manufacturer's fleet.
---------------------------------------------------------------------------

    \3433\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
---------------------------------------------------------------------------

    The agencies have considered the request from Volkswagen for 
credits or debits and fuel economy adjustments for SVLM vehicles and 
are denying the request. NHTSA notes that Congress prescribed 
alternative (reduced) CAFE standards for low-volume manufacturers, 
codified in 49 CFR part 525. Low-volume manufacturers' vehicles are 
often high-end sports cars and are not typically driven by their owners 
for long distances. Congress limited this exemption under the CAFE 
program to manufacturers of fewer than 10,000 passenger 
automobiles.\3434\ EPA has a similar program for smallvolume 
manufacturers which are defined as manufacturers with average sales for 
the three most recent consecutive model years of less than 5,000 
vehicles.\3435\ The flexibility proposed by Volkswagen would presumably 
be in addition to these existing provisions, but Volkswagen does not 
identify a source of authority for it. The agencies also have a number 
of questions about how specifically a SVLM concept might be 
implemented, such as whether every manufacturer would simply identify 
the 5,000 vehicles with the lowest projected VMT or lowest fuel economy 
and therefore qualify for credits for 5,000 vehicles every model year, 
or whether there should be additional criteria for vehicles to be 
included. The NPRM did not seek comment on a SVLM concept and the 
agencies did not receive other comments on the requested program. 
Therefore, the agencies are not adopting the SVLM concept suggested by 
Volkswagen.
---------------------------------------------------------------------------

    \3434\ 49 U.S.C. 32902(d)(1).
    \3435\ 40 CFR 86.1818-12(g).
---------------------------------------------------------------------------

E. CO2 and CAFE Compliance Issues Not Addressed in the NPRM

1. CO2 and CAFE Adjustments for 5-Cycle Testing
    EPA and NHTSA received several comments requesting that the 
agencies revise current CAFE test procedures to use EPA's 5-cycle test 
procedures in place of the 2-cycle test procedures that have been 
largely unchanged since the inception of the CAFE program, or offset 
measured 2-cycle test fuel economy and CO2 emissions for 
CO2 and CAFE compliance. Walter Kreucher commented ``some 
technologies (Hybrid Electric) have penalties on the road that are not 
reflected on the tests used to determine CAFE compliance. . . . If the 
Agencies want to provide adjustment factors for A/C and other `Off-
Cycle' conditions it must do so in both the positive and negative 
direction'' (sic).\3436\ AVE commented that the agencies should use 5-
cycle procedures rather than 2-cycle procedures, arguing that the 5-
cycle model better demonstrates real-world driving conditions and would 
lead to a more simplified credit allocation system.\3437\ BorgWarner 
echoed those comments, stating that the 5-cycle test is more accurate 
than the 2-cycle test and would reduce the need for credit 
adjustments.\3438\ Jeremy Michalek commented that the fuel economy 
values the public sees reflected on vehicles for purchase (e.g., on the 
Monroney label or in new car advertising) is calculated from the 5-
cycle test; updating the 2-cycle test to capture more of the vehicle's 
fuel efficiency factors would allow for better consistency and a more 
accurate fuel efficiency measure.\3439\ The Auto Alliance proposed that 
the EPA revise its methodology for calculating off-cycle improvements 
when using the 5-cycle methodology by subtracting the 2-cycle benefit 
from the 5-cycle benefit to ensure credits are calculated 
properly.\3440\
---------------------------------------------------------------------------

    \3436\ Walter Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
    \3437\ AVE, Detailed Comments, NHTSA-2018-0067-11696.
    \3438\ BorgWarner, Detailed Comments, NHTSA-2018-0067-11895.
    \3439\ Jeremy Michalek, et al., Detailed Comments, NHTSA-2018-
0067-11903.
    \3440\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
---------------------------------------------------------------------------

    The NPRM did not seek comment on revising compliance test 
procedures to use 5-cycle test procedures in place of 2-cycle test 
procedures, either entirely or broadly. Such a change would require 
extensive assessment and analysis to consider how changes could be 
implemented and what standards might be maximum feasible for CAFE and 
appropriate and reasonable for CO2 for new test procedures. 
There has been no analysis conducted to estimate the impacts of such a 
change on the levels of the standards. Therefore, making these 
requested changes is outside the scope of this rulemaking.
2. National Zero Emissions Vehicle Concept
    Although the agencies did not discuss or request comment on a 
National Zero Emissions Vehicle (NZEV) program concept, several 
organizations commented on that topic. Some discussed ideas from a task 
force that was formed by the governors of nine States who signed a 
memorandum of understanding (MOU) committing to undertake joint 
cooperative actions to build a robust market for ZEVs under their 
individual state programs. Collectively, these States have committed to 
having at least 3.3 million ZEVs operating on their roadways by 2025. 
ZEVs include battery-electric vehicles (BEVs), plug-in hybrid electric 
vehicles (PHEVs), and hydrogen fuel-cell electric vehicles (FCEVs). 
Comments on an NZEV concept were received from General Motors, CARB, 
Edison Electric Institute, Honda, NCAT, Workhorse Group, and Volvo.

[[Page 25246]]

    General Motors offered comments supporting an NZEV program, stating 
that it continues to expect California to be the leader of the EV 
market but hopes a national effort will be put forth, making the U.S. a 
global leader in EV technology development and deployment. \3441\ 
General Motors stated it believes an NZEV program would further U.S. 
national security interests, make the U.S. more competitive with China, 
which already has an NZEV program, and reduce U.S. dependence on 
foreign petroleum. General Motors requested that EPA incentivize EV 
deployment, including providing credits for autonomous EVs and EVs that 
are used in rideshare programs.\3442\ General Motors outlined their 
proposed NZEV program which would include increasing ZEV requirements 
annually, establishing credit banks for manufacturers based on national 
ZEV sales, and ZEV multipliers for vehicles over 5,250 lbs., autonomous 
vehicles using EV, and EVs in rideshare programs. General Motors also 
proposed that requirements would be revisited if EV battery cell were 
not available at the costs Argonne National Lab forecasts by 2025. 
General Motors also suggested implementing a Zero Emissions Task Force 
that would promote complementary policies. General Motors acknowledged 
that the NZEV program would have to be subject to acceleration or delay 
depending on how quickly technologies are incentivized like battery 
cost.
---------------------------------------------------------------------------

    \3441\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
    \3442\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
---------------------------------------------------------------------------

    CARB recommended a national ZEV multiplier, stating that a national 
incentive would help ensure ZEVs and PHEVs were being produced for sale 
beyond the ten States that have ZEV programs.\3443\ The Edison Electric 
Institute supported increasing stringency of fuel economy and 
CO2 standards and incorporating policies from ZEV States to 
create a ``One National Program.'' \3444\ Workhorse Group commented 
that a national ZEV mandate, where agencies progressively increase the 
mandated percentage of electric vehicles in every fleet, merits serious 
consideration by the agencies. They contended that an NZEV would have 
to work with the current State ZEV mandates and not preempt the 
progress already made.\3445\ Volvo, and Honda were proponents of 
incorporating ZEV standards into a national program. Volvo requested 
nationwide credits for ZEVs since there are 40 States without ZEV 
mandates.\3446\ Honda mentioned that incorporating California's ZEV 
credits into the national program would reduce compliance costs for 
manufacturers while incentivizing technological development.\3447\ NCAT 
recommended in their comment that EPA provide enhanced credits for EVs, 
PHEVs, and FCVs that are more stringent than California (and other 
States) ZEV mandates, making the national program credits 
``additional'' to state ZEV compliance credits.\3448\
---------------------------------------------------------------------------

    \3443\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
    \3444\ Edison Electric Institute, Detailed Comments, NHTSA-2018-
0067-11918.
    \3445\ Workhorse Group, Detailed Comments, NHTSA-2018-0067-
12215.
    \3446\ Volvo, Detailed Comments, NHTSA-2018-0067-12036.
    \3447\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
    \3448\ NCAT, Detailed Comments, NHTSA-2018-0067-11969.
---------------------------------------------------------------------------

    Northeast States for Coordinated Air Use Management (NESCAUM) 
commented that an aggressive reduction in emissions will not occur 
without national ZEV standards which will drive development of advanced 
clean vehicle technologies.\3449\
---------------------------------------------------------------------------

    \3449\ NESCAUM, Detailed Comments, NHTSA-2018-0067-11691.
---------------------------------------------------------------------------

    The NPRM did not propose or request comment on an NZEV concept or 
program, as such, and establishing such a program would be outside the 
scope of this rulemaking. Such a concept would require thorough 
assessment and full rulemaking notice and comment. There are also 
policy questions about what the appropriate level of potential 
incentives should be and whether certain technologies should receive 
greater incentives than other technologies, and if so, on what basis 
and by what amounts. Also, for the CAFE program, incentives for 
technologies are almost entirely prescribed by statute, and there are 
questions about how the CAFE program could implement an NZEV program in 
alignment with EPCA and EISA. Therefore, the agencies have decided not 
to implement an NZEV program as part of this rulemaking.
3. CO2 In-Use Requirements
    Current in-use regulations outlined in 86.1845-04 provide 
flexibility in determining the applicable number of test vehicles per 
test group. Each large volume manufacturer is provided the flexibility 
to employ small volume sampling allowances for a limited number of 
total annual production units. In response to the NPRM, Volkswagen is 
proposing to modify 86.1845-04 to provide a separate, additional small 
volume sampling allowance allocation of annual production volume for a 
manufacturer's plug-in hybrid vehicles. This additional allowance would 
only be applicable through the 2025 model year and would only be 
applicable to CO2 testing requirements under the in use 
regulations.
    The basis for this flexibility is rooted in the continuing 
evolution and development of traction drive battery cell chemistries 
and battery management systems. This ongoing development is aimed at 
continuously improving such features as energy density, power, cost, 
and durability. As such, the engineering processes for understanding 
and quantifying long-term performance are still developing and subject 
to reevaluation as new chemistries are examined. Manufacturers such as 
Volkswagen have allocated significant capital in battery testing to 
ensure that performance is maintained for consumers and are also 
providing longer term battery warranty provisions.
    Volkswagen believes that the targeted flexibility will provide 
additional time to continue evaluating chemistries and reduce 
administrative testing burdens for a very limited production allocation 
per manufacturer. This provision will further support plug-in hybrid 
technology development and deployment. Volkswagen proposed modifying 
86.1845-04 table SO4-07 footnote 2, to read as follows:
    \2\ Total annual production of groups eligible for testing under 
small volume sampling plan is capped at a maximum of 14,999 vehicle 49 
or 50 state annual sales, or a maximum of 4,500 vehicle California only 
sales per model year, per large volume manufacturer. Through model year 
2025, a separate total annual production of plug-in hybrid electric 
vehicle groups shall be eligible for testing under small volume 
sampling plan as described above. This allocation shall only be 
applicable to exhaust CO2 emission standards under this 
subpart.\3450\
---------------------------------------------------------------------------

    \3450\ See EPA-HQ-OAR-2018-0283-5689-A1, p.32.
---------------------------------------------------------------------------

    Regarding comments from VW on CO2 in-use requirements, 
EPA did not consider the change recommended by VW in the proposal and 
is not finalizing such a change. EPA believes the current program 
provides enough flexibility. EPA's general approach for this final rule 
is also to avoid providing incentives or other unique flexibilities to 
specific technologies.

[[Page 25247]]

F. Medium and Heavy-Duty Fuel Efficiency Technical Amendments

    NHTSA proposed in the NPRM to make minor technical revisions to 
correct typographical mistakes and improper references adopted in the 
agency's 2016 Phase 2 medium- and heavy-duty fuel efficiency 
rule.\3451\ The proposed changes were as follows:
---------------------------------------------------------------------------

    \3451\ 81 FR 73478 (Oct. 25, 2016).
---------------------------------------------------------------------------

     NHTSA heavy-duty vehicles and engine fuel consumption 
credit equations. In each credit equation in 49 CFR 535.7, the minus-
sign in each multiplication factor was omitted in the final version of 
the rule sent to the Federal Register. For example, the credit equation 
in Part 535.7(b)(1) should be specified as, Total MY Fleet FCC 
(gallons) = (Std-Act) x (Volume) x (UL) x (10-2) instead of (102), as 
currently exists. NHTSA proposed to correct these omissions.
     The CO2 to gasoline conversion factor: In 49 
CFR 535.6(a)(4)(ii) and (d)(5)(ii), NHTSA provides the methodology and 
equations for converting the CO2 FELs/FCLs for heavy-duty 
pickups and vans (gram per mile) and for engines (grams per hp-hr) to 
their gallon-of-gasoline equivalence. In each equation, NHTSA proposed 
to correct the conversion factor to 8,887 grams per gallon of gasoline 
fuel instead of a factor of 8,877 as currently specified.
     Curb weight definition: In 49 CFR 523.2, the reference in 
the definition for curb weight is incorrect. NHTSA proposed to correct 
the definition to incorporate a reference to 40 CFR 86.1803 instead of 
49 CFR 571.3.
    No public comments were received in response to NHTSA's proposed 
technical corrections. Therefore, NHTSA is finalizing these amendments 
and incorporating them into its heavy-duty regulations.

X. Regulatory Notices and Analyses

A. Executive Order 12866, Executive Order 13563

    Executive Order 12866, ``Regulatory Planning and Review'' (58 FR 
51735, Oct. 4, 1993), as amended by Executive Order 13563, ``Improving 
Regulation and Regulatory Review'' (76 FR 3821, Jan. 21, 2011), 
provides for making determinations whether a regulatory action is 
``significant'' and therefore subject to the Office of Management and 
Budget (OMB) review and to the requirements of the Executive Order. One 
comment requested that the agencies provide ``a far more robust cost/
benefit analysis as required by Executive Order (E.O.) 12866 and Office 
of Management and Budget Circular A-4.'' \3452\ The NPRM and this final 
rule satisfy the requirements of Executive Order 12866, ``Regulatory 
Planning and Review'' (58 FR 51735, Oct. 4, 1993), as amended by 
Executive Order 13563, ``Improving Regulation and Regulatory Review'' 
(76 FR 3821, Jan. 21, 2011). Under these Executive Orders, this action 
is an ``economically significant regulatory action'' because it is 
likely to have an annual effect on the economy of $100 million or more. 
Accordingly, EPA and NHTSA submitted this action to the OMB for review 
and any changes made in response to OMB recommendations have been 
documented in the docket for this action. The benefits and costs of 
this proposal are described above and in the Final Regulatory Impact 
Analysis (FRIA), which is located in the docket and on the agencies' 
websites.
---------------------------------------------------------------------------

    \3452\ See Anonymous Comment, Docket No. EPA-HQ-OAR-2018-0283-
3896, at 4-5 (footnote and citation omitted). As an example, the 
comment critiqued the NPRM's discussion of the ``diminishing 
returns'' of fuel economy benefits, alleging that the discussion 
``is not backed by reference to data or studies regarding how this 
conclusion was made.'' Id. at 5. Contrary to the comment's 
allegation, the conclusion is supported by the analysis from U.S. 
Energy Information Administration's (EIA's) Annual Energy Outlook 
(AEO) that was cited in the discussion. Id. As noted in the NPRM, 
the EIA--the statistical and analytical agency within the U.S. 
Department of Energy (DOE)--is the nation's premier source of energy 
information, and every fuel economy rulemaking since 2002 (and every 
joint CAFE and CO2 rulemaking since 2009) has applied 
fuel price projections from EIA's AEO. Id. at 42992 n.24.
---------------------------------------------------------------------------

B. DOT Regulatory Policies and Procedures

    The rule is also significant within the meaning of the Department 
of Transportation's Regulatory Policies and Procedures. The benefits 
and costs of this proposal are described above and in the FRIA, which 
is located in the docket and on NHTSA's website.

C. Executive Order 13771 (Reducing Regulation and Controlling 
Regulatory Costs)

    This rule is an E.O. 13771 deregulatory action. Per OMB Memorandum 
M-17-21, because this rule is deregulatory, it is not required to be 
offset by two deregulatory actions, as one comment suggested.\3453\
---------------------------------------------------------------------------

    \3453\ Anonymous Comment, Docket No. EPA-HQ-OAR-2018-0283-3896, 
at 8.
---------------------------------------------------------------------------

D. Executive Order 13211 (Energy Effects)

    Executive Order 13211 applies to any rule that: (1) is determined 
to be economically significant as defined under E.O. 12866, and is 
likely to have a significant adverse effect on the supply, 
distribution, or use of energy; or (2) that is designated by the 
Administrator of the Office of Information and Regulatory Affairs as a 
significant energy action. If the regulatory action meets either 
criterion, the agencies must evaluate the adverse energy effects of the 
rule and explain why the regulation is preferable to other potentially 
effective and reasonably feasible alternatives considered.
    The rule establishes passenger car and light truck fuel economy 
standards and tailpipe carbon dioxide and related emissions standards. 
An evaluation of energy effects of the action and reasonably feasible 
alternatives considered is provided in NHTSA's EIS and in the FRIA. To 
the extent that EPA's CO2 standards are substantially 
related to fuel economy and, accordingly, petroleum consumption, the 
EIS and FRIA analyses also provide an estimate of impacts of EPA's 
rule.

E. Environmental Considerations

1. National Environmental Policy Act (NEPA)
    Concurrently with this final rule, NHTSA is releasing a Final 
Environmental Impact Statement (FEIS), pursuant to the National 
Environmental Policy Act, 42 U.S.C. 4321-4347, and implementing 
regulations issued by the Council on Environmental Quality (CEQ), 40 
CFR part 1500, and NHTSA, 49 CFR part 520. NHTSA prepared the FEIS to 
analyze and disclose the potential environmental impacts of the 
proposed CAFE standards and a range of alternatives. The FEIS analyzes 
direct, indirect, and cumulative impacts and analyzes impacts in 
proportion to their significance. It describes potential environmental 
impacts to a variety of resources, including fuel and energy use, air 
quality, climate, land use and development, hazardous materials and 
regulated wastes, historical and cultural resources, noise, and 
environmental justice. The FEIS also describes how climate change 
resulting from global carbon emissions (including CO2 
emissions attributable to the U.S. light duty transportation sector 
under the alternatives considered) could affect certain key natural and 
human resources. Resource areas are assessed qualitatively and 
quantitatively, as appropriate, in the FEIS.
    Some commenters provided feedback on the ``flaws'' they identified 
in the CAFE model, concluding that because it played a significant role 
in modeling for the DEIS, the DEIS itself was flawed and

[[Page 25248]]

should be withdrawn and reissued.\3454\ The agencies address the 
comments regarding the CAFE model above in this preamble and in the 
FRIA. Ultimately, the findings on potential environmental impacts 
presented in the FEIS are of the same level of intensity and 
significance as those presented in the DEIS. While in some cases, the 
directionality of potential air quality emissions changed, the overall 
impact was generally small. NHTSA concludes that the CAFE model 
results, as used in the FEIS, do not result in the FEIS providing 
significant new information for the decisionmaker or the public 
compared to the DEIS.\3455\ NHTSA therefore concludes that a 
supplemental DEIS is not required.
---------------------------------------------------------------------------

    \3454\ States of California, Connecticut, Delaware, Hawaii, 
Iowa, Illinois, Maine, Maryland, Minnesota, North Carolina, New 
Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, and 
Washington; the Commonwealths of Massachusetts, Pennsylvania, and 
Virginia; the District of Columbia; and the Cities of Los Angeles, 
New York, Oakland, San Francisco, and San Jose (``California et. 
al.--Detailed NEPA Comments''), Docket No. NHTSA-2017-0069-0625, at 
6-11; Environmental Defense Fund, Docket No. NHTSA-2018-0067-11996, 
at 3-4; and Center for Biological Diversity, et al., Docket No. 
NHTSA-2018-0067-12123, at 19.
    \3455\ 40 CFR 1502.9(c)(1)(ii).
---------------------------------------------------------------------------

    NHTSA also performed a national-scale photochemical air quality 
modeling and health benefit assessment for the FEIS; it is included as 
Appendix E. The purpose of this assessment was to use air quality 
modeling and health-related benefits analysis tools to examine the 
potential air quality-related consequences of the alternatives 
considered in its Draft Environmental Impact Statement (DEIS). In a 
comment on the DEIS, the South Coast Air Quality Management District 
stated that performing the photochemical modeling for the FEIS ``comes 
too late for the public to be able to comment on that analysis,'' and 
that the EIS must be recirculated to allow such public comment.\3456\ 
However, NHTSA publicly stated its intent to conduct the analysis as 
part of the FEIS in its scoping notice published on July 26, 
2017.\3457\ The agency noted that this approach was consistent with 
past practice and resulted from the substantial time required to 
complete such an analysis. NHTSA also announced that, due to the 
substantial lead time required, the analysis would be based on the 
modeling of the alternatives presented in the DEIS, not of the 
alternatives as presented in the FEIS. NHTSA received no public 
comments in response to the scoping notice addressing this analytical 
approach, and the agency proceeded accordingly. Furthermore, while 
photochemical modeling provides spatial and temporal detail for 
estimating changes in ambient levels of air pollutants and their 
associated impacts on human health and welfare, the analysis affirms 
the estimates that appear in the EIS and does not provide significant 
new information for the decisionmaker or the public. For these reasons, 
NHTSA concludes that inclusion of the photochemical modeling and health 
benefit assessment in the FEIS is appropriate, and recirculation of the 
EIS is not required.
---------------------------------------------------------------------------

    \3456\ South Coast Air Quality Management District, Docket No. 
NHTSA-2018-0067-5666, at 10. See also North Carolina Department of 
Environmental Quality, Docket No. NHTSA-2018-0067-12025, at 35-37.
    \3457\ NHTSA, ``Notice of Intent to Prepare an Environmental 
Impact Statement for Model Year 2022-2025 Corporate Average Fuel 
Economy Standards,'' 82 FR 34740, 34743 fn. 15 (Jul. 26, 2017).
---------------------------------------------------------------------------

    NHTSA has considered the information contained in the FEIS in 
making the final decision described in this final rule.\3458\ This 
preamble and final rule constitute NHTSA's Record of Decision (ROD) 
under 40 CFR 1505.2 for its promulgation of CAFE standards for MYs 
2021-2026. NHTSA has authority to issue its FEIS and ROD simultaneously 
pursuant to 49 U.S.C. 304a(b) and U.S. Department of Transportation, 
Office of Transportation Policy, Guidance on the Use of Combined Final 
Environmental Impact Statements/Records of Decision and Errata Sheets 
in National Environmental Policy Act Reviews (April 25, 2019).\3459\ 
NHTSA has determined that neither the statutory criteria nor 
practicability considerations preclude simultaneous issuance.
---------------------------------------------------------------------------

    \3458\ The FEIS is available for review in the public docket for 
this action and in Docket No. NHTSA-2017-0069.
    \3459\ The guidance is available at https://www.transportation.gov/sites/dot.gov/files/docs/mission/transportation-policy/permittingcenter/337371/feis-rod-guidance-final-04302019.pdf.
---------------------------------------------------------------------------

    As required by the CEQ regulations,\3460\ this final rule (as the 
ROD) sets forth the following: (1) The agency's decision (Sections V 
and VIII above); (2) alternatives considered by NHTSA in reaching its 
decision, including the environmentally preferable alternative 
(Sections V, VII, and VIII above); (3) the factors balanced by NHTSA in 
making its decision, including essential considerations of national 
policy (Section VIII.B above); (4) how these factors and considerations 
entered into its decision (Section VIII.B above); and (5) the agency's 
preferences among alternatives based on relevant factors, including 
economic and technical considerations and agency statutory missions 
(Section VIII.B.4 above). This section also briefly addresses 
mitigation\3461\ and whether all practicable means to avoid or minimize 
environmental harm from the alternative selected have been adopted.
---------------------------------------------------------------------------

    \3460\ 40 CFR 1505.2.
    \3461\ See 40 CFR 1508.20(b) (``Mitigation includes . . . (b) 
Minimizing impacts by limiting the degree or magnitude of the action 
and its implementation. . .'')
---------------------------------------------------------------------------

    In the DEIS and in the FEIS, the agency identified a Preferred 
Alternative. In the DEIS, the Preferred Alternative was identified as 
Alternative 1 (0.0 Percent Annual Increase in Fuel Economy, MYs 2021-
2026), which were the standards the agency proposed in the NPRM. In the 
FEIS, the Preferred Alternative was identified as Alternative 3 (1.5 
Percent Annual Increase in Fuel Economy, MYs 2021-2026). As the FEIS 
notes, under the Preferred Alternative, on an mpg basis, the estimated 
annual increases in the average required fuel economy levels between 
MYs 2021 and 2026 is 1.5 percent for both passenger cars and light 
trucks.\3462\ After carefully reviewing and analyzing all of the 
information in the public record, the FEIS, and comments submitted on 
the DEIS and the NPRM, NHTSA has decided to finalize the Preferred 
Alternative described in the FEIS for the reasons described in this 
ROD.
---------------------------------------------------------------------------

    \3462\ Because the standards are attribute-based, average 
required fuel economy levels, and therefore rates of increase in 
those average mpg values, depend on the future composition of the 
fleet, which is uncertain and subject to change. When NHTSA 
describes a percent increase in stringency, we mean in terms of 
shifts in the footprint functions that form the basis for the actual 
CAFE standards (as in, on a gallon per mile basis, the CAFE 
standards change by a given percentage from one model year to the 
next).
---------------------------------------------------------------------------

    NHTSA has considered environmental considerations as part of its 
balancing of the statutory factors to set maximum feasible fuel economy 
standards. As a result, the agency has limited the degree or magnitude 
of the action as appropriate in light of its statutory 
responsibilities. NHTSA's authority to promulgate fuel economy 
standards does not allow it to regulate criteria polluants from 
vehicles or refineries, nor can NHTSA regulate other factors affecting 
those emissions, such as driving habits. Consequently, NHTSA must set 
CAFE standards but is unable to take further steps to mitigate the 
impacts of these standards. Chapter 9 of the FEIS provides a further 
discussion of mitigation measures in the context of NEPA.
    One commenter states that NHTSA, at a minimum, ``must include a 
thorough discussion of all reasonable mitigation measures and detail 
the appropriate agencies that could implement such

[[Page 25249]]

measures.'' \3463\ As examples, the commenter listed: ``creating tax 
breaks for transit and biking, expanding transportation demand 
management programs for federal employees, implementing a social 
marketing campaign regarding VMT reduction, increasing dedicated 
funding for transit and active modes, requiring VMT as a performance 
measure for federal funding, and providing NEPA guidance on evaluating 
VMT impacts of federal projects.'' Each of the examples listed is 
beyond NHTSA's statutory authority. Furthermore, documenting the myriad 
measures that could reduce VMT or address criteria pollutant or carbon 
dioxide emissions would provide no added benefit to the decisionmaker 
or the public. Each of these actions requires their own extensive cost-
benefit anlaysis, are beyond the purview of this action, and are beyond 
the legal responsibility of NHTSA. NHTSA concludes that the commenter's 
request is beyond the bounds of NEPA's ``rule of reason.'' \3464\
---------------------------------------------------------------------------

    \3463\ California et. al.--Detailed NEPA Comments, Docket No. 
NHTSA-2017-0069-0625, at 31.
    \3464\ Dep't of Transp. v. Pub. Citizen, 541 U.S. 752, 772 
(2004).
---------------------------------------------------------------------------

    Another commenter disputes NHTSA's conclusion that it lacks 
statutory authority to mitigate the impacts of its CAFE standards. 
Specifically, the commenter cites to its very authority to set fuel 
economy standards: ``It is axiomatic that fuel efficiency standards set 
at levels of the No Action Alternative or at more stringent levels 
would eliminate the additional pollution created by the proposed 
freeze.'' \3465\ This, however, mischaracterizes mitigation as nothing 
more than a choice among alternatives. NHTSA is already considering a 
range of reasonable alternatives and has concluded that alternatives 
more stringent than the No Action Alternative are beyond reasonable. 
Furthermore, NHTSA disputes that more stringent fuel economy standards 
will axiomatically lead to lower levels of criteria pollutant 
emissions. In fact, because of the rebound effect, higher levels of 
stringency may result in higher VMT, which may result in criteria 
pollutant emission increases.
---------------------------------------------------------------------------

    \3465\ Center for Biological Diversity, et al., Docket No. 
NHTSA-2018-0067-12123, at 55-56.
---------------------------------------------------------------------------

    The North Carolina Department of Environmental Quality commented 
that the proposed changes to the CAFE standards could undermine the 
integrity of many of the assumptions in various NEPA documents across 
the United States, in part because EPA required the use of the 
MOVES2014 model (or a subsequent revision) for transportation 
conformity determinations.\3466\ That version of MOVES incorporates 
CAFE and CO2 standards based on the agencies' actions in 
2012 and does not reflect the actions being finalized in this rule. The 
implication of the commenter's assertion, however, is that neither 
NHTSA nor EPA could take any regulatory action regarding CAFE or 
CO2 standards, regardless of whether such action was to 
increase or decrease such standards. Clearly neither agency can be 
paralyzed from undertaking its statutory obligations because of the 
independent NEPA obligations related to other ongoing Federal actions. 
For those actions, responsible officials may need to assess whether 
this final rule triggers the need for a supplemental NEPA document. 
However, it is not unique for Federal agencies to take actions or for 
new information to become available that affects the underlying inputs 
in models, such as EPA's MOVES model, on which NEPA and conformity 
analyses rely. Over time, those models will be updated to reflect these 
actions and information. EPA is responsible for approving the 
availability of models for the use in State implementation plans and 
transportation conformity analyses. EPA will evaluate and address, as 
appropriate, the impact of this action on future SIP approval actions. 
Currently approved emission factor models remain approved for SIPs and 
transportation conformity analyses, and EPA will work with DOT on the 
appropriate implementation of Federal requirements based on current and 
available information.
---------------------------------------------------------------------------

    \3466\ North Carolina Department of Environmental Quality, 
Docket No. NHTSA-2018-0067-12025, at 37. See also Southern 
Environmental Law Center, EPA-HQ-OAR-2018-0283-0887, at 2-4.
---------------------------------------------------------------------------

2. Clean Air Act (CAA) as Applied to NHTSA's Action
    The CAA (42 U.S.C.[thinsp]7401 et seq.) is the primary Federal 
legislation that addresses air quality. Under the authority of the CAA 
and subsequent amendments, EPA has established National Ambient Air 
Quality Standards (NAAQS) for six criteria pollutants, which are 
specifically identified pollutants that have recognized adverse effects 
on ambient air quality and that can accumulate in the atmosphere as a 
result of human activity. EPA is required to review each NAAQS every 
five years and to revise those standards as may be appropriate 
considering new scientific information.
    The air quality of a geographic region is usually assessed by 
comparing the levels of criteria air pollutants found in the ambient 
air to the levels established by the NAAQS (taking into account, as 
well, the other elements of a NAAQS: averaging time, form, and 
indicator). Concentrations of criteria pollutants within the air mass 
of a region are measured in parts of a pollutant per million parts 
(ppm) of air or in micrograms of a pollutant per cubic meter ([mu]g/
m\3\) of air present in repeated air samples taken at designated 
monitoring locations using specified types of monitors. These ambient 
concentrations of each criteria pollutant are compared to the levels, 
averaging time, and form specified by the NAAQS in order to assess 
whether the region's air quality is in attainment with the NAAQS.
    When the measured concentrations of a criteria pollutant within a 
geographic region are below those permitted by the NAAQS, EPA 
designates the region as an attainment area for that pollutant, while 
regions where concentrations of criteria pollutants exceed Federal 
standards are called nonattainment areas. Former nonattainment areas 
that are now in compliance with the NAAQS are designated as maintenance 
areas. Each State with a nonattainment area is required to develop and 
implement a State Implementation Plan (SIP) documenting how the region 
will reach attainment levels within time periods specified in the CAA. 
For maintenance areas, the SIP must document how the State intends to 
maintain compliance with the NAAQS. When EPA revises a NAAQS, each 
State must revise its SIP to address how it plans to attain the new 
standard.
    No Federal agency may ``engage in, support in any way or provide 
financial assistance for, license or permit, or approve'' any activity 
that does not ``conform'' to a SIP or Federal Implementation Plan after 
EPA has approved or promulgated it.\3467\ Further, no Federal agency 
may ``approve, accept, or fund'' any transportation plan, program, or 
project developed pursuant to title 23 or chapter 53 of title 49, 
U.S.C., unless the plan, program, or project has been found to 
``conform'' to any applicable implementation plan in effect.\3468\ The 
purpose of these conformity requirements is to ensure that Federally 
sponsored or conducted activities do not interfere with meeting the 
emissions targets in SIPs, do not cause or contribute to new violations 
of the NAAQS, and do not impede the ability of a State to attain or 
maintain the NAAQS or delay any interim milestones. EPA has issued two 
sets of

[[Page 25250]]

regulations to implement the conformity requirements:
---------------------------------------------------------------------------

    \3467\ 42 U.S.C. 7506(c)(1).
    \3468\ 42 U.S.C. 7506(c)(2).
---------------------------------------------------------------------------

    (1) The Transportation Conformity Rule\3469\ applies to 
transportation plans, programs, and projects that are developed, 
funded, or approved under title 23 or chapter 53 of title 49, U.S.C.
---------------------------------------------------------------------------

    \3469\ 40 CFR part 51, subpart T, and part 93, subpart A.
---------------------------------------------------------------------------

    (2) The General Conformity Rule\3470\ applies to all other federal 
actions not covered under transportation conformity. The General 
Conformity Rule establishes emissions thresholds, or de minimis levels, 
for use in evaluating the conformity of an action that results in 
emissions increases.\3471\ If the net increases of direct and indirect 
emissions are lower than these thresholds, then the project is presumed 
to conform and no further conformity evaluation is required. If the net 
increases of direct and indirect emissions exceed any of these 
thresholds, and the action is not otherwise exempt, then a conformity 
determination is required. The conformity determination can entail air 
quality modeling studies, consultation with EPA and state air quality 
agencies, and commitments to revise the SIP or to implement measures to 
mitigate air quality impacts.
---------------------------------------------------------------------------

    \3470\ 40 CFR part 51, subpart W, and part 93, subpart B.
    \3471\ 40 CFR 93.153(b).
---------------------------------------------------------------------------

    The CAFE standards and associated program activities are not 
developed, funded, or approved under title 23 or chapter 53 of title 
49, United States Code. Accordingly, this action and associated program 
activities are not subject to the Transportation Conformity Rule. Under 
the General Conformity Rule, a conformity determination is required 
where a Federal action would result in total direct and indirect 
emissions of a criteria pollutant or precursor originating in 
nonattainment or maintenance areas equaling or exceeding the rates 
specified in 40 CFR 93.153(b)(1) and (2). As explained below, NHTSA's 
action results in neither direct nor indirect emissions as defined in 
40 CFR 93.152.
    The General Conformity Rule defines direct emissions as ``those 
emissions of a criteria pollutant or its precursors that are caused or 
initiated by the Federal action and originate in a nonattainment or 
maintenance area and occur at the same time and place as the action and 
are reasonably foreseeable.'' \3472\ Because NHTSA's action would set 
fuel economy standards for light duty vehicles, it would cause no 
direct emissions consistent with the meaning of the General Conformity 
Rule.\3473\
---------------------------------------------------------------------------

    \3472\ 40 CFR 93.152.
    \3473\ Dep't of Transp. v. Pub. Citizen, 541 U.S. at 772 
(``[T]he emissions from the Mexican trucks are not `direct' because 
they will not occur at the same time or at the same place as the 
promulgation of the regulations.''). NHTSA's action is to establish 
fuel economy standards for MY 2021-2026 passenger car and light 
trucks; any emissions increases would occur in a different place and 
well after promulgation of the final rule.
---------------------------------------------------------------------------

    Indirect emissions under the General Conformity Rule are ``those 
emissions of a criteria pollutant or its precursors (1) That are caused 
or initiated by the federal action and originate in the same 
nonattainment or maintenance area but occur at a different time or 
place as the action; (2) that are reasonably foreseeable; (3) that the 
agency can practically control; and (4) for which the agency has 
continuing program responsibility.'' \3474\ Each element of the 
definition must be met to qualify as indirect emissions. NHTSA has 
determined that, for purposes of general conformity, emissions that may 
result from its final fuel economy standards would not be caused by 
NHTSA's action, but rather would occur because of subsequent activities 
the agency cannot practically control. ``[E]ven if a Federal licensing, 
rulemaking, or other approving action is a required initial step for a 
subsequent activity that causes emissions, such initial steps do not 
mean that a Federal agency can practically control any resulting 
emissions.'' \3475\
---------------------------------------------------------------------------

    \3474\ 40 CFR 93.152.
    \3475\ 40 CFR 93.152.
---------------------------------------------------------------------------

    As the CAFE program uses performance-based standards, NHTSA cannot 
control the technologies vehicle manufacturers use to improve the fuel 
economy of passenger cars and light trucks. Furthermore, NHTSA cannot 
control consumer purchasing (which affects average achieved fleetwide 
fuel economy) and driving behavior (i.e., operation of motor vehicles, 
as measured by VMT). It is the combination of fuel economy 
technologies, consumer purchasing, and driving behavior that results in 
criteria pollutant or precursor emissions. For purposes of analyzing 
the environmental impacts of the alternatives considered here and under 
NEPA, NHTSA has made assumptions regarding all of these factors. The 
agency's FEIS predicts that increases in air toxic and criteria 
pollutants would occur in some nonattainment areas under certain 
alternatives. However, the standards and alternatives do not mandate 
specific manufacturer decisions, consumer purchasing, or driver 
behavior, and NHTSA cannot practically control any of them.\3476\
---------------------------------------------------------------------------

    \3476\ See, e.g., Dep't of Transp. v. Pub. Citizen, 541 U.S. 
752, 772-73 (2004); S. Coast Air Quality Mgmt. Dist. v. Fed. Energy 
Regulatory Comm'n, 621 F.3d 1085, 1101 (9th Cir. 2010).
---------------------------------------------------------------------------

    In addition, NHTSA does not have the statutory authority to control 
the actual VMT by drivers. As the extent of emissions is directly 
dependent on the operation of motor vehicles, changes in any emissions 
that result from NHTSA's CAFE standards are not changes the agency can 
practically control or for which the agency has continuing program 
responsibility. Therefore, the final CAFE standards and alternative 
standards considered by NHTSA would not cause indirect emissions under 
the General Conformity Rule, and a general conformity determination is 
not required.
    As this analysis was presented in the NPRM, some commenters 
disagreed with NHTSA's conclusion. One commenter cited two reasons for 
concluding that the General Conformity Rule applies to NHTSA's 
action.\3477\ First, the commenter argues that NHTSA used 
``inappropriate modeling'' in its analysis. However, this is irrelevant 
to the agency's analysis, which is based on the Federal regulations and 
the applicable case law. Second, the commenter asserts that NHTSA 
``cannot have it both ways'' by alleging that it cannot control the 
technologies that automobile manufacturers would use or consumer 
purchasing behavior, yet justifies its rulemakings based on consumer 
purchasing and emissions implications.3478 3479 The 
rulemaking analysis presents a feasible pathway for manufacturers to 
comply with the rules, based on a series of assumptions about consumer 
behavior; it is not sufficiently foreseeable to trigger application of 
the General Conformity Rule. Furthermore, NHTSA cannot directly control 
these behaviors, and the chain of causation is too attenuated to be 
responsible for the resulting emissions. Another commenter stated that 
NHTSA has continuing

[[Page 25251]]

program responsibility for motor vehicle criteria pollutant emissions 
because it ``retain[s] authority to revise [its] standards in a way 
that affects future emission levels.'' \3480\ However, NHTSA disagrees 
with this assertion. First, the agency does not have statutory 
authority to regulate criteria pollutant emissions from motor vehicles. 
Second, the fact that NHTSA could establish CAFE standards for 
separate, future motor vehicles does not establish continuing program 
responsibility over emissions that could result from the vehicles 
regulated by this action.
---------------------------------------------------------------------------

    \3477\ California et. al.--Detailed NEPA Comments, Docket No. 
NHTSA-2017-0069-0625, at 21-22.
    \3478\ The commenter also quotes CBD v. NHTSA, 538 F.3d at 1217, 
for the proposition that NHTSA's regulations are the proximate cause 
of the emissions because they allow particular fuel economy levels 
that ``translate directly into particular tailpipe emissions.'' 
However, that quote was referencing carbon dioxide emissions, which 
are predictable based on fuel used. NHTSA can directly regulate fuel 
economy for passenger cars and light trucks. On the other hand, 
criteria pollutant emissions are more significantly impacted by VMT, 
technology choices, and other factors that are not directly within 
the control of NHTSA.
    \3479\ See also Joint Submission from the States of California 
et al. and the Cities of Oakland et al., Docket No. NHTSA-2018-0067-
11735, at 35.
    \3480\ Id.
---------------------------------------------------------------------------

    NHTSA and EPA further discuss their obligations under the General 
Conformity Rule, and further address comments received, in Section 
VI.D.3 above.
3. National Historic Preservation Act (NHPA)
    The NHPA (54 U.S.C. 300101 et seq.) sets forth government policy 
and procedures regarding ``historic properties''--that is, districts, 
sites, buildings, structures, and objects included on or eligible for 
the National Register of Historic Places. Section 106 of the NHPA 
requires Federal agencies to ``take into account'' the effects of their 
actions on historic properties.\3481\ In the NPRM, the agencies 
concluded that the NHPA is not applicable to this rulemaking because 
the promulgation of CAFE and CO2 emissions standards for 
light duty vehicles is not the type of activity that has the potential 
to cause effects on historic properties.
---------------------------------------------------------------------------

    \3481\ Section 106 is now codified at 54 U.S.C. 306108. 
Implementing regulations for the Section 106 process are located at 
36 CFR part 800.
---------------------------------------------------------------------------

    Two commenters wrote that ``[c]limate change and air pollution 
imperil historic properties throughout the country via direct 
degradation, sea level rise, fire, flood, and other forms of harm.'' 
Therefore, the commenters concluded that NHTSA and EPA must consult 
with the relevant Federal and State authorities and fully disclose any 
impacts to historic properties.\3482\ However, as this final rule 
establishes CAFE and CO2 standards that increase each year 
for MYs 2021-2026, this action will result in reductions in climate 
change-related impacts and most air pollutants compared to the absence 
of regulation. Furthermore, any impacts to particular historic 
properties that could be related to emissions changes associated with 
this rulemaking are not reasonably certain to occur, would be de 
minimis in their level of impact if they did occur, and are too 
attenuated to be attributed directly to this action. (See also Section 
X.E.6 below.) There is no evidence that the changes in air pollution or 
CO2 emissions associated with this rulemaking, in and of 
themselves, would alter the characteristics of a historic property 
qualifying it for inclusion in or eligibility for the National 
Register.\3483\ Nevertheless, NHTSA includes a brief, qualitative 
discussion of the impacts of the alternatives on historical and 
cultural resources in Section 7.3 of the FEIS. For the foregoing 
reasons, the agencies continue to conclude that any potential impacts 
have been accounted for in the associated analyses of this rulemaking 
and that no consultation is required under the NHPA.
---------------------------------------------------------------------------

    \3482\ CARB, Docket No. NHTSA-2018-0067-11873, at 411; 
California et. al.--Detailed NEPA Comments, Docket No. NHTSA-2017-
0069-0625, at 30.
    \3483\ 36 CFR 800.16(i).
---------------------------------------------------------------------------

4. Fish and Wildlife Conservation Act (FWCA)
    The FWCA (16 U.S.C. 2901 et seq.) provides financial and technical 
assistance to States for the development, revision, and implementation 
of conservation plans and programs for nongame fish and wildlife. In 
addition, the Act encourages all Federal departments and agencies to 
utilize their statutory and administrative authorities to conserve and 
to promote conservation of nongame fish and wildlife and their 
habitats. The agencies conclude that the FWCA is not applicable to this 
final rule because this rulemaking does not involve the conservation of 
nongame fish and wildlife and their habitats. NHTSA has, however, 
conducted a qualitative review in its FEIS of the related direct, 
indirect, and cumulative impacts, positive or negative, of the 
alternatives on potentially affected resources, including nongame fish 
and wildlife and their habitats.
5. Coastal Zone Management Act (CZMA)
    The Coastal Zone Management Act (16 U.S.C. 1451 et seq.) provides 
for the preservation, protection, development, and (where possible) 
restoration and enhancement of the Nation's coastal zone resources. 
Under the statute, States are provided with funds and technical 
assistance in developing coastal zone management programs. Each 
participating State must submit its program to the Secretary of 
Commerce for approval. Once the program has been approved, any activity 
of a Federal agency, either within or outside of the coastal zone, that 
affects any land or water use or natural resource of the coastal zone 
must be carried out in a manner that is consistent, to the maximum 
extent practicable, with the enforceable policies of the State's 
program.\3484\
---------------------------------------------------------------------------

    \3484\ 16 U.S.C. 1456(c)(1)(A).
---------------------------------------------------------------------------

    In the NPRM, the agencies concluded that the CZMA is not applicable 
to this rulemaking because this rulemaking does not involve an activity 
within, or outside of, the Nation's coastal zones that affects any land 
or water use or natural resource of the coastal zone. CARB commented 
that California's coast is vulnerable to sea level rise from climate 
change and that the proposal would exacerbate that threat. Therefore, 
the commenter claimed that the proposal violated California's policies 
and obligations in its management program to preserve, protect, and 
enhance its coastline.\3485\ However, in its FEIS, NHTSA estimates that 
the sea-level rise in 2100 associated with Alternative 1 (0 percent 
annual average increase for both passenger cars and light trucks for 
MYs 2021-2026), the least stringent alternative considered, would be 
0.7 mm. Such a level is too small to have any meaningful impact on land 
or water use or a natural resource of the coastal zone. Furthermore, as 
this final rule establishes CAFE and CO2 standards that 
increase each year for MYs 2021-2026, this action will result in 
reductions in sea level rise resulting from climate change compared to 
the absence of regulation. Therefore, the agencies continue to conclude 
that the CZMA is not applicable to this rulemaking. NHTSA has, however, 
conducted a qualitative review in its FEIS of the related direct, 
indirect, and cumulative impacts, positive or negative, of the 
alternatives on potentially affected resources, including coastal 
zones.
---------------------------------------------------------------------------

    \3485\ CARB, Docket No. NHTSA-2018-0067-11873, at 411.
---------------------------------------------------------------------------

6. Endangered Species Act (ESA)
    Under Section 7(a)(2) of the Endangered Species Act (ESA), Federal 
agencies must ensure that actions they authorize, fund, or carry out 
are ``not likely to jeopardize the continued existence'' of any 
Federally listed threatened or endangered species (collectively, 
``listed species'') or result in the destruction or adverse 
modification of the designated critical habitat of these species.\3486\ 
In general, if a Federal agency determines that an agency action may 
affect a listed species or designated critical habitat, it must 
initiate consultation with the

[[Page 25252]]

appropriate Service--the U.S. Fish and Wildlife Service (FWS) of the 
Department of the Interior (DOI) and/or the National Oceanic and 
Atmospheric Administration's National Marine Fisheries Service (NMFS) 
of the Department of Commerce (together, ``the Services''), depending 
on the species involved--in order to ensure that the action is not 
likely to jeopardize the species or destroy or adversely modify 
designated critical habitat.\3487\ Under this standard, the Federal 
agency taking action evaluates the possible effects of its action and 
determines whether to initiate consultation.\3488\
---------------------------------------------------------------------------

    \3486\ 16 U.S.C. 1536(a)(2).
    \3487\ See 50 CFR 402.14.
    \3488\ See 50 CFR 402.14(a) (``Each Federal agency shall review 
its actions at the earliest possible time to determine whether any 
action may affect listed species or critical habitat.'').
---------------------------------------------------------------------------

    In the NPRM, the agencies noted that they had considered the 
effects of the proposed standards and alternatives in light of 
applicable ESA regulations, case law, and guidance to determine what, 
if any, impact there might be to listed species or designated critical 
habitat. The agencies also considered the discussion in the DEIS, where 
NHTSA incorporated by reference its response to a public comment on 
page 9-101 of the MY 2017-2025 CAFE Standards Final EIS.\3489\ Based on 
that assessment, the agencies determined that the actions of setting 
CAFE and CO2 emissions standards did not require 
consultation under Section 7(a)(2) of the ESA. Accordingly, the 
agencies wrote that they had concluded their review of this action 
under Section 7 of the ESA.
---------------------------------------------------------------------------

    \3489\ For the final rule for MY 2017 and beyond CAFE standards, 
NHTSA concluded that a Section 7(a)(2) consultation was not required 
because any potential for a specific impact on particular listed 
species and their habitats associated with emission changes achieved 
by that rulemaking were too uncertain and remote to trigger the 
threshold for such a consultation. In the Draft EIS, NHTSA wrote 
that this conclusion, based on the discussion and analysis cited, 
applied equally to the current rulemaking.
---------------------------------------------------------------------------

    Several commenters disagreed with the agencies' assessment. In 
general, commenters stated that the agencies' proposed action would 
increase emissions of CO2 and criteria air pollutants (e.g., 
nitrogen oxide [NOX] and sulfur dioxide 
[SO2]\3490\), that these emissions would have direct or 
indirect (i.e., through climate change) impacts on listed species and 
critical habitats, that the threshold for a finding of ``may affect'' 
is extremely low, and that the agencies therefore have a duty to 
consult with the Services under the ESA.\3491\
---------------------------------------------------------------------------

    \3490\ In fact, in Section 4.2.1.1 of NHTSA's FEIS, the agency 
reports that any of the action alternatives would result in 
decreased emissions of sulfur dioxide in 2025, 2035, and 2050 
compared to the No Action Alternative.
    \3491\ See Center for Biological Diversity, Earthjustice, 
Natural Resources Defense Council, and Sierra Club, Docket Nos. 
NHTSA-2017-0069-0605 and NHTSA-2018-0067-12127; Center for 
Biological Diversity, Sierra Club, and Public Citizen, Inc., Docket 
No. NHTSA-2018-0067-12378; Center for Biological Diversity, 
Earthjustice, Environmental Law and Policy Center, Natural Resources 
Defense Council, Public Citizen, Inc., Safe Climate Campaign, Sierra 
Club, Southern Environmental Law Center, and Union of Concerned 
Scientists, Docket No. NHTSA-2018-0067-12123, at 69; States of 
California, Connecticut, Delaware, Hawaii, Iowa, Illinois, Maine, 
Maryland, Minnesota, New Jersey, New Mexico, New York, North 
Carolina, Oregon, Rhode Island, Vermont, and Washington, the 
Commonwealths of Massachusetts, Pennsylvania, and Virginia, the 
District of Columbia, and the Cities of Los Angeles, New York, 
Oakland, San Francisco, and San Jose, Docket Nos. NHTSA-2018-0067-
11735, at 47-48; and California Air Resources Board, Docket Nos. 
NHTSA-2018-0067-11873, at 411.
---------------------------------------------------------------------------

    In light of these comments, the agencies re-evaluated their 
obligations under the ESA and applicable regulations, case law, and 
guidance. Ultimately, for the following reasons, the agencies arrive at 
the same conclusion. Although there is a general association between 
the actions undertaken in this final rule and environmental impacts, as 
described in this preamble and the FEIS, the agencies' actions result 
in no effects on listed species or designated critical habitat and 
therefore do not require consultation under Section 7(a)(2) of the ESA. 
Furthermore, the agencies lack sufficient discretion or control to 
bring these actions under the consultation requirement of the ESA. The 
agencies' review under the ESA is concluded.
a) The Agencies' Actions Have No Effects on Listed Species or Critical 
Habitat and Do Not Trigger ESA Consultation
    Commenters have stated that CO2 and criteria air 
pollutant emissions are relevant to Section 7(a)(2) consultation 
because of the potential impacts of climate change or the pollutants 
themselves on listed species or critical habitat. The agencies have 
considered the potential impacts of this action to listed species or 
designated critical habitat of these species and conclude that any such 
impacts cannot be attributed to the agencies' actions (e.g., they are 
too uncertain and attenuated). Because the agencies conclude there are 
``no effects,'' Section 7(a)(2) consultation is not required. The 
agencies base this conclusion both on the language of the Section 
7(a)(2) implementing regulations and on the long history of actions and 
guidance provided by DOI.
    The Section 7(a)(2) implementing regulations require consultation 
if a Federal agency determines its action ``may affect'' listed species 
or critical habitat.\3492\ The recently revised regulations define 
``effects of the action'' as ``all consequences to listed species or 
critical habitat that are caused by the proposed action, including the 
consequences of other activities that are caused by the proposed 
action. A consequence is caused by the proposed action if it would not 
occur but for the proposed action and it is reasonably certain to 
occur.'' \3493\ The revised definition made explicit a ``but for'' test 
and the concept of ``reasonably certain to occur'' for all 
effects.\3494\ However, in the preamble to the final rule, the Services 
emphasized that the ``but for'' test and ``reasonably certain to 
occur'' are not new or heightened standards.\3495\ In this context, 
```but for' causation means that the consequence in question would not 
occur if the proposed action did not go forward . . . . In other words, 
if the agency fails to take the proposed action and the activity would 
still occur, there is no `but for' causation. In that event, the 
activity would not be considered an effect of the action under 
consultation.'' \3496\
---------------------------------------------------------------------------

    \3492\ 50 CFR 402.14(a). The Services recently issued a final 
rule revising the regulations governing the ESA Section 7 
consultation process. 84 FR 44976 (Aug. 27, 2019). The effective 
date of the new regulations was subsequently delayed to October 28, 
2019. 84 FR 50333 (Sep. 25, 2019). As discussed in the text that 
follows, the agencies believe that their conclusion would be the 
same under both the current and prior regulations.
    \3493\ 50 CFR 402.02 (emphasis added), as amended by 84 FR 
44976, 45016 (Aug. 27, 2019).
    \3494\ The Services' prior regulations defined ``effects of the 
action'' in relevant part as ``the direct and indirect effects of an 
action on the species or critical habitat, together with the effects 
of other activities that are interrelated or interdependent with 
that action, that will be added to the environmental baseline.'' 50 
CFR 402.02 (as in effect prior to Oct. 28, 2019). Indirect effects 
were defined as ``those that are caused by the proposed action and 
are later in time, but still are reasonably certain to occur.'' Id.
    \3495\ 84 FR at 44977 (``As discussed in the proposed rule, the 
Services have applied the `but for' test to determine causation for 
decades. That is, we have looked at the consequences of an action 
and used the causation standard of `but for' plus an element of 
foreseeability (i.e., reasonably certain to occur) to determine 
whether the consequence was caused by the action under 
consultation.'').
    \3496\ Id. We note that as the Services do not consider this to 
be a change in their longstanding application of the ESA, this 
interpretation applies equally under the prior regulations (which 
were effective through October 28, 2019, and the current 
regulations.
---------------------------------------------------------------------------

    The revised ESA regulations also provide a framework for 
determining whether consequences are caused by a proposed action and 
are therefore ``effects'' that may trigger consultation. The 
regulations provide in part:


[[Page 25253]]


    To be considered an effect of a proposed action, a consequence 
must be caused by the proposed action (i.e., the consequence would 
not occur but for the proposed action and is reasonably certain to 
occur). A conclusion of reasonably certain to occur must be based on 
clear and substantial information, using the best scientific and 
commercial data available. Considerations for determining that a 
consequence to the species or critical habitat is not caused by the 
proposed action include, but are not limited to:
    (1) The consequence is so remote in time from the action under 
consultation that it is not reasonably certain to occur; or
    (2) The consequence is so geographically remote from the 
immediate area involved in the action that it is not reasonably 
certain to occur; or
    (3) The consequence is only reached through a lengthy causal 
chain that involves so many steps as to make the consequence not 
reasonably certain to occur.\3497\
---------------------------------------------------------------------------

    \3497\ 50 CFR 402.17(b).

The regulations go on to make clear that the action agency must factor 
these considerations into its assessments of potential effects.\3498\
---------------------------------------------------------------------------

    \3498\ 50 CFR 402.17(c) (``Required consideration. The 
provisions in paragraphs (a) and (b) of this section must be 
considered by the action agency and the Services.'').

    DOI, the agency charged with co-administering the ESA, previously 
evaluated whether CO2 emissions associated with a specific 
proposed Federal action triggered ESA Section 7(a)(2) consultation. The 
agencies have reviewed the long history of actions and guidance 
provided by DOI. To that point, the agencies incorporate by reference 
Appendix G of the MY 2012-2016 CAFE standards EIS.\3499\ That analysis 
relied on the significant legal and technical analysis undertaken by 
FWS and DOI. Specifically, NHTSA looked at the history of the Polar 
Bear Special Rule and several guidance memoranda provided by FWS and 
the U.S. Geological Survey. Ultimately, DOI concluded that a causal 
link could not be made between CO2 emissions associated with 
a proposed Federal action and specific effects on listed species; 
therefore, no Section 7(a)(2) consultation would be required.
---------------------------------------------------------------------------

    \3499\ Available on NHTSA's Corporate Average Fuel Economy 
website at https://one.nhtsa.gov/Laws-&-Regulations/CAFE-%E2%80%93-Fuel-Economy/Final-EIS-for-CAFE-Passenger-Cars-and-Light-Trucks,-Model-Years-2012%E2%80%932016.
---------------------------------------------------------------------------

    Subsequent to the publication of that Appendix, a court vacated the 
Polar Bear Special Rule on NEPA grounds, though it upheld the ESA 
analysis as having a rational basis.\3500\ FWS then issued a revised 
Final Special Rule for the Polar Bear.\3501\ In that final rule, FWS 
provided that for ESA Section 7, the determination of whether 
consultation is triggered is narrow and focused on the discrete effect 
of the proposed agency action. FWS wrote, ``[T]he consultation 
requirement is triggered only if there is a causal connection between 
the proposed action and a discernible effect to the species or critical 
habitat that is reasonably certain to occur. One must be able to 
`connect the dots' between an effect of a proposed action and an impact 
to the species and there must be a reasonable certainty that the effect 
will occur.'' \3502\ The statement in the revised Final Special Rule is 
consistent with the prior guidance published by FWS and remains valid 
today.\3503\ Likewise, the current regulations identify remoteness in 
time, geography, and the causal chain as factors to be considered in 
assessing whether a consequence is ``reasonably certain to occur.'' If 
the consequence is not reasonably certain to occur, it is not an 
``effect of a proposed action'' and does not trigger the consultation 
requirement.
---------------------------------------------------------------------------

    \3500\ In re: Polar Bear Endangered Species Act Listing and 
Section 4(D) Rule Litigation, 818 F.Supp.2d 214 (D.D.C. Oct. 17, 
2011).
    \3501\ 78 FR 11766 (Feb. 20, 2013).
    \3502\ 78 FR at 11784-11785.
    \3503\ See DOI Solicitor's Opinion No. M-37017, ``Guidance on 
the Applicability of the Endangered Species Act Consultation 
Requirements to Proposed Actions Involving the Emissions of 
Greenhouse Gases'' (Oct. 3, 2008).
---------------------------------------------------------------------------

    The agencies' actions establishing CAFE and CO2 
standards for passenger cars and light trucks do not directly affect 
listed species or critical habitat. The regulations promulgated by the 
agencies are used to calculate average standards for manufacturers 
based on the vehicles they produce for sale in the United States. Any 
potential effects of this action on listed species or designated 
critical habitat would be a result of changes to CO2 or air 
pollutant emissions that are caused by the individual choices of 
manufacturers in producing these vehicles and of consumers in 
purchasing and operating those vehicles. The agencies are not 
requiring, authorizing, funding, or carrying out the operation of motor 
vehicles (i.e., the proximate cause of downstream emissions), the 
production or refining of fuel (i.e., a proximate cause of upstream 
emissions),\3504\ the use of any land that is critical habitat for any 
purpose, or the taking of any listed species or other activity that may 
affect any listed species. Ultimately, the relevant decisions that 
result in emissions are taken by third parties, and any on-the-ground 
activities to implement and carry out those decisions are undertaken by 
such third parties. These decisions are influenced by a complex series 
of market factors that, though influenced by the agencies' actions, 
independently could result in the same series of decisions by consumers 
that commenters attribute to the agencies' actions (such as increased 
VMT and therefore increased emissions). This complex and lengthy chain 
of causality, which is highly dependent on market factors and therefore 
uncertain, leads the agencies to conclude that the resulting impacts of 
their actions to listed species or critical habitat do not satisfy the 
``but for'' test or are ``reasonably certain to occur.''
---------------------------------------------------------------------------

    \3504\ The agencies note that upstream emissions sources, such 
as oil extraction sites and fuel refineries, remain subject to the 
ESA. As future non-federal activities become reasonably certain, 
Section 7 and/or other sections of the ESA may provide protection 
for listed species and designated critical habitats. For example, 
new oil exploration or extraction activity may result in permitting 
or construction activities that would trigger consultation or other 
activities for the protection of listed species or designated 
critical habitat, as impacts may be more direct and more certain to 
occur.
---------------------------------------------------------------------------

    With regard to climate change, EPA and NHTSA are not able to make a 
causal link for purposes of Section 7(a)(2) that would ``connect the 
dots'' between their actions, vehicle emissions from motor vehicles 
affected by their actions, climate change, and particular impacts to 
listed species or critical habitats. The agencies' actions are to set 
standards that are effectively footprint curves, which are used as part 
of a complex calculation based on the vehicles produced by 
manufacturers for sale in the United States to determine a corporate 
average standard for each manufacturer. This approach, dictated by the 
Federal statute, gives manufacturers significant discretion to design, 
produce, and sell motor vehicles to meet consumer demand. Because 
manufacturers could choose to produce more vehicles with larger 
footprints (and therefore less stringent standards), fleet-average 
CO2 emissions could increase to some extent year-over-year 
independently of where the agencies set standards. Or the opposite may 
be true, and a shift in consumer preferences could lead to increased 
production of vehicles with smaller footprints (and therefore more 
stringent standards), resulting in overall declines in CO2 
emissions in the future compared to what the agencies are forecasting. 
Importantly, consumers not only choose which vehicles to purchase 
across a range of available fuel economies, they also choose how much 
to operate those vehicles (and therefore the quantity of fuel used and 
CO2 emitted)

[[Page 25254]]

independently of any action undertaken by the 
agencies.3505 3506
---------------------------------------------------------------------------

    \3505\ While VMT is affected by the cost of driving associated 
with fuel economy (i.e., the rebound effect), it is also affected by 
several market factors, such as economic conditions, that are far 
beyond the agencies' control and arguably have a greater influence 
than this rulemaking.
    \3506\ The fact that overall CO2 emissions are 
influenced so heavily by consumer preferences and behavior further 
supports the agencies' conclusion that impacts are not ``reasonably 
certain to occur.''
---------------------------------------------------------------------------

    Even with so many third parties in the causal chain making 
independent choices influenced by independent factors, the mechanics of 
climate change further break the chain of causality between the 
agencies' actions and specific effects on listed species or designated 
critical habitat. Climate change is a global phenomenon, impacted by 
greenhouse gas emissions that could occur anywhere throughout the 
world. As these gases accumulate in the atmosphere, radiative forcing 
increases, resulting in various potential impacts to the global climate 
system (e.g., warming temperatures, droughts, and changes in ocean pH) 
over long time scales. These changes could directly or indirectly 
impact listed species and/or designated critical habitat over time. 
Although this is a simplified explanation of a complex phenomenon 
subject to a significant degree of scientific study, it illustrates 
that the potential climate change-related consequences of this 
rulemaking on listed species and designated critical habitat are not 
``reasonably certain to occur'' under any of the three tests in the ESA 
regulations and listed above. Not only are the consequences to listed 
species or designated critical habitat geographically and temporally 
remote from the emissions that result from regulated vehicles, the 
chain of causality is simply too lengthy and complex. Because impacts 
to listed species and designated critical habitat result from climate 
shifts that, in and of themselves, result from the accumulation over 
time of greenhouse gas emissions from anywhere in the world, there is 
simply no way to ``connect the dots'' between the emissions from a 
regulated vehicle and those impacts. While the potential impacts of 
climate change have been well-documented, there is no degree of 
certainty that this action (as distinct from any other source of 
CO2 emissions) would be the cause of any particular impact 
to listed species or critical habitats. Because greenhouse gas 
emissions continue to occur from other sectors within the U.S. and from 
other sources globally, there is simply no scientific way to apportion 
any impact to a listed species or designated critical habitat to the 
agencies' actions.\3507\
---------------------------------------------------------------------------

    \3507\ See 50 CFR 402.17(b) (``A conclusion of reasonably 
certain to occur must be based on clear and substantial information, 
using the best scientific and commercial data available.'')
---------------------------------------------------------------------------

    One comment to the NPRM documented the potential impacts of climate 
change on Federally protected species and included a five-page table of 
species listed during 2006 to 2015 for which the commenters claim 
climate change was a listing factor.\3508\ This conflates the 
requirements of ESA Section 4 (governing ESA listing) and ESA Section 7 
(addressing the obligations of Federal agencies). Section 4 requires 
FWS or NMFS to assess all threats to species regardless of the origin 
of those threats. 16 U.S.C. 1533(a)(1). In contrast, the focus of 
Section 7(a)(2) is narrower and requires agencies to assess only 
effects on species that are attributable to the specific agency action. 
16 U.S.C. 1536(a)(2). That climate change was considered as a factor in 
a determination to list a species does not speak to the separate 
inquiry of whether the specific agency action is impacting a listed 
species. Here, the agencies believe this comment inappropriately 
attributes the entire issue of climate change, including all 
CO2 emissions no matter which sector generated them, to 
NHTSA and EPA's actions. In fact, NHTSA and EPA's actions would have 
only very small impacts on climate attributes, such as average 
temperatures, precipitation, and sea-level rise. The likelihood that 
these very small impacts, which are described above and in NHTSA's 
FEIS, would jeopardize listed species or adversely modify designated 
critical habitat is simply too remote to be cognizable under the ESA 
consultation requirements.\3509\ The fact that the agencies would 
exacerbate the impacts of climate change to a very small degree is not 
enough to determine that impacts on listed species or designated 
critical habitat are reasonably certain to occur.3510 3511
---------------------------------------------------------------------------

    \3508\ Center for Biological Diversity, Sierra Club, and Public 
Citizen, Inc., Docket No. NHTSA-2018-0067-12378, at 25-30.
    \3509\ Ground Zero Center for Non-Violent Action v. U.S. Dept. 
of Navy, 383 F.3d 1082 (2004).
    \3510\ Such a broad interpretation of the ESA would ensnare 
every Federal action that resulted in even an additional ounce of 
additional carbon dioxide emissions into the Section 7(a)(2) 
consultation process. See, e.g., 78 FR 11766, 11785 (Feb. 20, 2013) 
(``Without the requirement of a causal connection between the action 
under consultation and effects to species, literally every agency 
action that contributes CO2 emissions to the atmosphere 
would arguably result in consultation with respect to every listed 
species that may be affected by climate change.'').
    \3511\ The agencies also disagree that, for purposes of 
compliance with the ESA, this action would exacerbate climate change 
impacts on listed species or critical habitat. This final rule 
establishes CAFE and CO2 standards that increase in 
stringency on a year-by-year basis. While these standards are less 
stringent than the standards considered and set forth in the 2012 
rulemaking, the ESA does not serve as a one-way ratchet when 
agencies use their inherent authority to reconsider decisions that 
have not yet taken effect.
---------------------------------------------------------------------------

    As noted above, for consultation to be required, there must exist a 
sufficient nexus between the agency activity and the impact on listed 
species that the ESA intends to avoid. The Services have defined that 
nexus as ``but for'' causation. However, there is no ``but for'' 
causation associated with this final rule as the impacts of climate 
change will occur regardless of this action. In fact, even if the 
agencies were to set CAFE and CO2 standards at levels that 
would eliminate all CO2 emissions from motor vehicles made 
available for sale in the United States, the impacts of climate change 
are still projected to occur due to emissions from other sectors in the 
United States and other sources globally. Changes to tailpipe 
greenhouse gas emissions or associated upstream emissions related to 
this rulemaking and the alternatives considered would be very small 
compared to global CO2 emissions, which would continue. The 
agencies also note that because third parties (as described above) 
undertake most of the decisions that result in emissions, increased 
greenhouse gas emissions could occur regardless of the agencies' 
actions in this final rule. This further demonstrates the lack of ``but 
for'' causality in this case.
    Criteria air pollutant emissions from passenger cars and light 
trucks differ from greenhouse gas emissions in many ways. Most 
significantly, because passenger cars and light trucks are subject to 
gram-per-mile emissions standards for criteria pollutants, more fuel-
efficient (and, correspondingly, less CO2-intensive) 
vehicles are not necessarily, from the standpoint of air quality, 
``cleaner'' vehicles. Therefore, to the extent that CAFE and 
CO2 standards lead to changes in overall quantities of 
vehicular emissions that impact air quality, these are dominated by 
induced changes in highway travel. Changes in overall fuel consumption 
do lead to changes in emissions from ``upstream'' processes involved in 
supplying fuel to vehicles. Depending on how total vehicular emissions 
and total upstream emissions change in response to less stringent 
standards, overall emissions could increase or decrease.
    While small in magnitude, net impacts could also vary considerably

[[Page 25255]]

among different geographic areas depending on the locations of upstream 
emission sources and where changes in highway travel occur. This is 
important because of another significant difference between criteria 
air pollutant emissions and greenhouse gas emissions: Criteria air 
pollutant emissions are localized \3512\ whereas CO2 
emissions contribute to global atmospheric concentrations and climate 
change no matter where they occur. As reported in Section 4.1.1 of the 
FEIS, concentrations of many air pollutants emitted from motor vehicles 
are elevated in ambient air within approximately 1,000 to 2,000 feet of 
major roadways. With meteorological conditions that tend to inhibit the 
dispersion of emissions, concentrations of traffic-generated air 
pollutants can be elevated for as much as about 8,500 feet downwind of 
roads.3513 3514 But this means that impacts of criteria 
pollutant emissions are dependent on where they occur, to a degree much 
more significant than greenhouse gas emissions. Although the agencies 
anticipate increased fuel use as a result of this final rule (compared 
to the standards described in the 2012 final rule),\3515\ NHTSA and EPA 
have no way to know with reasonable certainty where additional fuel 
extraction and refining will occur. The agencies also cannot calculate 
with reasonable certainty where changes in highway travel will occur, 
as those impacts may not be uniform across the country. In fact, 
changes in land use patterns could exacerbate or reduce criteria 
pollutant emissions in any particular area, and such local changes are 
more uncertain. Therefore, even with the best scientific and commercial 
data available, the agencies cannot draw conclusions on impacts on 
particular listed species or designated critical habitat.
---------------------------------------------------------------------------

    \3512\ Criteria pollutant emissions contribute to local, 
regional, cross-state, and cross-national air pollution. Ultimately, 
however, the physical distance impacted by the pollutants is much 
smaller than for CO2 emissions, which affect the global 
atmosphere.
    \3513\ Hu, S., S. Fruin, K. Kozawa, S. Mara, S.E. Paulson, and 
A.M. Winer. A Wide Area of Air Pollutant Impact Downwind of a 
Freeway during Pre-sunrise Hours. Atmospheric Environment. 
43(16):2541-49 (2009). doi:10.1016/j.atmosenv.2009.02.033.
    \3514\ Hu, S., S.E. Paulson, S. Fruin, K. Kozawa, S. Mara, and 
A.M. Winer. Observation of Elevated Air Pollutant Concentrations in 
a Residential Neighborhood of Los Angeles California Using a Mobile 
Platform. Atmospheric Environment. 51:311-319 (2012). doi:10.1016/
j.atmosenv.2011.12.055. Available at: http://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC3755476&blobtype=pdf.
    \3515\ Although, again, the agencies note that average fleet-
wide fuel economy is projected to improve under any of the 
alternatives considered in this action.
---------------------------------------------------------------------------

    In short, the impacts of CAFE and CO2 standards on 
criteria pollutant emissions is indirect, and the impacts on air 
quality at any particular location (such as where a listed species or 
designated critical habitat is located) are more ambiguous than for 
global atmospheric concentrations of CO2 over the long term. 
Therefore, the agencies reach the same conclusion for criteria 
pollutant emissions as for CO2 emissions and climate change. 
For example, the causal chain between the agencies' actions and any 
impacts to listed species or designated critical habitat is attenuated 
by the fact that independent third parties must choose not only how 
much to operate their motor vehicles, but where to operate those motor 
vehicles as well. And the agencies cannot meaningfully conclude that 
any impact to a listed species and designated critical habitat would be 
caused by criteria pollutant emissions from the vehicles regulated by 
this rule rather than by another source. Finally, the impacts on 
criteria pollutant emissions as a result of this rule, especially in 
light of other emissions sources besides the regulated vehicles, are 
small\3516\ and the likelihood of jeopardy or the adverse modification 
of designated critical habitat is too remote. Current modeling tools 
available are not designed to trace fluctuations in ambient 
concentration levels of criteria and toxic air pollutants to potential 
impacts on particular endangered species. The agencies therefore cannot 
conclude that impacts are ``reasonably certain to occur.'' \3517\
---------------------------------------------------------------------------

    \3516\ For more information, see Chapter 4 of the FEIS.
    \3517\ See 50 CFR 402.17 (``A conclusion of reasonably certain 
to occur must be based on clear and substantial information, using 
the best scientific and commercial data available'').
---------------------------------------------------------------------------

    Finally, the agencies also note the potential uncertainty related 
to changes in total air pollutant and CO2 emissions as a 
result of the flexibilities in the CAFE and CO2 programs. 
Both programs allow manufacturers to trade and apply credits that have 
been earned from over-compliance in lieu of meeting the applicable 
standards for a particular model year, and manufacturers may have 
planned to rely on credits to comply with the standards for the model 
years regulated by this action. This could offset any changes in 
emissions that would result from the agencies' final decision. 
Furthermore, NHTSA's CAFE program allows manufacturers to pay civil 
penalties to cover any shortfall in compliance, further offsetting 
potential improvements in fuel economy (and, therefore, changes in air 
pollutant and CO2 emissions) that might have occurred under 
the augural standards. The existence of these flexibilities further 
supports the agencies' conclusion that they can establish neither ``but 
for'' causation nor a reasonable certainty that impacts will occur on 
listed species or designated critical habitat.
    The agencies have considered this analysis and conclude that any 
consequence to specific listed species or designated critical habitats 
from climate change or other air pollutant emissions is too remote and 
uncertain to be attributable to the agencies' actions here. These 
consequences are not ``effects'' for purposes of consultation under 
Section 7(a)(2). NHTSA and EPA therefore conclude that this final rule 
has no effect on listed species or their critical habitats.
(b) The Agencies Lack Sufficient Discretion or Control To Bring These 
Actions Under the Consultation Requirement of the ESA
    The primary purpose of EPCA, as amended by EISA, and codified at 49 
U.S.C. chapter 329, is energy conservation, and NHTSA is statutorily 
obligated to set attribute-based CAFE standards for each model year at 
the levels it determines are ``maximum feasible.'' \3518\ But ``maximum 
feasible'' is a balancing of several factors, and Congress clearly did 
not envision that the CAFE program would ``solve'' energy conservation 
in a single rulemaking action.\3519\ Fuel economy standards have the 
related benefit of reducing CO2 emissions, and may also 
result in reduced emissions of many criteria air pollutants. Similarly, 
EPA has found that the elevated concentrations of greenhouse gases in 
the atmosphere may reasonably be anticipated to endanger public health 
and welfare. As a result of these findings, CAA section 202(a) requires 
the agency to issue standards applicable to emissions of such gases 
from motor vehicles. Although not a statutory requirement, EPA has 
given weight to the policy goal of establishing CO2

[[Page 25256]]

standards that are coordinated with NHTSA's CAFE standards.\3520\
---------------------------------------------------------------------------

    \3518\ See 49 U.S.C. 32902(a) (``At least 18 months before the 
beginning of each model year, the Secretary of Transportation shall 
prescribe by regulation average fuel economy standards for 
automobiles manufactured by a manufacturer in that model year. Each 
standard shall be the maximum feasible average fuel economy level 
that the Secretary decides the manufacturers can achieve in that 
model year.'').
    \3519\ See, e.g., 49 U.S.C. 32902(b)(2) (setting separate 
requirements for CAFE standards for MYs 2011 through 2020 and MYs 
2021 through 2030).
    \3520\ See Mass. v. EPA, 549 U.S. 497, 532 (2007) (``. . .there 
is no reason to think the two agencies cannot both administer their 
obligations and yet avoid inconsistency.'')
---------------------------------------------------------------------------

    As previously indicated, commenters assert that CO2 and 
criteria air pollutant emissions are relevant to Section 7(a)(2) 
consultation because of the potential impacts of climate change or the 
pollutants themselves on listed species or designated critical habitat. 
However, it is not clear whether their comments are based on the fact 
that the agencies predict increases in CO2 emissions and 
most criteria pollutant emissions under all action alternatives 
compared to the MY 2022-2025 CO2 and augural CAFE standards, 
or the fact that any emissions from passenger cars or light trucks will 
continue under any of the alternatives considered.
    With regard to the latter, NHTSA does not interpret EPCA/EISA to 
mean that Congress expected the CAFE program to take the U.S. auto 
fleet off of oil entirely--indeed, EISA renders doing so impossible 
because it amended EPCA to prohibit NHTSA from considering the fuel 
economy of dedicated alternative fuel vehicles, including electric 
vehicles, when setting maximum feasible standards. This means that 
standards cannot be set that assume increased usage of full 
electrification for compliance. As a result, no matter the level at 
which NHTSA sets CAFE standards in accordance with EPCA, CO2 
and criteria pollutant emissions will continue. So long as NHTSA's 
obligation to set CAFE standards remains in place, it is reasonable to 
assume that Congress's expectation for EPA, in coordinating with NHTSA, 
is similar.
    The purpose of Section 7(a)(2) consultation is to ensure that 
Federal agencies are not undertaking, funding, permitting, or 
authorizing actions that are likely to jeopardize the continued 
existence of listed species or destroy or adversely modify designated 
critical habitat. However, no matter what standards the agencies set 
under the CAFE and CO2 programs, Americans will continue to 
drive. Neither NHTSA nor EPA has authority to control vehicle miles 
traveled. As long as there is driving, there will be emissions--whether 
from vehicle tailpipes or from the stationary sources that create the 
energy that the vehicles consume. Moreover, both agencies have 
concluded that significant further electrification of the fleet is not 
practicable at this time due to concerns about consumer acceptance in a 
time of foreseeably low fuel prices. The fact that CO2 and 
criteria pollutant emissions will continue after NHTSA and EPA actions 
on standards cannot, alone, trigger Section 7(a)(2) consultation as the 
agencies lack the discretion or control over these emissions to simply 
regulate them away entirely in this action.\3521\ Consultation is not 
required where an agency lacks discretion to take action that will 
inure to the benefit of listed species.\3522\ Since elimination of oil 
from the fleet is inconsistent with the agencies' statutory authorities 
and the clear intent of Congress, consultation is not triggered under 
this scenario.
---------------------------------------------------------------------------

    \3521\ National Ass'n of Home Builders v. Defenders of Wildlife, 
551 U.S. 644, 673 (2007) (``Applying Chevron, we defer to the 
Agency's reasonable interpretation of ESA [section] 7(a)(2) as 
applying only to `actions in which there is discretionary Federal 
involvement or control.''' (quoting 50 CFR 402.03)).
    \3522\ Id.; Sierra Club v. Babbitt, 65 F.3d 1502, 1509 (9th Cir. 
1995) (ESA Section 7(a)(2) consultation is not required where an 
agency lacks discretion to influence private conduct in a manner 
that will inure to the benefit of listed species).
---------------------------------------------------------------------------

    Commenters may instead be referring to the trend in CO2 
and criteria air pollutant emissions under the action alternatives 
considered in this rulemaking (e.g., whether and by how much emissions 
increase or decrease). To that point, all of the action alternatives 
considered result in increases in CO2 and most criteria air 
pollutant emissions compared to the standards considered and set forth 
in the 2012 rulemaking. However, the agencies do not believe this is 
the relevant comparison for purposes of determining the applicability 
of Section 7 of the ESA to this action. Model years 2021 through 2026, 
for the most part, have not yet arrived. So it is not appropriate to 
compare the current action to a prior action that has not been 
implemented and which the agencies are reconsidering. When compared to 
standards through MY 2020, under any of the alternatives considered, 
fuel economy will improve and CO2 and most criteria 
pollutant emissions will decrease over time, either as stringency 
increases or from the turnover in the fleet to newer, cleaner vehicles.
    As detailed above, however, there is no way to meaningfully 
differentiate between the alternatives in terms of outcomes for listed 
species and designated critical habitat. The agencies cannot reasonably 
calculate how incrementally less emissions resulting from more 
stringent standards would benefit those species or habitats; rather, at 
most, the agencies can only posit that more stringent standards 
hypothetically could lead to better outcomes. But where to draw any 
line in terms of impacts to species and habitats is an impossible 
exercise. Yet, as noted above, NHTSA is mandated by Congress to set 
``maximum feasible'' standards and EPA's mission is to protect public 
health and welfare. Under these circumstances, where the agencies must 
issue standards pursuant to statutory mandate that under any scenario 
will involve emissions, yet they lack the commensurate ability to take 
action that will inure to the benefit of species in any meaningful way, 
Section 7(a)(2) consultation is not required.
    Finally, regardless of the level of stringency at which the 
agencies set CAFE and CO2 standards, criteria pollutant and 
CO2 emissions from motor vehicles will change to a greater 
or lesser degree because of several independent factors. Because of the 
complex relationships between fuel economy, vehicle sales, driver 
behavior (e.g., VMT and driving location), and technology choices by 
manufacturers, emissions will never uniformly increase or decrease for 
all future model years, across all regulated pollutants, and in all 
locations throughout the country. For example, increased stringency may 
result in greater VMT, resulting in larger downstream emissions of some 
criteria pollutants. On the other hand, decreased stringency may result 
in greater fuel refining, result in larger upstream emissions of some 
pollutants. Because vehicle operation and refinery activity depends 
upon independent market forces, impacts to particular listed species or 
designated critical habitat are dependent upon where vehicle operation 
or increased fuel refining occur, but neither agency can control such 
decisions. Regardless of whether NHTSA and EPA engage in Section 
7(a)(2) consultation, the agencies lack the control necessary to negate 
all emissions increases in whatever years and locations they occur 
(e.g., ensure ideal technology choices by manufacturers, control 
consumer purchasing behavior, or regulate driving locations or VMT), or 
otherwise mitigate impacts associated with these particular emissions. 
But setting stringency is, in fact, what the agencies are statutorily 
obligated to do.
    For the foregoing reasons, NHTSA and EPA conclude that they lack 
sufficient discretion or control to bring these actions under the 
consultation requirement of the ESA.
7. Floodplain Management (Executive Order 11988 and DOT Order 5650.2)
    These Orders require Federal agencies to avoid the long- and short-
term adverse impacts associated with the

[[Page 25257]]

occupancy and modification of floodplains, and to restore and preserve 
the natural and beneficial values served by floodplains. Executive 
Order 11988 also directs agencies to minimize the impact of floods on 
human safety, health, and welfare, and to restore and preserve the 
natural and beneficial values served by floodplains through evaluating 
the potential effects of any actions the agency may take in a 
floodplain and ensuring that its program planning and budget requests 
reflect consideration of flood hazards and floodplain management. DOT 
Order 5650.2 sets forth DOT policies and procedures for implementing 
Executive Order 11988. The DOT Order requires that the agency determine 
if a proposed action is within the limits of a base floodplain, meaning 
it is encroaching on the floodplain, and whether this encroachment is 
significant. If significant, the agency is required to conduct further 
analysis of the proposed action and any practicable alternatives. If a 
practicable alternative avoids floodplain encroachment, then the agency 
is required to implement it.
    In this rulemaking, the agencies are not occupying, modifying and/
or encroaching on floodplains. The agencies, therefore, conclude that 
the Orders are not applicable to this action. NHTSA has, however, 
conducted a review of the alternatives on potentially affected 
resources, including floodplains, in its FEIS.
8. Preservation of the Nation's Wetlands (Executive Order 11990 and DOT 
Order 5660.1a)
    These Orders require Federal agencies to avoid, to the extent 
possible, undertaking or providing assistance for new construction 
located in wetlands unless the agency head finds that there is no 
practicable alternative to such construction and that the proposed 
action includes all practicable measures to minimize harm to wetlands 
that may result from such use. Executive Order 11990 also directs 
agencies to take action to minimize the destruction, loss, or 
degradation of wetlands in ``conducting Federal activities and programs 
affecting land use, including but not limited to water and related land 
resources planning, regulating, and licensing activities.'' DOT Order 
5660.1a sets forth DOT policy for interpreting Executive Order 11990 
and requires that transportation projects ``located in or having an 
impact on wetlands'' should be conducted to assure protection of the 
Nation's wetlands. If a project does have a significant impact on 
wetlands, an EIS must be prepared.
    In the NPRM, the agencies noted that they are not undertaking or 
providing assistance for new construction located in wetlands. The 
agencies, therefore, concluded that these Orders do not apply to this 
rulemaking. One commenter disagreed with this conclusion, noting the 
potential land use impacts of the rule and the agencies' obligation to 
consider all factors relevant to the proposal's effect on the survival 
and quality of wetlands.\3523\ The agencies do not believe that it is 
feasible to establish the requisite causal chain between the impacts of 
this action and impacts on wetlands, nor would such impacts be 
reasonably foreseeable as a direct or indirect result of this 
rulemaking. The agencies therefore continue to conclude that these 
Orders do not apply to this rulemaking. Regardless, NHTSA addresses the 
potential effects of the alternatives on resources, including wetlands, 
in its FEIS.
---------------------------------------------------------------------------

    \3523\ Joint Submission from the States of California et al. and 
the Cities of Oakland et al., Docket No. NHTSA-2018-0067-11735, at 
46-47.
---------------------------------------------------------------------------

9. Migratory Bird Treaty Act (MBTA), Bald and Golden Eagle Protection 
Act (BGEPA), Executive Order 13186
    The MBTA (16 U.S.C. 703-712) provides for the protection of certain 
migratory birds by making it illegal for anyone to ``pursue, hunt, 
take, capture, kill, attempt to take, capture, or kill, possess, offer 
for sale, sell, offer to barter, barter, offer to purchase, purchase, 
deliver for shipment, ship, export, import, cause to be shipped, 
exported, or imported, deliver for transportation, transport or cause 
to be transported, carry or cause to be carried, or receive for 
shipment, transportation, carriage, or export'' any migratory bird 
covered under the statute.\3524\
---------------------------------------------------------------------------

    \3524\ 16 U.S.C. 703(a).
---------------------------------------------------------------------------

    The BGEPA (16 U.S.C. 668-668d) makes it illegal to ``take, possess, 
sell, purchase, barter, offer to sell, purchase or barter, transport, 
export or import'' any bald or golden eagles.\3525\ Executive Order 
13186, ``Responsibilities of Federal Agencies to Protect Migratory 
Birds,'' helps to further the purposes of the MBTA by requiring a 
Federal agency to develop a Memorandum of Understanding (MOU) with the 
Fish and Wildlife Service when it is taking an action that has (or is 
likely to have) a measurable negative impact on migratory bird 
populations.
---------------------------------------------------------------------------

    \3525\ 16 U.S.C. 668(a).
---------------------------------------------------------------------------

    The agencies conclude that the MBTA, BGEPA, and Executive Order 
13186 do not apply to this action because there is no disturbance, 
take, measurable negative impact, or other covered activity involving 
migratory birds or bald or golden eagles involved in this rulemaking.
10. Department of Transportation Act (Section 4(f))
    Section 4(f) of the Department of Transportation Act of 1966 (49 
U.S.C. 303), as amended, is designed to preserve publicly owned park 
and recreation lands, waterfowl and wildlife refuges, and historic 
sites. Specifically, Section 4(f) provides that DOT agencies cannot 
approve a transportation program or project that requires the use of 
any publicly owned land from a public park, recreation area, or 
wildlife or waterfowl refuge of national, State, or local significance, 
or any land from a historic site of national, State, or local 
significance, unless a determination is made that:
    (1) There is no feasible and prudent alternative to the use of 
land, and
    (2) The program or project includes all possible planning to 
minimize harm to the property resulting from the use.
    These requirements may be satisfied if the transportation use of a 
Section 4(f) property results in a de minimis impact on the area.
    NHTSA concludes that Section 4(f) is not applicable to this action 
because this rulemaking is not an approval of a transportation program 
or project that requires the use of any publicly owned land.
11. Executive Order 12898: ``Federal Actions To Address Environmental 
Justice in Minority Populations and Low-Income Populations''
    Executive Order 12898 (59 FR 7629 (Feb. 16, 1994)) establishes 
Federal executive policy on environmental justice. It directs Federal 
agencies, to the greatest extent practicable and permitted by law, to 
make environmental justice part of their mission by identifying and 
addressing, as appropriate, disproportionately high and adverse human 
health or environmental effects of their programs, policies, and 
activities on minority and low-income populations in the United States. 
DOT Order 5610.2(a) \3526\ sets forth the Department of 
Transportation's policy to consider environmental justice principles in 
all its programs, policies, and activities.
---------------------------------------------------------------------------

    \3526\ Department of Transportation Updated Environmental 
Justice Order 5610.2(a), 77 FR 27534 (May 10, 2012).
---------------------------------------------------------------------------

    Environmental justice is a principle asserting that all people 
deserve fair treatment and meaningful involvement with respect to 
environmental laws,

[[Page 25258]]

regulations, and policies. EPA seeks to provide the same degree of 
protection from environmental health hazards for all people. DOT shares 
this goal and is informed about the potential environmental impacts of 
its rulemakings through the NEPA process. One comment on the NPRM 
claimed that the agencies ``failed to recognize the benefits of the 
existing standards'' for disadvantaged communities. Specifically, the 
commenter claimed that the agencies did not provide an underlying 
analysis of environmental justice issues and thereby failed to meet the 
requirements of E.O. 12898.\3527\ However, the agencies addressed their 
obligations under E.O. 12898 in the preamble to the NPRM and in Section 
7.5 of the DEIS. The agencies received a number of comments regarding 
the analysis it presented. NHTSA responds to those comments in Section 
10.7 of the FEIS, and the agencies have revised their environmental 
justice analysis based on the information contained in those comments. 
The revised analysis is presented here and in the FEIS.
---------------------------------------------------------------------------

    \3527\ CARB, Docket No. NHTSA-2018-0067-11873, at 411-12.
---------------------------------------------------------------------------

    There is evidence that proximity to oil refineries could be 
correlated with incidences of cancer and 
leukemia.3528 3529 3530 Proximity to high-traffic roadways 
could result in adverse cardiovascular and respiratory impacts, among 
other possible impacts.3531 3532 3533 3534 3535 3536 3537 
Climate change affects overall global temperatures, which could, in 
turn, affect the number and severity of outbreaks of vector-borne 
illnesses.3538 3539 In the context of this rulemaking, the 
environmental justice concern is the extent to which minority and low-
income populations could be more exposed or vulnerable to such 
environmental and health impacts.
---------------------------------------------------------------------------

    \3528\ Pukkala, E. Cancer incidence among Finnish oil refinery 
workers, 1971-1994. Journal of Occupational and Environmental 
Medicine. 40(8):675-79 (1998). doi:10.1023/A:1018474919807.
    \3529\ Chan, C.-C.; Shie, R.H.; Chang, T.Y.; Tsai, D.H. Workers' 
exposures and potential health risks to air toxics in a 
petrochemical complex assessed by improved methodology. 
International Archives of Occupational and Environmental Health. 
79(2):135-142 (2006). doi:10.1007/s00420-005-0028-9. Online at: 
https://www.researchgate.net/publication/7605242_Workers'_exposures_and_potential_health_risks_to_air_toxics_i
n_a_petrochemical_complex_assessed_by_improved_methodology.
    \3530\ Bulka, C.; Nastoupil, L.J.; McClellan, W.; Ambinder, A.; 
Phillips, A.; Ward, K.; Bayakly, A.R.; Switchenko, J.M.; Waller, L.; 
Flowers, C.R. Residence proximity to benzene release sites is 
associated with increased incidence of non-Hodgkin lymphoma. Cancer. 
119(18):3309-17 (2013). doi:10.1002/cncr.28083. Online at: http://onlinelibrary.wiley.com/doi/10.1002/cncr.28083/pdf;jsessionid=1520A90A764A95985316057D7D76A362.f02t02.
    \3531\ HEI (Health Effects Institute). 2010. Traffic-Related Air 
Pollution: A Critical Review of the Literature on Emissions, 
Exposure and Health Effects. Special Report 17. Health Effects 
Institute: Boston, MA:. HEI Panel on the Health Effects of Traffic-
Related Air Pollution, 386 pp. Available at: https://www.healtheffects.org/system/files/SR17Traffic%20Review.pdf. 
(Accessed: March 3, 2018).
    \3532\ Heinrich, J. and H.-E. Wichmann. 2004. Traffic Related 
Pollutants in Europe and their Effect on Allergic Disease. Current 
Opinion in Allergy and Clinical Immunology 4(5):341-348.
    \3533\ Salam, M.T., T. Islam, and F.D. Gilliland. 2008. Recent 
Evidence for Adverse Effects of Residential Proximity to Traffic 
Sources on Asthma. Current Opinion in Pulmonary Medicine 14(1):3-8. 
doi:10.1097/MCP.0b013e3282f1987a.
    \3534\ Samet, J.M. 2007. Traffic, Air Pollution, and Health. 
Inhalation Toxicology 19(12):1021-27. doi:10.1080/08958370701533541.
    \3535\ Adar, S. and J. Kaufman. 2007. Cardiovascular Disease and 
Air Pollutants: Evaluating and Improving Epidemiological Data 
Implicating Traffic Exposure. Inhalation Toxicology 19(S1):135-49. 
doi:10.1080/08958370701496012.
    \3536\ Wilker, E.H., E. Mostofsky, S.H. Lue, D. Gold, J. 
Schwartz, G.A. Wellenius, and M.A. Mittleman. 2013. Residential 
Proximity to High-Traffic Roadways and Poststroke Mortality. Journal 
of Stroke and Cerebrovascular Diseases 22(8): e366-e372. 
doi:10.1016/j.jstrokecerebrovasdis.2013.03.034. Available at: 
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4066388/. (Accessed: 
March 6, 2018).
    \3537\ Hart, J.E., E.B. Rimm, K.M. Rexrode, and F. Laden. 2013. 
Changes in Traffic Exposure and the Risk of Incident Myocardial 
Infarction and All-cause Mortality. Epidemiology 24(5):734-42.
    \3538\ U.S. Global Change Research Program (GCRP). Global 
Climate Change Impacts in the United States: The Third National 
Climate Assessment. U.S. Global Change Research Program. Melillo, 
J.M, T.C. Richmond, and G.W. Yohe (Eds.). U.S. Government Printing 
Office: Washington, DC 841 pp (2014). doi:10.7930/J0Z31WJ2. 
Available at: http://nca2014.globalchange.gov/report. (Accessed: 
February 27, 2018).
    \3539\ GCRP. The Impacts of Climate Change on Human Health in 
the United States, A Scientific Assessment (2016). April 2016. 
Available at: https://health2016.globalchange.gov. (Accessed: 
February 28, 2018).
---------------------------------------------------------------------------

    Numerous studies have found that some environmental hazards are 
more prevalent in areas where racial/ethnic minorities and people with 
low socioeconomic status represent a higher proportion of the 
population compared with the general population. In addition, compared 
to non-Hispanic whites, some subpopulations defined by race and 
ethnicity have been shown to have a greater incidence of some health 
conditions during certain life stages. For example, in 2014, about 13 
percent of Black, non-Hispanic and 24 percent of Puerto Rican children 
were estimated to have asthma, compared with 8 percent of white, non-
Hispanic children.\3540\ The agencies have therefore considered areas 
nationwide that could contain minority and low-income communities who 
would most likely be exposed to the environmental and health impacts of 
oil production, distribution, and consumption or the potential impacts 
of climate change. These include areas where oil production and 
refining occur, areas near roadways, coastal flood-prone areas, and 
urban areas that are subject to the heat island effect.\3541\
---------------------------------------------------------------------------

    \3540\ http://www.cdc.gov/asthma/most_recent_data.htm.
    \3541\ The heat island effect refers to developed areas having 
higher temperatures than surrounding rural areas.
---------------------------------------------------------------------------

    The following discussion addresses environmental justice 
implications related to air quality and to climate change and carbon 
emissions in the context of this final rulemaking. Emissions of air 
pollutants may be affected by this rulemaking due to changes in fuel 
use and VMT, which are described above. To the degree to which minority 
and low-income populations may be present in proximity to the locations 
described in this section, they may be exposed disproportionately to 
these emissions changes. In addition, the following analysis also 
discusses other potential reasons why minority and low-income 
populations may be susceptible to the health impacts of air pollutants. 
NHTSA also discusses environmental justice in Chapter 7.5 of its FEIS.
a) Proximity to Oil Production and Refining
    As stated above, numerous studies have found that some 
environmental hazards are more prevaluent in areas where minority and 
low-income populations represent a higher proportion of the population 
compared with the general population. For example, one study found that 
survey respondents who were black and, to a lesser degree, had lower 
income levels, were significantly more likely to live within 1 mile of 
an industrial facility listed in the EPA's 1987 Toxic Release Inventory 
(TRI) national database.\3542\
---------------------------------------------------------------------------

    \3542\ Mohai, P., P.M. Lantz, J. Morenoff, J.S. House, and R.P. 
Mero. Racial and Socioeconomic Disparities in Residential Proximity 
to Polluting Industrial Facilities: Evidence from the Americans' 
Changing Lives Study. American Journal of Public Health 99(S3): 
S649-S656 (2009). doi:10.2105/AJPH.2007.131383. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2774179/pdf/S649.pdf. 
(Accessed: March 2, 2018).
---------------------------------------------------------------------------

    A meta-analysis of 49 environmental equity studies concluded that 
evidence of race-based environmental inequities is statistically 
significant (although the average magnitude of these inequities is 
small), while evidence supporting the existence of income-based 
environmental inequities is substantially weaker.\3543\ Considering 
poverty-based class effects, that meta-

[[Page 25259]]

analysis found an inverse relationship between environmental risk and 
poverty, concluding that environmental risks are less likely to be 
located in areas of extreme poverty.\3544\ However, individual studies 
may reach contradictory conclusions in relation to race- and income-
based inequities across a range of environmental risks. Therefore, the 
meta-analysis also sought to examine the reasons why conclusions vary 
across studies of environmental inequity. Possible explanations for why 
studies reach contrary conclusions include variability in the source of 
potential environmental risk that the study considers (e.g., the type 
of facility or the associated level of pollution or risk); variability 
in the methodology applied to aggregate demographic data and to define 
the comparison population; and the degree to which statistical models 
control for other variables that may explain the distribution of 
potential environmental risk.
---------------------------------------------------------------------------

    \3543\ Ringquist, E.J. Evidence of Environmental Inequities: A 
Meta-Analysis. Journal of Policy Analysis and Management 24(2):223-
47 (2005).
    \3544\ Ringuist (2005).
---------------------------------------------------------------------------

    To test whether there are disparate impacts from hazardous 
industrial facilities on racial/ethnic minorities, the disadvantaged, 
the working class, and manufacturing workers, one study tested the 
relationship between hazard scores of Philadelphia-area facilities in 
EPA's Risk-Screening Environmental Indicators (RSEI) database and the 
demographics of populations near those facilities using multivariate 
regression.\3545\ This study concluded that racial/ethnic minorities, 
the most socioeconomically disadvantaged, and those employed in 
manufacturing suffer a disparate impact from the highest-hazard 
facilities (primarily manufacturing plants).
---------------------------------------------------------------------------

    \3545\ Sicotte, D. and S. Swanson. Whose Risk in Philadelphia? 
Proximity to Unequally Hazardous Industrial Facilities. Social 
Science Quarterly 88(2):516-534 (2007).
---------------------------------------------------------------------------

    Other commissioned reports and case studies provide additional 
evidence of the presence of low-income and minority populations near 
industrial facilities and of racial or socioeconomic disparities in 
exposure to environmental risk, although these sources were not 
published in peer-reviewed scientific 
journals.3546 3547 3548 3549
---------------------------------------------------------------------------

    \3546\ UCC (United Church of Christ). Toxic Wastes and Race at 
Twenty: 1987--2007. A Report Prepared for the United Church of 
Christ Justice and Witness Ministries. Available at: https://www.nrdc.org/sites/default/files/toxic-wastes-and-race-at-twenty-1987-2007.pdf (2007). (Accessed: April 9, 2018).
    \3547\ National Association for the Advancement of Colored 
People and Clean Air Task Force. Fumes Across the Fence-line: The 
Health Impacts of Air Pollution from Oil & Gas Facilities on African 
American Communities (2017). Available at: http://www.catf.us/wp-content/uploads/2017/11/CATF_Pub_FumesAcrossTheFenceLine.pdf. 
(Accessed: February 24, 2019).
    \3548\ Ash, M., J.K. Boyce, G. Chang, M. Pastor, J. Scoggins, 
and J. Tran. Justice in the Air: Tracking Toxic Pollution from 
America's Industries and Companies to our States, Cities, and 
Neighborhoods. Political Economy Research Institute at the 
University of Massachusetts, Amherst and the Program for 
Environmental and Regional Equity at the University of Southern 
California (2009). Available at: https://dornsife.usc.edu/assets/sites/242/docs/justice_in_the_air_web.pdf. (Accessed: February 24, 
2019).
    \3549\ Kay, J. and C. Katz. Pollution, Poverty and People of 
Color: Living With Industry. Scientific American. Available at: 
https://www.scientificamerican.com/article/pollution-poverty-people-color-living-industry/ (2012). (Accessed: March 4, 2018).
---------------------------------------------------------------------------

    Few studies address disproportionate exposure to environmental risk 
associated with oil refineries specifically. One study found that the 
populations surrounding oil refineries are more often minorities, 
concluding that ``56 percent of people living within three miles of 
[oil] refineries in the United States are minorities--almost double the 
national average.'' \3550\ Another examined whether findings of 
environmental inequity varied between coke production plants and oil 
refineries, both of which are significant sources of air 
pollution.\3551\ This study concluded that census tracts near coke 
plants had a disproportionate share of poor and nonwhite residents, and 
that existing inequities were primarily economic in nature. However, 
the findings for oil refineries did not strongly support an 
environmental inequity hypothesis. A more recent study of environmental 
justice in the oil refinery industry found evidence of environmental 
injustice as a result of unemployment levels in areas around refineries 
and, to a slightly lesser extent, as a result of income 
inequality.\3552\ This study did not test for race-based environmental 
inequities.
---------------------------------------------------------------------------

    \3550\ O'Rourke, D. and S. Connolly. Just Oil? The Distribution 
of Environmental and Social Impacts of Oil Production and 
Consumption. Annual Review of Environment and Resources 28(1):587-
617 (2003). doi:10.1146/annurev.energy.28.050302.105617.
    \3551\ Graham, J.D., N.D. Beaulieu, D. Sussman, M. Sadowitz, and 
Y.C. Li. Who Lives Near Coke Plants and Oil Refineries? An 
Exploration of the Environmental Inequity Hypothesis. Risk Analysis 
19(2):171-86 (1999). doi:10.1023/A:1006965325489. Green, R.S., S. 
Smorodinsky, J.J. Kim, R. McLaughlin, and B. Ostro. Proximity of 
California public schools to busy roads. Environmental Health 
Perspectives 112 (1):61-66 (2004). Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241798/. (Accessed: May 31, 
2018).
    \3552\ Carpenter, A. and M. Wagner. Environmental Justice in the 
Oil Refinery Industry: A Panel Analysis Across United States 
Counties. Ecological Economics 159:101-109 (2019).
---------------------------------------------------------------------------

    Overall, the body of scientific literature points to 
disproportionate representation of minority and low-income populations 
in proximity to a range of industrial, manufacturing, and hazardous 
waste facilities that are stationary sources of air pollution; although 
results of individual studies may vary. While the scientific literature 
specific to oil refineries is limited, disproportionate exposure of 
minority and low-income populations to air pollution from oil 
refineries is suggested by other broader studies of racial and 
socioeconomic disparities in proximity to industrial facilities 
generally.
    The potential increase in fuel production and consumption projected 
as a result of this rulemaking (compared to the No Action Alternative) 
could lead to an increase in upstream emissions of criteria and toxic 
air pollutants due to increased extraction, refining, and 
transportation of fuel. As described in Section VII.A.4.c.3.b.i, total 
upstream emissions of criteria and toxic air pollutants in 2035 are 
projected to increase under all action alternatives compared to the No 
Action Alternative, with the exception that total upstream emissions of 
SO2 are projected to decrease under all action alternatives 
under the CAFE program (but not under the CO2 program). As 
noted, a correlation between proximity to oil refineries and the 
prevalence of minority and low-income populations is suggested in the 
scientific literature. To the extent that minority and low-income 
populations live closer to oil refining facilities, these populations 
may be more likely to be adversely affected by these emissions. 
However, the magnitude of the change in emissions relative to the 
baseline is minor and would not be characterized as high and adverse.
Proximity to High-Traffic Roadways
    Studies have more consistently demonstrated a disproportionate 
prevalence of minority and low-income populations living near mobile 
sources of pollutants. In certain locations in the United States, for 
example, there is consistent evidence that populations or schools near 
roadways typically include a greater percentage of minority or low-
income residents.3553 3554 3555 3556 3557 3558 3559 In

[[Page 25260]]

California, studies demonstrate that minorities and low-income 
populations are disproportionately likely to live near a major roadway 
or in areas of high traffic density compared to the general 
population.3560 3561 A study of traffic, air pollution, and 
socio-economic status inside and outside the Minneapolis-St. Paul 
metropolitan area similarly found that populations on the lower end of 
the socioeconomic spectrum and minorities are disproportionately 
exposed to traffic and air pollution and at higher risk for adverse 
health outcomes.\3562\ Near-road exposure to vehicle emissions can 
cause or exacerbate health conditions such as 
asthma.3563 3564 3565 3566 One study demonstrated that 
students at schools in Michigan closer to major highways had a higher 
risk of respiratory and neurological disease and were more likely to 
fail to meet state educational standards, after controlling for other 
variables.\3567\ In general, studies such as these demonstrate trends 
in specific locations in the United States that may be indicative of 
broader national trends.
---------------------------------------------------------------------------

    \3553\ Green, R.S., S. Smorodinsky, J.J. Kim, R. McLaughlin, and 
B. Ostro. Proximity of California public schools to busy roads. 
Environmental Health Perspectives 112 (1):61-66 (2004). Available 
at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241798/. Last 
accessed: May 31, 2018.
    \3554\ Wu, Y-C.; Batterman, S.A. Proximity of schools in 
Detroit, Michigan to automobile and truck traffic. Journal of 
Exposure Science and Environmental Epidemiology 16(5): 457-470 
(2006). doi:10.1038/sj.jes.7500484. Available at: http://www.nature.com/articles/7500484. Last accessed: May 31, 2018.
    \3555\ Chakraborty, J., and P.A. Zandbergen. Children at risk: 
measuring racial/ethnic disparities in potential exposure to air 
pollution at school and home. Journal of Epidemiology & Community 
Health 61:1074-1079 (2007). doi: 10.1136/jech.2006.054130.
    \3556\ Depro, B., and C. Timmins. Mobility and Environmental 
Equity: Do Housing Choices Determine Exposure to Air Pollution? 
North Carolina State University and RTI International, Duke 
University and NBER (2008). Available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.586.7164&rep=rep1&type=pdf. (Accessed: May 31, 
2018).
    \3557\ Marshall, J.D. Environmental inequality: air pollution 
exposures in California's South Coast Air Basin. Atmospheric 
Environment 42(21):5499-5503 (2008).
    \3558\ Su, J. G., T. Larson, T. Gould, M. Cohen, and M. 
Buzzelli. Transboundary air pollution and environmental justice: 
Vancouver and Seattle compared. GeoJournal 75(6):595-608 (2010). 
doi: 10.1007/s10708-009-9269-6.
    \3559\ Su, J. G., M. Jarrett, A. de Nazelle, and J. Wolch. Does 
exposure to air pollution in urban parks have socioeconomic, racial 
or ethnic gradients? Environmental Research 111 (3):319-328 (2011). 
doi: 10.1016/j.envres.2011.01.002.
    \3560\ Carlson, A.E. The Clean Air Act's Blind Spot: 
Microclimates and Hotspot Pollution. 65 UCLA Law Review 1036 (2018).
    \3561\ Gunier, R.B., A. Hertz, J. Von Behren, and P. Reynolds. 
Traffic density in California: socioeconomic and ethnic differences 
among potentially exposed children. Journal of Exposure Analysis and 
Environmental Epidemiology 13(3):240-46 (2003). doi:10.1038/
sj.jea.7500276.
    \3562\ Pratt, G.C., M.L. Vadali, D.L. Kvale, and K.M. Ellickson, 
Traffic, air pollution, minority, and socio-economic status: 
addressing inequities in exposure and risk. International Journal of 
Environmental research and Public Health 12(5):53555372 (2015). 
doi:10.3390/ijerph120505355.
    \3563\ Carlson (2018).
    \3564\ Gunier et al. (2003).
    \3565\ Meng, Y-Y., M. Wilhelm, R.P. Rull, P. English, S. Nathan, 
and B. Ritz. Are frequent asthma symptoms among low-income 
individuals related to heavy traffic near homes, vulnerabilities, or 
both? Annals of Epidemiology 18:343-350 (2008). doi:10.1016/
j.annepidem.2008.01.006.
    \3566\ Khreis, H., C. Kelly, J. Tate, R. Parslow, K. Lucas, and 
M. Nieuwenhuijsen. Exposure to traffic-related air pollution and 
risk of development of childhood asthma: A systematic review and 
meta-analysis. Environment International 100:1-31 (2017). https://doi.org/10.1016/j.envint.2016.11.012.
    \3567\ Kweon, B-S., P. Mohai, S. Lee, and A.M. Sametshaw. 2016. 
Proximity of Public Schools to Major Highways and Industrial 
Facilities, and Students' School Performance and Health Hazards. 
Environment and Planning B: Urban Analytics and City Science 
45(2):312-329. doi.org/10.1177/0265813516673060.
---------------------------------------------------------------------------

    Fewer studies have been conducted at the national level, yet those 
that do exist also demonstrate a correlation between minority and low-
income status and proximity to roadways.3568 3569 For 
example, one study found that greater traffic volumes and densities at 
the national level are associated with larger shares of minority and 
low-income populations living in the vicinity.\3570\ Another study 
found that schools with minority and underprivileged \3571\ children 
were disproportionately located within 250 meters of a major 
roadway.\3572\
---------------------------------------------------------------------------

    \3568\ Tian, N., J. Xue, and T. M. Barzyk. Evaluating 
socioeconomic and racial differences in traffic-related metrics in 
the United States using a GIS approach. Journal of Exposure Science 
and Environmental Epidemiology 23 (2):215 (2013). doi: 10.1038/
jes.2012.83. Available at: http://www.nature.com/articles/jes201283. 
(Accessed: May 31, 2018).
    \3569\ Boehmer, T.K., S.L. Foster, J.R. Henry, E.L. Woghiren-
Akinnifesi, and F.Y. Yip. Residential Proximity to Major Highways--
United States, 2010. Morbidity and Mortality Weekly Report 62(3):46-
50 (2013). Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/su6203a8.htm. (Accessed: February 26, 2018).
    \3570\ Rowangould, G.M. A Census of the US Near-roadway 
Population: Public Health and Environmental Justice Considerations. 
Transportation Research Part D: Transport and Environment 25:59-67 
(2013). doi:10.1016/j.trd.2013.08.003.
    \3571\ Public schools were determined to serve predominantly 
underprivileged students if they were eligible for Title I programs 
(federal programs that provide funds to school districts and schools 
with high numbers or high percentages of children who are 
disadvantaged) or had a majority of students who were eligible for 
free/reduced-price meals under the National School Lunch and 
Breakfast Programs.
    \3572\ Kingsley, S.L., M.N. Eliot, L. Carlson, J. Finn, D.L. 
MacIntosh, H.H. Suh, and G.A. Wellenius. Proximity of US Schools to 
Major Roadways: A Nationwide Assessment. Journal of Exposure Science 
and Environmental Epidemiology 24(3):253-59 (2014). doi:10.1038/
jes.2014.5.
---------------------------------------------------------------------------

    As detailed in Section 10.3.8 of the PRIA and Section X.E.11.a.2 of 
the FRIA, NHTSA and EPA analyzed two national databases that allowed 
evaluation of whether homes and schools were located near a major road 
and whether disparities in exposure may be occurring in these 
environments. The American Housing Survey (AHS) includes descriptive 
statistics of over 70,000 housing units across the nation. The study 
survey is conducted every two years by the U.S. Census Bureau. The 
second database the agencies analyzed was the U.S. Department of 
Education's Common Core of Data, which includes enrollment and location 
information for schools across the U.S.
    In analyzing the 2009 AHS, the focus was on whether or not a 
housing unit was located within 300 feet of a ``4-or-more lane highway, 
railroad, or airport.'' \3573\ Whether there were differences between 
households in such locations compared with those in locations farther 
from these transportation facilities was analyzed.\3574\ Other 
variables, such as land use category, region of country, and housing 
type, were included. Homes with a nonwhite householder were found to be 
22 to 34 percent more likely to be located within 300 feet of these 
large transportation facilities than homes with white householders. 
Homes with a Hispanic householder were 17 to 33 percent more likely to 
be located within 300 feet of these large transportation facilities 
than homes with non-Hispanic householders. Households near large 
transportation facilities were, on average, lower in income and 
educational attainment, more likely to be a rental property, and more 
likely to be located in an urban area compared with households more 
distant from transportation facilities.
---------------------------------------------------------------------------

    \3573\ This variable primarily represents roadway proximity. 
According to the Central Intelligence Agency's World Factbook, in 
2010, the United States had 6,506,204 km of roadways, 224,792 km of 
railways, and 15,079 airports. Highways thus represent the 
overwhelming majority of transportation facilities described by this 
factor in the AHS.
    \3574\ Bailey, C. (2011) Demographic and Social Patterns in 
Housing Units Near Large Highways and other Transportation Sources. 
Memorandum to docket.
---------------------------------------------------------------------------

    In examining schools near major roadways, the Common Core of Data 
(CCD) from the U.S. Department of Education, which includes information 
on all public elementary and secondary schools and school districts 
nationwide, was examined.\3575\ To determine school proximities to 
major roadways, a geographic information system (GIS) to map each 
school and roadways based on the U.S. Census's TIGER roadway file was 
used.\3576\ Minority students were found to be overrepresented at 
schools within 200 meters of the largest roadways, and schools within 
200 meters of the largest roadways also had higher than expected 
numbers of

[[Page 25261]]

students eligible for free or reduced-price lunches. For example, Black 
students represent 22 percent of students at schools located within 200 
meters of a primary road, whereas Black students represent 17 percent 
of students in all U.S. schools. Hispanic students represent 30 percent 
of students at schools located within 200 meters of a primary road, 
whereas Hispanic students represent 22 percent of students in all U.S. 
schools.
---------------------------------------------------------------------------

    \3575\ http://nces.ed.gov/ccd/.
    \3576\ Pedde, M.; Bailey, C. Identification of Schools within 
200 Meters of U.S. Primary and Secondary Roads. Memorandum to the 
docket (2011).
---------------------------------------------------------------------------

    Overall, there is substantial evidence that the population who 
lives or attends school near major roadways are more likely to be 
minority or low income. As described in Section VII.A.4.c.3.b.i, total 
downstream (tailpipe) emissions of criteria and toxic air pollutants 
for cars and light trucks in 2035 are projected to remain relatively 
unchanged or decrease under all action alternatives compared to the No 
Action Alternative, with the following exceptions: total downstream 
emissions of SO2 would increase under all action 
alternatives under both the CAFE and CO2 programs; total 
downstream emissions of acrolein would increase under Alternatives 5, 
6, and 7 under the CAFE program (but not under the CO2 
program); and total downstream emissions of acetaldehyde and butadiene 
would increase under Alternatives 6 and 7 under the CAFE program (but 
not under the CO2 program). To the extent minority and low-
income populations disproportionately live or attend schools near major 
roadways, these populations may be more likely to be affected by these 
emissions. However, because some pollutant emissions are expected to 
decrease and others are expected to increase, health impacts are mixed. 
Overall, as the magnitude of the emissions changes is anticipated to be 
minor compared to total tailpipe emissions for these vehicles, the 
impacts to minority or low-income populations are not considered high 
and adverse.
    The agencies used the standards that were discussed in the 2012 
rulemaking as the baseline for this rulemaking. Therefore, the agencies 
project increases in certain air pollutants for purposes of this 
analysis. However, as discussed above, one impact of the standards 
finalized in this rulemaking is to reduce the up-front cost of new and 
used vehicles. Low income populations may benefit most from the 
reduction in cost of acquiring newer vehicles, which generally are more 
fuel efficient and have lower air pollutant emissions than older 
vehicles. This cost reduction may have the effect of encouraging the 
quicker adoption of cleaner vehicles in low income communities, which 
could result in air quality and health benefits for those who live or 
attend school in proximity to the roadways where they are operated. To 
the degree to which minority populations may also live in proximity to 
these roadways, they would also experience benefits, thereby mitigating 
the disparity in racial, ethnic, and economically based exposures.
c) Other Vulnerabilities to Climate Change and Health Impacts of Air 
Pollutants
    Some areas most vulnerable to climate change tend to have a higher 
concentration of minority and low-income populations, potentially 
putting these communities at higher risk from climate variability and 
climate-related extreme weather events.\3577\ For example, urban areas 
tend to have pronounced social inequities that could result in 
disproportionately larger minority and low-income populations than 
those in the surrounding nonurban areas.\3578\ Urban areas are also 
subject to the most substantial temperature increases from climate 
change because of the urban heat island 
effect.3579 3580 3581 Taken together, these tendencies 
demonstrate a potential for disproportionate impacts on minority and 
low-income populations in urban areas. Low-income populations in 
coastal urban areas, which are vulnerable to increases in flooding as a 
result of projected sea-level rise, larger storm surges, and human 
settlement in floodplains, could also be disproportionately affected by 
climate change because they are less likely to have the means to 
evacuate quickly in the event of a natural disaster and, therefore, are 
at greater risk of injury and loss of life.3582 3583
---------------------------------------------------------------------------

    \3577\ U.S. Global Change Research Program (GCRP). Global 
Climate Change Impacts in the United States: The Third National 
Climate Assessment. U.S. Global Change Research Program. Melillo, 
J.M, T.C. Richmond, and G.W. Yohe (Eds.)]. U.S. Government Printing 
Office: Washington, DC 841 pp (2014). doi:10.7930/J0Z31WJ2. 
Available at: http://nca2014.globalchange.gov/report. (Accessed: 
February 27, 2018).
    \3578\ GCRP (2014).
    \3579\ GCRP (2014).
    \3580\ Knowlton, K., B. Lynn, R.A. Goldberg, C. Rosenzweig, C. 
Hogrefe, J.K. Rosenthal, and P.L. Kinney. Projecting Heat-related 
Mortality Impacts under a Changing Climate in the New York City 
Region. American Journal of Public Health 97(11):2028-34 (2007). 
doi:10.2105/AJPH.2006.102947. Available in: http://ajph.aphapublications.org/cgi/content/full/97/11/2028. Last 
accessed: March 4, 2018.
    \3581\ EPA. Heat Island Effect. U.S. Environmental Protection 
Agency (2017). Last revised: February 20, 2018. Available at: 
https://www.epa.gov/heat-islands. (Accessed: February 28, 2018.).
    \3582\ GCRP. Global Climate Impacts in the United States (2009). 
Cambridge, United Kingdom and New York, NY, USA. Karl, T.R., J.M. 
Melillo, and T.C. Peterson (Eds.). Cambridge University Press: 
Cambridge, UK. pp. 196.
    \3583\ GCRP (2014).
---------------------------------------------------------------------------

    Independent of their proximity to pollution sources or climate 
change, locations of potentially high impact, minority and low-income 
populations could be more vulnerable to the health impacts of 
pollutants and climate change. Reports from the U.S. Department of 
Health and Human Services have stated that minority and low-income 
populations tend to have less access to health care services, and the 
services received are more likely to suffer with respect to 
quality.3584 3585 3586 Other studies show that low 
socioeconomic position can modify the health effects of air pollution, 
with higher effects observed in groups with lower socioeconomic 
position.3587 3588 Possible explanations for this 
observation include that low socioeconomic position groups may be 
differentially exposed to air pollution or may be differentially 
vulnerable to effects of exposure.\3589\
---------------------------------------------------------------------------

    \3584\ U.S. Department of Health and Human Services (HHS). 
National Healthcare Disparities Report. U.S. Department of Health 
and Human Service. Rockville, MD, Agency for Healthcare Research and 
Quality (2003). Available at: http://archive.ahrq.gov/qual/nhdr03/nhdr03.htm. (Accessed: March 3, 2018).
    \3585\ HHS. Minority Health: Recent Findings. Agency for 
Healthcare Research Quality (2013). Last revised: February 2013. 
Available at: https://www.ahrq.gov/research/findings/factsheets/minority/minorfind/index.html. (Accessed: March 3, 2018).
    \3586\ HHS. 2016 National Healthcare Disparities Report. U.S. 
Department of Health and Human Service (2017). Rockville, MD. Agency 
for Healthcare Research and Quality. Available at: https://www.ahrq.gov/research/findings/nhqrdr/nhqdr16/summary.html. 
(Accessed: September 20, 2017).
    \3587\ O'Neill, M.S., M. Jerrett, I. Kawachi, J.I. Levy, A.J. 
Cohen, N. Gouveia, P. Wilkinson, T. Fletcher, L. Cifuentes, and J. 
Schwartz. Health, Wealth, and Air Pollution: Advancing Theory and 
Methods. Environmental Health Perspectives 111(16):1861-70 (2003). 
doi: 10.1289/ehp.6334. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241758/pdf/ehp0111-001861.pdf. (Accessed: February 
24, 2019).
    \3588\ Finkelstein, M.M.; Jerrett, M.; DeLuca, P.; Finkelstein, 
N.; Verma, D.K.; Chapman, K.; Sears, M.R. Relation between income, 
air pollution and mortality: a cohort study. Canadian Med Assn J 
169: 397-402 (2003).
    \3589\ O'Neill et al. (2003).
---------------------------------------------------------------------------

    In terms of climate change, increases in heat-related morbidity and 
mortality because of higher overall and extreme temperatures are likely 
to affect minority and low-income populations disproportionately, 
partially because of limited access to air conditioning and high energy 
costs.3590 3591 3592 3593 Native

[[Page 25262]]

American tribes and Alaskan Native villages are also more susceptible 
to the impacts of climate change, as these groups often 
disproportionately rely on natural resources for livelihoods, 
medicines, and cultural and spiritual purposes.\3594\ Moreover, coastal 
tribal communities may have to relocate because of sea-level rise, 
erosion, and permafrost thaw.\3595\ NHTSA's FEIS provides additional 
discussion of health and societal impacts of climate change on 
indigenous communities in Section 8.6.5.2, Sectoral Impacts of Climate 
Change, under Human Health and Human Security.
---------------------------------------------------------------------------

    \3590\ EPA. 2009. Technical Support Document for Endangerment 
and Cause or Contribute Findings for Greenhouse Gases under Section 
202(a) of the Clean Air Act. December 7, 2009. U.S. Environmental 
Protection Agency, Office of Atmospheric Programs, Climate Change 
Division: Washington, DC Available at: https://www.epa.gov/sites/production/files/2016-08/document/endangerment_tsd.pdf. (Accessed: 
February 28, 2018).
    \3591\ O'Neill, M.S., A. Zanobetti, and J. Schwartz. Disparities 
by Race in Heat-Related Mortality in Four US Cities: The Role of Air 
Conditioning Prevalence. Journal of Urban Health 82(2):191-97 
(2005). doi:10.1093/jurban/jti043. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3456567/pdf/11524_2006_Article_375.pdf. (Accessed: March 4, 2018).
    \3592\ GCRP (2014).
    \3593\ Harlan, S.L. and D.M. Ruddell. Climate Change and Health 
in Cities: Impacts of Heat and Air Pollution and Potential Co-
Benefits from Mitigation and Adaptation. Current Opinion in 
Environmental Sustainability 3(3):126-34 (2011). doi: 10.1016/
j.cosust.2011.01.001.
    \3594\ National Tribal Air Association. 2009. Impacts of climate 
change on Tribes in the United States. Submitted December 11, 2009 
to Assistant Administrator Gina McCarthy, USEPA, Office of Air and 
Radiation. Available at: http://www.epa.gov/air/tribal/pdfs/Impacts%20of%20Climate%20Change%20on%20Tribes%20in%20the%20United%20States.pdf. Last accessed: February 24, 2019.
    \3595\ Maldonado, J., C. Shearer, R. Bronen, K. Peterson, and H. 
Lazrus. The Impact of Climate Change on Tribal Communities in the 
US: Displacement, Relocation, and Human Rights. Climatic Change 
120(3):601-14 (2013).
---------------------------------------------------------------------------

    Together, this information indicates that the same set of potential 
environmental effects (e.g., air pollutants, heat increases, and sea-
level rise) may disproportionately affect minority and low-income 
populations because of socioeconomic circumstances or histories of 
discrimination and inequity.
    As described in Chapter 5 of NHTSA's FEIS, the action alternatives 
are projected to increase CO2 emissions from passenger cars 
and light trucks by 4 to 10 percent by 2100 compared to the No Action 
Alternative. Impacts of climate change could disproportionately affect 
minority and low-income populations in urban areas that are subject to 
the most substantial temperature increases from climate change. These 
impacts are largely because of the urban heat island effect. 
Additionally, minority and low-income populations that live in flood-
prone coastal areas could be disproportionately affected. However, the 
contribution of the action alternatives to climate change impacts would 
be very minor rather than high and adverse. Compared to the annual U.S. 
CO2 emissions of 7,193 MMTCO2e from all sources 
by the end of the century projected by the GCAM Reference scenario, the 
action alternatives are projected to increase annual U.S. 
CO2 emissions by 0.4 to 1.2 percent in 2100. Compared to 
annual global CO2 emissions, the action alternatives would 
represent an even smaller percentage increase and ultimately, by 2100, 
are projected to result in percentage increases in global mean surface 
temperature, atmospheric CO2 concentrations, and sea level, 
and decreases in ocean pH, ranging from 0.09 percent to less than 0.01 
percent. Any impacts of this rulemaking on low-income and minority 
communities would be attenuated by a lengthy causal chain; but if one 
could attempt to draw those links, the changes to climate values would 
be very small and incremental compared to the expected changes 
associated with the emissions trajectories in the GCAM Reference 
scenario.
    As reported in Section VII.A.4.c.3.c above, adverse health impacts 
over the lifetimes of vehicles through MY 2029 are projected to 
increase nationwide under each of the action alternatives (except 
Alternative 6 and Alternative 7 under the CAFE program, which show 
decreases) compared to the No Action Alternative. Increases in these 
pollutant emissions, however, would be primarily the result of 
increases in upstream emissions (emissions near refineries, power 
plants, and extraction sites), while downstream emissions (tailpipe 
emissions near roadways) are anticipated to decrease or increase by 
smaller amounts. The health impacts reported in that section occur over 
a long period of time, would be incremental in magnitude, and would not 
be characterized as high. Those impacts would also be borne nationwide, 
so impacts to minority and low-income populations would be smaller.
d) Conclusion
    Based on the foregoing, the agencies have determined that this 
rulemaking (and alternatives considered) would not result in 
disproportionately high and adverse human health or environmental 
effects on minority or low-income populations. This rulemaking would 
set standards nationwide, and although minority and low-income 
populations may experience some disproportionate effects, in particular 
locations, the overall impacts on human health and the environment 
would not be ``high and adverse'' under E.O. 12898.
    Furthermore, the agencies note that there are no mitigation 
measures or alternatives available as part of this action that could 
fulfill the respective statutory missions of the agencies and that 
would address the considerations discussed in Section VIII (e.g., 
economic practicability) or avoid or reduce any disproportionate 
effects in particular locations experienced by minority and low-income 
populations. The impacts described in this analysis would result from 
air pollutant and CO2 emissions that may occur from the 
levels of stringency selected by the agencies. However, for the reasons 
described in Section VIII, the agencies cannot select a higher level of 
stringency. While the agencies have considered the potential impacts 
described in this analysis, there is a substantial need, based on the 
overall public interest, to address the costs associated with the 
standards discussed in the 2012 rulemaking. More stringent alternatives 
would have severe adverse social and economic costs, as described in 
Section VIII, and necessitate the level of standards finalized in this 
rulemaking.
12. Executive Order 13045: ``Protection of Children From Environmental 
Health Risks and Safety Risks''
    This action is subject to E.O. 13045 (62 FR 19885, April 23, 1997) 
because it is an economically significant regulatory action as defined 
by E.O. 12866, and the agencies have reason to believe that the 
environmental health or safety risks related to this action may have a 
disproportionate effect on children. Specifically, children are more 
vulnerable to adverse health effects related to mobile source 
emissions, as well as to the potential long-term impacts of climate 
change. Pursuant to E.O. 13045, NHTSA and EPA must prepare an 
evaluation of the environmental health or safety effects of the planned 
regulation on children and an explanation of why the planned regulation 
is preferable to other potentially effective and reasonably feasible 
alternatives considered by the agencies. Further, this analysis may be 
included as part of any other required analysis.
    This preamble and NHTSA's Final EIS discuss air quality, climate 
change, and their related environmental and health effects, noting 
where these would disproportionately affect children. The EPA 
Administrator has also discussed the impact of climate-related health 
effects on children in the Endangerment and Cause or Contribute 
Findings for

[[Page 25263]]

Greenhouse Gases Under Section 202(a) of the Clean Air Act (74 FR 
66496, December 15, 2009). In addition, this preamble explains why the 
agencies' final standards are preferable to other alternatives 
considered. Together, this preamble and NHTSA's Final EIS satisfy the 
agencies' responsibilities under E.O. 13045.

F. Regulatory Flexibility Act

    Pursuant to the Regulatory Flexibility Act (5 U.S.C. 601 et seq., 
as amended by the Small Business Regulatory Enforcement Fairness Act 
(SBREFA) of 1996), whenever an agency is required to publish a notice 
of proposed rulemaking or final rule, it must prepare and make 
available for public comment a regulatory flexibility analysis that 
describes the effect of the rule on small entities (i.e., small 
businesses, small organizations, and small governmental jurisdictions). 
No regulatory flexibility analysis is required if the head of an agency 
certifies the rule will not have a significant economic impact on a 
substantial number of small entities. SBREFA amended the Regulatory 
Flexibility Act to require Federal agencies to provide a statement of 
the factual basis for certifying that a rule will not have a 
significant economic impact on a substantial number of small entities.
    Two comments argued that the agencies should prepare a regulatory 
flexibility analysis and convene a small business review panel to 
assess the impacts in accordance with the Regulatory Flexibility Act, 5 
U.S.C. 601 et seq., as amended by SBREFA.\3596\ The agencies considered 
these comments and the impacts of this rule under the Regulatory 
Flexibility Act and certify that this rule will not have a significant 
economic impact on a substantial number of small entities. The 
following is the agencies' statement providing the factual basis for 
this certification pursuant to 5 U.S.C. 605(b).
---------------------------------------------------------------------------

    \3596\ See National Coalition for Advanced Transportation (NCAT) 
Comment, Docket No. NHTSA-2018-0067-11969, at 64-65; Workhorse 
Group, Inc. Comment, Docket No. NHTSA-2018-0067-12215, at 1-2.
---------------------------------------------------------------------------

    Small businesses are defined based on the North American Industry 
Classification System (NAICS) code.\3597\ One of the criteria for 
determining size is the number of employees in the firm. For 
establishments primarily engaged in manufacturing or assembling 
automobiles, as well as light duty trucks, the firm must have less than 
1,500 employees to be classified as a small business. This rule would 
affect motor vehicle manufacturers. As shown in Table X-1, the agencies 
have identified 15 small manufacturers of passenger cars, light trucks, 
and SUVs of electric, hybrid, and internal combustion engines.\3598\ 
The agencies acknowledge that some newer manufacturers may not be 
listed. However, those new manufacturers tend to have transportation 
products that are not part of the light-duty vehicle fleet and have yet 
to start production of light-duty vehicles. Moreover, NHTSA does not 
believe that there are a ``substantial number'' of these newer 
companies.\3599\
---------------------------------------------------------------------------

    \3597\ Classified in NAICS under Subsector 336--Transportation 
Equipment Manufacturing for Automobile Manufacturing (336111), Light 
Truck (336112), and Heavy Duty Truck Manufacturing (336120). https://www.sba.gov/document/support-table-size-standards.
    \3598\ Two comments pointed out that Workhorse Group Inc. was 
not listed as a small domestic vehicle manufacturer in Table XII-1 
of the proposal. See National Coalition for Advanced Transportation 
(NCAT) Comment, Docket No. NHTSA-2018-0067-11969, at 64-65; 
Workhorse Group, Inc. Comment, Docket No. NHTSA-2018-0067-12215, at 
1-2. Workhorse Group has been added to the table here, but neither 
its addition nor the existence of a small number of other new small 
manufacturers does not alter the conclusion that this rule will not 
have a significant economic impact on a substantial number of small 
entities.
    \3599\ 5 U.S.C. 605(b).
    [GRAPHIC] [TIFF OMITTED] TR30AP20.758
    

[[Page 25264]]


    NHTSA believes that the rulemaking would not have a significant 
economic impact on the small vehicle manufacturers because under 49 CFR 
part 525, passenger car manufacturers making less than 10,000 vehicles 
per year can petition NHTSA to have alternative standards set for those 
manufacturers. These manufacturers do not currently meet the 27.5 mpg 
standard and must already petition the agency for relief. If the 
standard is raised, it has no meaningful impact on these 
manufacturers--they still must go through the same process and petition 
for relief. Given there already is a mechanism for relieving burden on 
small businesses, which is the purpose of the Regulatory Flexibility 
Act, a regulatory flexibility analysis was not prepared.
---------------------------------------------------------------------------

    \3600\ Estimated number of employees as of 2018, source: 
Linkedin.com.
    \3601\ Rough estimate of light duty vehicle production for model 
year 2017.
---------------------------------------------------------------------------

    Two comments argued that small manufacturers of electric vehicles 
would face a significant economic impact because their ability to earn 
credits would be ``substantially diminished.'' \3602\ The method for 
earning credits applies equally across manufacturers and does not place 
small entities at a significant competitive disadvantage. In any event, 
even if the rule had a ``significant economic impact'' on these small 
EV manufacturers, the amount of these companies is not ``a substantial 
number.'' \3603\ For these reasons, their existence does not alter the 
agencies' analysis of the applicability of the Regulatory Flexibility 
Act. EPA believes this rulemaking would not have a significant economic 
impact on a substantial number of small entities under the Regulatory 
Flexibility Act, as amended by the Small Business Regulatory 
Enforcement Fairness Act. EPA is exempting from the CO2 
standards any manufacturer, domestic or foreign, meeting SBA's size 
definitions of small business as described in 13 CFR 121.201. EPA 
adopted the same type of exemption for small businesses in the 2017 and 
later rulemaking. EPA estimates that small entities comprise less than 
0.1 percent of total annual vehicle sales and exempting them will have 
a negligible impact on the CO2 emissions reductions from the 
standards. Because EPA is exempting small businesses from the 
CO2 standards, the agency certifies that the rule will not 
have a significant economic impact on a substantial number of small 
entities. Therefore, EPA has not conducted a Regulatory Flexibility 
Analysis or a SBREFA SBAR Panel for the rule.
---------------------------------------------------------------------------

    \3602\ National Coalition for Advanced Transportation (NCAT) 
Comment, Docket No. NHTSA-2018-0067-11969, at 65; Workhorse Group, 
Inc. Comment, Docket No. NHTSA-2018-0067-12215, at 2.
    \3603\ 5 U.S.C. 605.
---------------------------------------------------------------------------

    EPA regulations allow small businesses voluntarily to waive their 
small business exemption and optionally to certify to the 
CO2 standards. This option allows small entity manufacturers 
to earn CO2 credits under the CO2 program, if 
their actual fleetwide CO2 performance is better than their 
fleetwide CO2 target standard. However, the exemption waiver 
is optional for small entities and thus the agency believes that 
manufacturers opt into the CO2 program if it is economically 
advantageous for them to do so, for example in order to generate and 
sell CO2 credits. Therefore, EPA believes this voluntary 
option does not affect EPA's determination that the standards will 
impose no significant adverse impact on small entities.

G. Executive Order 13132 (Federalism)

    Executive Order 13132 requires Federal agencies to develop an 
accountable process to ensure ``meaningful and timely input by State 
and local officials in the development of regulatory policies that have 
federalism implications.'' The Order defines the term ``[p]olicies that 
have federalism implications'' to include regulations that have 
``substantial direct effects on the States, on the relationship between 
the national government and the States, or on the distribution of power 
and responsibilities among the various levels of government.'' Under 
the Order, agencies may not issue a regulation that has federalism 
implications, that imposes substantial direct compliance costs, unless 
the Federal government provides the funds necessary to pay the direct 
compliance costs incurred by State and local governments, or the 
agencies consult with State and local officials early in the process of 
developing the proposed regulation. The agencies complied with the 
Order's requirements.
    NHTSA also addressed the federalism implications of its proposal in 
The Safer Affordable Fuel-Efficient Vehicles Rule Part One: One 
National Program final rulemaking.\3604\
---------------------------------------------------------------------------

    \3604\ 84 FR 51310 (Sep. 27, 2019).
---------------------------------------------------------------------------

H. Executive Order 12988 (Civil Justice Reform)

    Pursuant to Executive Order 12988, ``Civil Justice Reform,'' \3605\ 
NHTSA has considered whether this rulemaking would have any retroactive 
effect. This proposed rule does not have any retroactive effect.
---------------------------------------------------------------------------

    \3605\ 61 FR 4729 (Feb. 7, 1996).
---------------------------------------------------------------------------

I. Executive Order 13175 (Consultation and Coordination With Indian 
Tribal Governments)

    This final rule does not have tribal implications, as specified in 
Executive Order 13175 (65 FR 67249, November 9, 2000). This rule will 
be implemented at the Federal level and impose compliance costs only on 
vehicle manufacturers. Thus, Executive Order 13175 does not apply to 
this rule. Some comments complained that the agencies have not 
consulted or coordinated with Native American communities and Indian 
Tribes in promulgating this rule.\3606\ Executive Order 13175 requires 
consultation with Tribal officials when agencies are developing 
policies that have ``substantial direct effects'' on Tribes and Tribal 
interests.\3607\ Even accepting the comments' description of the 
effects of the rule, they have identified only indirect effects of the 
standards on Tribal interests.\3608\
---------------------------------------------------------------------------

    \3606\ See, e.g., CARB Comment, Docket No. NHTSA-2018-0067-
11873, at 412; National Tribal Air Association Comment, Docket No. 
NHTSA-2018-0067-11948, at 4; Keweenaw Bay Indian Community Comment, 
Docket No. EPA-HQ-OAR-2018-0283-3325, at 1-2; Fond du Lac Band of 
Lake Superior Chippewa Comment, Docket No. EPA-HQ-OAR-2018-0283-
4030, at 3; Sac and Fox Nation, Docket No. EPA-HQ-OAR-2018-0283-
4159, at 4-5; The Leech Lake Band of Ojibwe Comment, Docket No. EPA-
HQ-OAR-2018-0283-5931, at 4-5.
    \3607\ 65 FR 67249, 67249 (Nov. 6, 2000).
    \3608\ See, e.g., National Tribal Air Association Comment, 
Docket No. NHTSA-2018-0067-11948, at 4.
---------------------------------------------------------------------------

J. Unfunded Mandates Reform Act

    Section 202 of the Unfunded Mandates Reform Act of 1995 (UMRA) 
requires Federal agencies to prepare a written assessment of the costs, 
benefits, and other effects of a proposed or final rule that includes a 
Federal mandate likely to result in the expenditure by State, local, or 
Tribal governments, in the aggregate, or by the private sector, of more 
than $100 million in any one year (adjusted for inflation with base 
year of 1995). Adjusting this amount by the implicit gross domestic 
product price deflator for 2016 results in $148 million (111.416/75.324 
= 1.48).\3609\ Before promulgating a rule for which a written statement 
is needed, section 205 of UMRA generally requires NHTSA and EPA to 
identify and consider a reasonable number of regulatory

[[Page 25265]]

alternatives and adopt the least costly, most cost-effective, or least 
burdensome alternative that achieves the objective of the rule. The 
provisions of section 205 do not apply when they are inconsistent with 
applicable law. Moreover, section 205 allows NHTSA and EPA to adopt an 
alternative other than the least costly, most cost-effective, or least 
burdensome alternative if the agency publishes with the rule an 
explanation of why that alternative was not adopted.
---------------------------------------------------------------------------

    \3609\ Bureau of Economic Analysis, National Income and Product 
Accounts (NIPA), Table 1.1.9 Implicit Price Deflators for Gross 
Domestic Product. https://bea.gov/iTable/index_nipa.cfm.
---------------------------------------------------------------------------

    This rule will not result in the expenditure by State, local, or 
Tribal governments, in the aggregate, of more than $148 million 
annually, but it will result in the expenditure of that magnitude by 
vehicle manufacturers and/or their suppliers. In developing this rule, 
NHTSA and EPA considered a variety of alternative average fuel economy 
standards lower and higher than those previously proposed. The fuel 
economy standards for MYs 2021-2026 are the least costly, most cost-
effective, and least burdensome alternative that achieve the objectives 
of the rule.

K. Regulation Identifier Number

    The Department of Transportation assigns a regulation identifier 
number (RIN) to each regulatory action listed in the Unified Agenda of 
Federal Regulations. The Regulatory Information Service Center 
publishes the Unified Agenda in April and October of each year. The RIN 
contained in the heading at the beginning of this document may be used 
to find this action in the Unified Agenda.

L. National Technology Transfer and Advancement Act

    Section 12(d) of the National Technology Transfer and Advancement 
Act (NTTAA) requires NHTSA and EPA to evaluate and use existing 
voluntary consensus standards in its regulatory activities unless doing 
so would be inconsistent with applicable law (e.g., the statutory 
provisions regarding NHTSA's vehicle safety authority, or EPA's testing 
authority) or otherwise impractical.\3610\
---------------------------------------------------------------------------

    \3610\ 15 U.S.C. 272.
---------------------------------------------------------------------------

    Voluntary consensus standards are technical standards developed or 
adopted by voluntary consensus standards bodies. Technical standards 
are defined by the NTTAA as ``performance-based or design-specific 
technical specification and related management systems practices.'' 
They pertain to ``products and processes, such as size, strength, or 
technical performance of a product, process or material.''
    Examples of organizations generally regarded as voluntary consensus 
standards bodies include the American Society for Testing and Materials 
(ASTM), the Society of Automotive Engineers (SAE), and the American 
National Standards Institute (ANSI). If the agencies do not use 
available and potentially applicable voluntary consensus standards, 
they are required by the Act to provide Congress, through OMB, an 
explanation of the reasons for not using such standards.
    For CO2 emissions, EPA will collect data over the same 
tests that are used for the MY 2012-2016 CO2 standards and 
for the CAFE program. This unified data collection will minimize the 
amount of testing done by manufacturers because manufacturers are 
already required to run these tests. For A/C credits, EPA will use a 
consensus methodology developed by the Society of Automotive Engineers 
(SAE) and also a new A/C test. EPA knows of no consensus standard 
available for the A/C test.
    There are currently no voluntary consensus standards that NHTSA 
administers relevant to today's CAFE standards.

M. Department of Energy Review

    In accordance with 49 U.S.C. 32902(j)(2), NHTSA submitted this rule 
to the Department of Energy for review.

N. Paperwork Reduction Act

    The Paperwork Reduction Act (PRA) of 1995, Public Law 104-13,\3611\ 
gives OMB authority to regulate matters regarding the collection, 
management, storage, and dissemination of certain information by and 
for the Federal government. It seeks to reduce the total amount of 
paperwork handled by the government and the public. NHTSA strives to 
reduce the public's information collection burden hours each fiscal 
year by streamlining external and internal processes.
---------------------------------------------------------------------------

    \3611\ Codified at 44 U.S.C. 3501 et seq.
---------------------------------------------------------------------------

    To this end, NHTSA will continue to collect information to ensure 
compliance with its CAFE program. NHTSA will reinstate its previously-
approved collection of information for Corporate Average Fuel Economy 
(CAFE) reports specified in 49 CFR part 537 (OMB control number 2127-
0019), add the additional burden for reporting changes adopted in the 
October 15, 2012 final rule that recently came into effect (see 77 FR 
62623), and account for the change in burden in this rule as well as 
for other CAFE reporting provisions required by Congress and NHTSA. 
NHTSA is also changing the name of this collection to represent more 
accurately the breadth of all CAFE regulatory reporting. Although NHTSA 
is adding additional burden hours to its CAFE report requirement in 49 
CFR 537, the agency believes there will be a reduction in the overall 
paperwork burden due to the standardization of data and the streamlined 
process.
    In compliance with the PRA, the information collection request 
(ICR) abstracted below was forwarded to OMB for review and comment. The 
ICR describes the nature of the information collection and its expected 
burden.
    Title: Corporate Average Fuel Economy.
    Type of Request: Reinstatement and amendment of a previously 
approved collection.
    OMB Control Number: 2127-0019.
    Form Numbers: NHTSA Form 1474 (CAFE Projections Reporting Template) 
and NHTSA Form 1475 (CAFE Credit Template).
    Requested Expiration Date of Approval: Three years from date of 
approval.
    Summary of the collection of information: As part of this 
rulemaking, NHTSA is reinstating and modifying its previously-approved 
collection for CAFE-related collections of information. NHTSA and EPA 
have coordinated their compliance and reporting requirements in an 
effort not to impose duplicative burdens on regulated entities. This 
information collection contains three different components: Burden 
related to NHTSA's CAFE reporting requirements; burden related to CAFE 
compliance, but not via reporting requirements; and information 
gathered by NHTSA to help inform CAFE analyses. All templates 
referenced in this section will be available in the rulemaking docket 
and the NHTSA public information center.\3612\
---------------------------------------------------------------------------

    \3612\ https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
---------------------------------------------------------------------------

CAFE Compliance Reports
    NHTSA is reinstating \3613\ its collection related to the reporting 
requirements in 49 U.S.C. 32907, ``Reports and tests of 
manufacturers.'' In that section, manufacturers are statutorily 
required to submit CAFE compliance reports to the Secretary of 
Transportation.\3614\ The reports must state if a manufacturer will 
comply with its applicable fuel economy standard(s), describe what 
actions the manufacturer

[[Page 25266]]

intends to take to comply with the standard(s), and include other 
information as required by NHTSA. Manufacturers are required to submit 
two CAFE compliance reports--a pre-model year report (PMY) and a mid-
model year (MMY) report--each year. In the event a manufacturer needs 
to correct previously-submitted information, a manufacturer may need to 
file additional reports.\3615\
---------------------------------------------------------------------------

    \3613\ This collection expired on April 30, 2016.
    \3614\ 49 U.S.C. 32907 (delegated to the NHTSA Administrator at 
49 CFR 1.95). Because of this delegation, for purposes of 
discussion, statutory references to the Secretary of Transportation 
in this section will be discussed in terms of NHTSA or the NHTSA 
Administrator.
    \3615\ Specifically, a manufacturer shall submit a report 
containing the information during the 30 days before the beginning 
of each model year, and during the 30 days beginning the 180th day 
of the model year. When a manufacturer decides that actions reported 
are not sufficient to ensure compliance with that standard, the 
manufacturer shall report additional actions it intends to take to 
comply with the standard and include a statement about whether those 
actions are sufficient to ensure compliance.
---------------------------------------------------------------------------

    To implement this statute, NHTSA issued 49 CFR part 537, 
``Automotive Fuel Economy Reports,'' which adds additional definition 
to the terms of section 32907. The first report, the PMY report must be 
submitted to NHTSA before December 31 of the calendar year prior to the 
corresponding model year and contain manufacturers' projected 
information for that upcoming model year. The second report, the MMY 
report must be submitted by July 31 of the given model year and contain 
updated information from manufacturers based on actual and projected 
information known midway through the model year. Finally, the last 
report, a supplementary report, is required to be submitted anytime a 
manufacturer needs to correct information previously submitted to 
NHTSA.
    Compliance reports must include information on passenger and non-
passenger automobiles (trucks) describing the projected and actual fuel 
economy standards, fuel economy performance values, production sales 
volumes and information on vehicle design features (e.g., engine 
displacement and transmission class) and other vehicle attribute 
characteristics (e.g., track width, wheel base, and other light truck 
off-road features). Manufacturers submit confidential and non-
confidential versions of these reports to NHTSA. Confidential reports 
differ by including estimated or actual production sales information, 
which is withheld from public disclosure to protect each manufacturer's 
competitive sales strategies. NHTSA uses the reports as the basis for 
vehicle auditing and testing, which helps manufacturers correct 
reporting errors prior to the end of the model year and facilitate 
acceptance of their final CAFE report by the Environmental Protection 
Agency (EPA). The reports also help the agency, as well as the 
manufacturers who prepare them, anticipate potential compliance issues 
as early as possible, and help manufacturers plan their compliance 
strategies.
    Further, NHTSA is modifying this collection to account for 
additional information manufacturers are required to include in their 
reports. In the CAFE standards previously promulgated for MY 2017 and 
beyond,\3616\ NHTSA allowed for manufacturers to gain additional fuel 
economy benefits by installing certain technologies on their vehicles 
beginning with MY 2017.\3617\ These technologies include air-
conditioning systems with increased efficiency, off-cycle technologies 
whose benefits are not adequately captured on the Federal Test 
Procedure and/or the Highway Fuel Economy Test,\3618\ and hybrid 
electric technologies installed on full-size pickup trucks. Prior to MY 
2017, manufacturers were unable to earn a fuel economy benefit for 
these technologies, so NHTSA's reporting requirements did not include 
an opportunity to report them. Now, manufacturers must provide 
information on these technologies in their CAFE reports. NHTSA requires 
manufacturers to provide detailed information on the model types using 
these technologies to gain fuel economy benefits. These details are 
necessary to facilitate NHTSA's technical analyses and to ensure the 
agency can perform random enforcement audits when necessary.
---------------------------------------------------------------------------

    \3616\ 77 FR 62623 (Oct. 15, 2012).
    \3617\ These technologies were not included in the burden for 
part 537 at the time as the additional reporting requirements would 
not take effect until years later.
    \3618\ E.g., engine idle stop-start systems, active transmission 
warmup systems, etc.
---------------------------------------------------------------------------

    In addition to a list of all fuel consumption improvement 
technologies utilized in their fleet, 49 CFR 537 requires manufacturers 
to report the make, model type, compliance category, and production 
volume of each vehicle equipped with each technology and the associated 
fuel consumption improvement value (FCIV). NHTSA is adding the 
reporting and enforcement burden hours and cost for these new 
incentives to this collection. Manufacturers can also petition the EPA 
and NHTSA, in accordance with 40 CFR 86.1868-12 or 40 CFR 86.1869-12, 
to gain additional credits based upon the improved performance of any 
of the new incentivized technologies allowed starting in model year 
2017. EPA approves these petitions in collaboration with NHTSA and any 
adjustments are taken into account for both programs. As a part the 
agencies' coordination, NHTSA provides EPA with an evaluation of each 
new technology to ensure its direct impact on fuel economy and an 
assessment on the suitability of each technology for use in increasing 
a manufacturer's fuel economy performance. Furthermore, at times, NHTSA 
may independently request additional information from a manufacturer to 
support its evaluations. This information along with any research 
conclusions shared with EPA and NHTSA in the petitions is required to 
be submitted in manufacturer's CAFE reports.
    NHTSA is also changing the burden hours for its CAFE reporting 
requirements in 49 CFR part 537 by adjusting the total amount of time 
spent collecting the required reporting information through the use of 
a standardized reporting template to streamline the collection process. 
The standardized template will be used by manufacturers to collect all 
the required CAFE information under 49 CFR 537.7(b) and (c) and 
provides a format which ensures accuracy, completeness, and better 
alignment with the final data provided to EPA.
2. Other CAFE Compliance Collections
    NHTSA is adopting a new standardized template for manufacturers 
buying CAFE credits and for manufacturers submitting credit 
transactions in accordance with 49 CFR part 536. In 49 CFR part 
536.5(d), NHTSA is required to assess compliance with fuel economy 
standards each year, utilizing the certified and reported CAFE data 
provided by the EPA for enforcement of the CAFE program pursuant to 49 
U.S.C. 32904(e). Credit values are calculated based on the CAFE data 
from the EPA. If a manufacturer's vehicles in a particular compliance 
category performs better than its required fuel economy standard, NHTSA 
adds credits to the manufacturer's account for that compliance 
category. If a manufacturer's vehicles in a particular compliance 
category perform worse than the required fuel economy standard, NHTSA 
will add a credit deficit to the manufacturer's account and will 
provide written notification to the manufacturer concerning its failure 
to comply. The manufacturer will be required to confirm the shortfall 
and must either: Submit a plan indicating how it will allocate existing 
credits or earn, transfer, and/or acquire credits or pay the equivalent 
civil penalty. The manufacturer must submit a plan or

[[Page 25267]]

payment within 60 days of receiving notification from NHTSA.
    Manufacturers should use the credit transaction template any time a 
credit transaction request is sent to NHTSA. For example, manufacturers 
that purchase credits and want to apply them to their credit accounts 
will use the credit transaction template. The template NHTSA is 
adopting is a simple spreadsheet that credit entities fill out. When 
completed, credit entities will have an organized list of credit 
transactions and will be able to click a button on the spreadsheet to 
generate a joint transaction letter for trading parties to sign and 
submit to NHTSA, along with the spreadsheet. Entities trading credits 
are also required to provide to NHTSA all the confidential information 
associated with the monetary and non-monetary price of credit trades. 
NHTSA believes these changes will significantly reduce the burden on 
manufacturers in managing their CAFE credit accounts and provide better 
oversight of the CAFE credit program for NHTSA.
    Finally, NHTSA is accounting for the additional burden due to 
existing CAFE program elements. In 49 CFR part 525, small volume 
manufacturers submit petitions to NHTSA for exemption from an 
applicable average fuel economy standard and to request to comply with 
a less stringent alternative average fuel economy standard. In 49 CFR 
part 534, manufacturers are required to submit information to NHTSA 
when establishing a corporate controlled relationship with another 
manufacturer. A controlled relationship exists between manufacturers 
that control, are controlled by, or are under common control with, one 
or more other manufacturers. Accordingly, manufacturers that have 
entered into written contracts transferring rights and responsibilities 
to other manufacturers in controlled relationships for CAFE purposes 
are required to provide reports to NHTSA. There are additional 
reporting requirements for manufacturers submitting carry back plans 
and when manufacturers split apart from controlled relationships and 
must designate how credits are to be allocated between the 
parties.\3619\ Manufacturers with credit deficits at the end of the 
model year, can carry back future earned credits up to three model 
years in advance of the deficit to resolve a current shortfall. The 
carryback plan proving the existence of a manufacturer's future earned 
credits must be submitted and approved by NHTSA, pursuant to 49 U.S.C. 
32903(b).
---------------------------------------------------------------------------

    \3619\ See 49 CFR part 536.
---------------------------------------------------------------------------

3. Analysis Fleet Composition
    As discussed in Section VI.B, in setting CAFE standards, NHTSA 
creates an analysis fleet from which to model potential future economy 
improvements. To compose this fleet, the agency uses a mixture of 
compliance data and information from other sources to replicate more 
closely the fleet from a recent model year. While refining the analysis 
fleet, NHTSA occasionally asks manufacturers for information that is 
similar to information submitted as part of EPA's final model year 
report (e.g., final model year vehicle volumes). Periodically, NHTSA 
may ask manufacturers for more detailed information than what is 
required for compliance (e.g., what engines are shared across vehicle 
models). Often, NHTSA requests this information from manufacturers 
after manufacturers have submitted their final model year reports to 
EPA, but before EPA processes and releases final model year reports.
    Information like this, which is used to verify and supplement the 
data used to create the analysis fleet, is tremendously valuable to 
generating an accurate analysis fleet, and setting maximum feasible 
standards. The more accurate the analysis fleet is, the more accurate 
the modeling of what technologies could be applied will be. Therefore, 
NHTSA is accounting for the burden on manufacturers to provide the 
agency with this additional information. In almost all instances, 
manufacturers already have the information NHTSA seeks, but it might 
need to be reformatted or recompiled. Because of this, NHTSA believes 
the burden to provide this information will often be minimal.
    Affected Public: Respondents are manufacturers of engines and 
vehicles within the North American Industry Classification System 
(NAICS) and use the coding structure as defined by NAICS including 
codes 33611, 336111, 336112, 33631, 33631, 33632, 336320, 33635, and 
336350 for motor vehicle and parts manufacturing.
    Respondent's obligation to respond: Regulated entities are required 
to respond to inquiries covered by this collection. 49 U.S.C. 32907. 49 
CFR part 525, 534, 536, and 537.
    Frequency of response: Variable, based on compliance obligation. 
Please see PRA supporting documentation in the docket for more detailed 
information.
    Average burden time per response: Variable, based on compliance 
obligation. Please see PRA supporting documentation in the docket for 
more detailed information.
    Number of respondents: 23.
4. Estimated Total Annual Burden Hours and Costs:
[GRAPHIC] [TIFF OMITTED] TR30AP20.759

[GRAPHIC] [TIFF OMITTED] TR30AP20.760


[[Page 25268]]



O. Privacy Act

    In accordance with 5 U.S.C. 553(c), the agencies solicited comments 
from the public to inform the rulemaking process better. These comments 
are posted, without edit, to www.regulations.gov, as described in DOT's 
system of records notice, DOT/ALL-14 FDMS, accessible through 
www.transportation.gov/privacy. In order to facilitate comment tracking 
and response, the agencies encouraged commenters to provide their 
names, or the names of their organizations; however, submission of 
names is completely optional.

List of Subjects

40 CFR Part 86

    Administrative practice and procedure, Confidential business 
information, Incorporation by reference, Labeling, Motor vehicle 
pollution, Reporting and recordkeeping requirements.

40 CFR Part 600

    Administrative practice and procedure, Electric power, Fuel 
economy, Labeling, Reporting and recordkeeping requirements.

49 CFR Parts 523, 531, and 533

    Fuel economy.

49 CFR Parts 536 and 537

    Fuel economy, Reporting and recordkeeping requirements.

Environmental Protection Agency

40 CFR Chapter I

    For the reasons set forth in the preamble, the Environmental 
Protection Agency is amending part 86 of title 40, Chapter I of the 
Code of Federal Regulations as follows:

PART 86--CONTROL OF EMISSIONS FROM NEW AND IN-USE HIGHWAY VEHICLES 
AND ENGINES

0
1. The authority citation for part 86 continues to read as follows:

    Authority: 42 U.S.C. 7401-7671q.

0
2. Section 86.1818-12 is amended by revising paragraphs (c)(2)(i)(A) 
through (C) and (c)(3)(i)(A), (B), and (D), to read as follows:


Sec.  86.1818-12  Greenhouse gas emission standards for light-duty 
vehicles, light-duty trucks, and medium-duty passenger vehicles.

* * * * *
    (c) * * *
    (2) * * *
    (i) * * *
    (A) For passenger automobiles with a footprint of less than or 
equal to 41 square feet, the gram/mile CO2 target value 
shall be selected for the appropriate model year from the following 
table:

------------------------------------------------------------------------
                                                            CO2 target
                       Model year                         value  (grams/
                                                               mile)
------------------------------------------------------------------------
2012....................................................           244.0
2013....................................................           237.0
2014....................................................           228.0
2015....................................................           217.0
2016....................................................           206.0
2017....................................................           195.0
2018....................................................           185.0
2019....................................................           175.0
2020....................................................           166.0
2021....................................................           161.8
2022....................................................           159.0
2023....................................................           156.4
2024....................................................           153.7
2025....................................................           151.2
2026 and later..........................................           148.6
------------------------------------------------------------------------

    (B) For passenger automobiles with a footprint of greater than 56 
square feet, the gram/mile CO2 target value shall be 
selected for the appropriate model year from the following table:

------------------------------------------------------------------------
                                                            CO2 target
                       Model year                          value (grams/
                                                               mile)
------------------------------------------------------------------------
2012....................................................           315.0
2013....................................................           307.0
2014....................................................           299.0
2015....................................................           288.0
2016....................................................           277.0
2017....................................................           263.0
2018....................................................           250.0
2019....................................................           238.0
2020....................................................           226.0
2021....................................................           220.9
2022....................................................           217.3
2023....................................................           213.7
2024....................................................           210.2
2025....................................................           206.8
2026 and later..........................................           203.4
------------------------------------------------------------------------

    (C) For passenger automobiles with a footprint that is greater than 
41 square feet and less than or equal to 56 square feet, the gram/mile 
CO2 target value shall be calculated using the following 
equation and rounded to the nearest 0.1 grams/mile, except that for any 
vehicle footprint the maximum CO2 target value shall be the 
value specified for the same model year in paragraph (c)(2)(i)(B) of 
this section:
Target CO2 = [a x f] + b

Where: f is the vehicle footprint, as defined in Sec.  86.1803; and 
a and b are selected from the following table for the appropriate 
model year:

------------------------------------------------------------------------
               Model year                        a               b
------------------------------------------------------------------------
2012....................................            4.72            50.5
2013....................................            4.72            43.3
2014....................................            4.72            34.8
2015....................................            4.72            23.4
2016....................................            4.72            12.7
2017....................................            4.53             8.9
2018....................................            4.35             6.5
2019....................................            4.17             4.2
2020....................................            4.01             1.9
2021....................................            3.94             0.2
2022....................................            3.88            -0.1
2023....................................            3.82            -0.4
2024....................................            3.77            -0.6
2025....................................            3.71            -0.9
2026 and later..........................            3.65            -1.2
------------------------------------------------------------------------

* * * * *
    (3) * * *
    (i) * * *
    (A) For light trucks with a footprint of less than or equal to 41 
square feet, the gram/mile CO2 target value shall be 
selected for the appropriate model year from the following table:

[[Page 25269]]



------------------------------------------------------------------------
                                                            CO2 target
                       Model year                          value (grams/
                                                               mile)
------------------------------------------------------------------------
2012....................................................           294.0
2013....................................................           284.0
2014....................................................           275.0
2015....................................................           261.0
2016....................................................           247.0
2017....................................................           238.0
2018....................................................           227.0
2019....................................................           220.0
2020....................................................           212.0
2021....................................................           206.5
2022....................................................           203.0
2023....................................................           199.6
2024....................................................           196.2
2025....................................................           193.2
2026 and later..........................................           189.9
------------------------------------------------------------------------

    (B) For light trucks with a footprint that is greater than 41 
square feet and less than or equal to the maximum footprint value 
specified in the table below for each model year, the gram/mile 
CO2 target value shall be calculated using the following 
equation and rounded to the nearest 0.1 grams/mile, except that for any 
vehicle footprint the maximum CO2 target value shall be the 
value specified for the same model year in paragraph (c)(3)(i)(D) of 
this section:

Target CO2 = (a x f) + b

Where:
    f is the footprint, as defined in Sec.  86.1803; and a and b are 
selected from the following table for the appropriate model year:

----------------------------------------------------------------------------------------------------------------
                                                                      Maximum
                           Model year                                footprint           a               b
----------------------------------------------------------------------------------------------------------------
2012............................................................            66.0            4.04           128.6
2013............................................................            66.0            4.04           118.7
2014............................................................            66.0            4.04           109.4
2015............................................................            66.0            4.04            95.1
2016............................................................            66.0            4.04            81.1
2017............................................................            50.7            4.87            38.3
2018............................................................            60.2            4.76            31.6
2019............................................................            66.4            4.68            27.7
2020............................................................            68.3            4.57            24.6
2021............................................................            68.3            4.51            21.5
2022............................................................            68.3            4.44            20.6
2023............................................................            68.3            4.37            20.2
2024............................................................            68.3            4.31            19.6
2025............................................................            68.3            4.23            19.6
2026 and later..................................................            68.3            4.17            19.0
----------------------------------------------------------------------------------------------------------------

* * * * *
    (D) For light trucks with a footprint greater than the minimum 
value specified in the table below for each model year, the gram/mile 
CO2 target value shall be selected for the appropriate model 
year from the following table:

------------------------------------------------------------------------
                                                            CO2 target
               Model year                     Minimum      value (grams/
                                             footprint         mile)
------------------------------------------------------------------------
2012....................................            66.0           395.0
2013....................................            66.0           385.0
2014....................................            66.0           376.0
2015....................................            66.0           362.0
2016....................................            66.0           348.0
2017....................................            66.0           347.0
2018....................................            66.0           342.0
2019....................................            66.4           339.0
2020....................................            68.3           337.0
2021....................................            68.3           329.4
2022....................................            68.3           324.1
2023....................................            68.3           318.9
2024....................................            68.3           313.7
2025....................................            68.3           308.7
2026 and later..........................            68.3           303.7
------------------------------------------------------------------------

* * * * *

0
3. Section 86.1866-12 is amended by revising paragraph (a)(2), removing 
paragraph (a)(3), and revising (b) introductory text, (b)(1), and 
(b)(2)(i) to read as follows:


Sec.  86.1866-12  CO2 credits for advanced technology vehicles.

* * * * *
    (a) * * *
    (2) Model years 2017 through 2026: For electric vehicles, plug-in 
hybrid electric vehicles, and fuel cell vehicles produced for U.S. 
sale, where ``U.S.'' means the states and territories of the United 
States, in the 2017 through 2026 model years, such use of zero (0) 
grams/mile CO2 is unrestricted.
    (b) For electric vehicles, plug-in hybrid electric vehicles, fuel 
cell vehicles, dedicated natural gas vehicles, and dual-fuel natural 
gas vehicles as those terms are defined in Sec.  86.1803-01, that are 
certified and produced for U.S. sale in the specified model years and 
that meet the additional specifications in this section, the 
manufacturer may use the production multipliers in this paragraph (b) 
when determining additional credits for advanced technology vehicles. 
Full size pickup trucks eligible for and using a production multiplier 
are not eligible for the performance-based credits described in Sec.  
86.1870-12(b).

[[Page 25270]]

    (1) The production multipliers, by model year, for model year 2017 
through 2021 electric vehicles and fuel cell vehicles are as follows:

------------------------------------------------------------------------
                                                            Production
                       Model year                           multiplier
------------------------------------------------------------------------
2017....................................................             2.0
2018....................................................             2.0
2019....................................................             2.0
2020....................................................            1.75
2021....................................................             1.5
------------------------------------------------------------------------

    (2)(i) The production multipliers, by model year, for model year 
2017 through 2021 plug-in hybrid electric vehicles and model year 2017 
through 2026 dedicated natural gas vehicles and dual-fuel natural gas 
vehicles are as follows:

------------------------------------------------------------------------
                                                            Production
                       Model year                           multiplier
------------------------------------------------------------------------
2017....................................................             1.6
2018....................................................             1.6
2019....................................................             1.6
2020....................................................            1.45
2021....................................................             1.3
2022-2026 (dedicated and dual fuel natural gas vehicles              2.0
 only)..................................................
------------------------------------------------------------------------

* * * * *

0
4. Section 86.1868-12 is amended by adding an entry to the end of the 
table in paragraph (a)(2) and by adding paragraph (h)(7) to read as 
follows:


Sec.  86.1868-12  CO2 credits for improving the efficiency of air 
conditioning systems.

* * * * *
    (a) * * *
    (2) * * *

------------------------------------------------------------------------
                                             Passenger
       Air conditioning technology          automobiles    Light  trucks
                                              (g/mi)          (g/mi)
------------------------------------------------------------------------
 
                              * * * * * * *
Advanced technology air conditioning                 1.1             1.1
 compressor with improved efficiency
 relative to fixed-displacement
 compressors achieved through the
 addition of a variable crankcase
 suction valve..........................
------------------------------------------------------------------------

* * * * *
    (h) * * *
    (7) Advanced technology air conditioning compressor means an air 
conditioning compressor with improved efficiency relative to fixed-
displacement compressors. Efficiency gains are derived from improved 
internal valve systems that optimize the internal refrigerant flow 
across the range of compressor operator conditions through the addition 
of a variable crankcase suction valve.

0
5. Section 86.1869-12 is amended by revising paragraph (a), by adding 
paragraphs (b)(1)(ix), (b)(1)(x), (b)(4)(xiii) and (b)(4)(xiv), and by 
revising paragraph (d)(2) to read as follows:


Sec.  86.1869-12  CO2 credits for off-cycle CO2 reducing technologies.

* * * * *
    (a) Manufacturers may generate credits for CO2-reducing 
technologies where the CO2 reduction benefit of the 
technology is not adequately captured on the Federal Test Procedure 
and/or the Highway Fuel Economy Test such that the technology would not 
be otherwise installed for purposes of reducing emissions (directly or 
indirectly) over those test cycles for compliance with the GHG 
standards. These technologies must have a measurable, demonstrable, and 
verifiable real-world CO2 reduction that occurs outside the 
conditions of the Federal Test Procedure and the Highway Fuel Economy 
Test. These optional credits are referred to as ``off-cycle'' credits. 
The technologies must not be integral or inherent to the basic vehicle 
design, such as engine, transmission, mass reduction, passive 
aerodynamic design, and tire technologies. Technologies installed for 
non-off-cycle emissions related reasons are also not eligible as they 
would be considered part of the baseline vehicle design. The technology 
must not be inherent to the design of occupant comfort and 
entertainment features except for technologies related to reducing 
passenger air conditioning demand and improving air conditioning system 
efficiency. Notwithstanding the provisions of this paragraph (a), off-
cycle menu technologies included in paragraph (b) of this section 
remain eligible for credits. Off-cycle technologies used to generate 
emission credits are considered emission-related components subject to 
applicable requirements and must be demonstrated to be effective for 
the full useful life of the vehicle. Unless the manufacturer 
demonstrates that the technology is not subject to in-use 
deterioration, the manufacturer must account for the deterioration in 
their analysis. Durability evaluations of off-cycle technologies may 
occur at any time throughout a model year, provided that the results 
can be factored into the data provided in the model year report. Off-
cycle credits may not be approved for crash-avoidance technologies, 
safety critical systems or systems affecting safety-critical functions, 
or technologies designed for the purpose of reducing the frequency of 
vehicle crashes. Off-cycle credits may not be earned for technologies 
installed on a motor vehicle to attain compliance with any vehicle 
safety standard or any regulation set forth in Title 49 of the Code of 
Federal Regulations. The manufacturer must use one of the three options 
specified in this section to determine the CO2 gram per mile 
credit applicable to an off-cycle technology. Note that the option 
provided in paragraph (b) of this section applies only to the 2014 and 
later model years. The manufacturer should notify EPA in their pre-
model year report of their intention to generate any credits under this 
section.
    (b) * * *
    (1) * * *
    (ix) High efficiency alternator. The credit for a high efficiency 
alternator for passenger automobiles and light trucks shall be 
calculated using the following equation, and rounded to the nearest 0.1 
grams/mile:
[GRAPHIC] [TIFF OMITTED] TR30AP20.761


[[Page 25271]]


Where:

VDAHEA is the ratio of the alternator output power to the 
power supplied to the alternator, as measured using the Verband der 
Automobilindustrie (VDA) efficiency measurement methodology and 
expressed as a whole number percent from 68 to 100.
* * * * *
    (4) * * *
    (xiii) High efficiency alternator means an alternator where the 
ratio of the alternator output power to the power supplied to the 
alternator is greater than 67 percent, as measured using the Verband 
der Automobilindustrie (VDA) efficiency measurement methodology.
* * * * *
    (d) * * *
    (2) Notice and opportunity for public comment. (i) The 
Administrator will publish a notice of availability in the Federal 
Register notifying the public of a manufacturer's proposed alternative 
off-cycle credit calculation methodology. The notice will include 
details regarding the proposed methodology but will not include any 
Confidential Business Information. The notice will include instructions 
on how to comment on the methodology. The Administrator will take 
public comments into consideration in the final determination and will 
notify the public of the final determination. Credits may not be 
accrued using an approved methodology until the first model year for 
which the Administrator has issued a final approval.
    (ii) The Administrator may waive these notice and comment 
requirements for technologies for which EPA has previously approved a 
methodology for determining credits. To qualify for this waiver, the 
new application must be substantially identical in form, content, and 
methodology to the application for a previously approved methodology, 
and must include the following:
    (A) A cite to the appropriate previously approved methodology, 
including the appropriate Federal Register Notice and any subsequent 
EPA documentation of the Administrator's decision;
    (B) All necessary manufacturer- and vehicle-specific test data, 
modeling, and credit calculations; and,
    (C) Any other vehicle- or technology-specific details required 
pursuant to the previously approved methodology to assess and support 
an appropriate credit value.
    (iii) A waiver of the notice and comment requirements does not 
imply a determination that a specific credit value for a given 
technology is appropriate, and nor does it imply a waiver from the 
requirements in paragraphs (d)(1) and (e) of this section.
    (iv) The Administrator retains the option to require a notice and 
opportunity for public comment in cases where a new application 
deviates in significant respects from a previously approved methodology 
or raises novel substantive issues.
* * * * *

0
6. Section 86.1870-12 is amended by revising paragraphs (a)(2) and 
(b)(2) to read as follows:


Sec.  86.1870-12  CO2 credits for qualifying full-size light pickup 
trucks.

* * * * *
    (a) * * *
    (2) Full size pickup trucks that are strong hybrid electric 
vehicles and that are produced in the 2017 through 2021 model years are 
eligible for a credit of 20 grams/mile. To receive this credit in a 
model year, the manufacturer must produce a quantity of strong hybrid 
electric full size pickup trucks such that the proportion of production 
of such vehicles, when compared to the manufacturer's total production 
of full size pickup trucks, is not less than 10 percent in that model 
year.
* * * * *
    (b) * * *
    (2) Full size pickup trucks that are produced in the 2017 through 
2021 model years and that achieve carbon-related exhaust emissions less 
than or equal to the applicable target value determined in Sec.  
86.1818-12(c)(3) multiplied by 0.80 (rounded to the nearest gram/mile) 
in a model year are eligible for a credit of 20 grams/mile. A pickup 
truck that qualifies for this credit in a model year may claim this 
credit for a maximum of four subsequent model years (a total of five 
consecutive model years) if the carbon-related exhaust emissions of 
that pickup truck do not increase relative to the emissions in the 
model year in which the pickup truck first qualified for the credit. 
This credit may not be claimed in any model year after 2021. To qualify 
for this credit in a model year, the manufacturer must produce a 
quantity of full size pickup trucks that meet the emission requirements 
of this paragraph (b)(2) such that the proportion of production of such 
vehicles, when compared to the manufacturer's total production of full 
size pickup trucks, is not less than 10 percent in that model year.
* * * * *

PART 600--FUEL ECONOMY AND GREENHOUSE GAS EXHAUST EMISSIONS OF 
MOTOR VEHICLES

0
7. The authority citation for part 600 continues to read as follows:

    Authority:  49 U.S.C. 32901--23919q, Pub. L. 109-58.

0
8. Section 600.113-12 is amended by revising paragraphs (n) 
introductory text, (n)(1), and (n)(3) to read as follows:


Sec.  600.113-12  Fuel economy, CO2 emissions, and carbon-related 
exhaust emission calculations for FTP, HFET, US06, SC03 and cold 
temperature FTP tests.


* * * * *
    (n) Manufacturers shall determine CO2 emissions and 
carbon-related exhaust emissions for electric vehicles, fuel cell 
vehicles, and plug-in hybrid electric vehicles according to the 
provisions of this paragraph (n). Subject to the limitations on the 
number of vehicles produced and delivered for sale as described in 
Sec.  86.1866 of this chapter, the manufacturer may be allowed to use a 
value of 0 grams/mile to represent the emissions of fuel cell vehicles 
and the proportion of electric operation of a electric vehicles and 
plug-in hybrid electric vehicles that is derived from electricity that 
is generated from sources that are not onboard the vehicle, as 
described in paragraphs (n)(1) through (3) of this section. For 
purposes of labeling under this part, the CO2 emissions for 
electric vehicles shall be 0 grams per mile. Similarly, for purposes of 
labeling under this part, the CO2 emissions for plug-in 
hybrid electric vehicles shall be 0 grams per mile for the proportion 
of electric operation that is derived from electricity that is 
generated from sources that are not onboard the vehicle. For all 2027 
and later model year electric vehicles, fuel cell vehicles, and plug-in 
hybrid electric vehicles, the provisions of this paragraph (n) shall be 
used to determine the non-zero value for CREE for purposes of meeting 
the greenhouse gas emission standards described in Sec.  86.1818 of 
this chapter.
    (1) For electric vehicles, but not including fuel cell vehicles, 
the carbon-related exhaust emissions in grams per mile is to be 
calculated using the following equation and rounded to the nearest one 
gram per mile:

CREE = CREEUP - CREEGAS

Where:

CREE means the carbon-related exhaust emission value as defined in 
Sec.  600.002, which may be set equal to zero for eligible 2012 
through 2026 model year electric vehicles as described in Sec.  
86.1866-12(a) of this chapter.

[[Page 25272]]

[GRAPHIC] [TIFF OMITTED] TR30AP20.762

[GRAPHIC] [TIFF OMITTED] TR30AP20.763

Where:

EC = The vehicle energy consumption in watt-hours per mile, for 
combined FTP/HFET operation, determined according to procedures 
established by the Administrator under Sec.  600.116-12.
GRIDLOSS = 0.935 (to account for grid transmission losses).
AVGUSUP = 0.534 (the nationwide average electricity greenhouse gas 
emission rate at the powerplant, in grams per watt-hour).
2478 is the estimated grams of upstream greenhouse gas emissions per 
gallon of gasoline.
8887 is the estimated grams of CO2 per gallon of 
gasoline.
TargetCO2 = The CO2 Target Value for the fuel 
cell or electric vehicle determined according to Sec.  86.1818 of 
this chapter for the appropriate model year.
* * * * *
    (3) For 2012 and later model year fuel cell vehicles, the carbon-
related exhaust emissions in grams per mile shall be calculated using 
the method specified in paragraph (n)(1) of this section, except that 
CREEUP shall be determined according to procedures established by the 
Administrator under Sec.  600.111-08(f). As described in Sec.  86.1866 
of this chapter, the value of CREE may be set equal to zero for 2012 
through 2026 model year fuel cell vehicles.
* * * * *

0
9. Section 600.510-12 is amended by revising paragraphs (c)(2)(vi) 
introductory text, adding paragraph (c)(2)(vii) introductory text, 
revising the introductory text of paragraphs (c)(2)(vii)(B), (j)(2)(v), 
(vii)(A) and (vii)(B) to read as follows:


Sec.  600.510-12  Calculation of average fuel economy and average 
carbon-related exhaust emissions.

* * * * *
    (c) * * *
    (2) * * *
    (vi) For natural gas dual fuel model types, for model years 1993 
through 2016, and optionally for 2021 and later model years, the 
harmonic average of the following two terms; the result rounded to the 
nearest 0.1 mpg:
* * * * *
    (vii) This paragraph (c)(2)(vii) applies to model year 2017 through 
2020 natural gas dual fuel model types. Model year 2021 and later 
natural gas dual fuel model types may use the provisions of paragraph 
(c)(2)(vi) of this section or this paragraph (c)(2)(vii).
* * * * *
    (B) Model year 2017 through 2020 natural gas dual fuel model types 
must meet the following criteria to qualify for use of a Utility Factor 
greater than 0.5:
* * * * *
    (j) * * *
    (2) * * *
    (v) For natural gas dual fuel model types, for model years 2012 
through 2015, and optionally for 2021 and later model years, the 
arithmetic average of the following two terms; the result rounded to 
the nearest gram per mile:
* * * * *
    (vii)(A) This paragraph (j)(2)(vii) applies to model year 2016 
through 2020 natural gas dual fuel model types. Model year 2021 and 
later natural gas dual fuel model types may use the provisions of 
paragraph (j)(2)(v) of this section or this paragraph (j)(2)(vii).
* * * * *
    (B) Model year 2016 through 2020 natural gas dual fuel model types 
must meet the following criteria to qualify for use of a Utility Factor 
greater than 0.5:
* * * * *

National Highway Transportation Administration

Chapter V

    For the reasons discussed in the preamble, the National Highway 
Traffic Safety Administration amends 49 CFR chapter V as follows:

PART 523--VEHICLE CLASSIFICATION

0
10. The authority citation for part 523 continues to read as follows:

    Authority: 49 U.S.C 32901; delegation of authority at 49 CFR 
1.95.

0
11. Amend Sec.  523.2 by revising the definitions of ``Curb weight'' 
and ``Full-size pickup truck'' to read as follows:


Sec.  523.2  Definitions.

* * * * *
    Curb weight has the meaning given in 40 CFR 86.1803-01.
* * * * *
    Full-size pickup truck means a light truck or medium duty passenger 
vehicle that meets the specifications in 40 CFR 86.1803-01.
* * * * *

PART 531--PASSENGER AUTOMOBILE AVERAGE FUEL ECONOMY STANDARDS

0
12. The authority citation for part 531 is revised to read as follows:

    Authority:  49 U.S.C. 32902; delegation of authority at 49 CFR 
1.95.

0
13. Amend Sec.  531.5 by revising the introductory text of paragraph 
(c), Table III to paragraph (c), and paragraph (d), removing paragraph 
(e), and redesignating paragraph (f) as paragraph (e) to read as 
follows:


Sec.  531.5  Fuel economy standards.

* * * * *
    (c) For model years 2012-2026, a manufacturer's passenger 
automobile fleet shall comply with the fleet average fuel economy level 
calculated for that model year according to this Figure 2 and the 
appropriate values in this Table III.
* * * * *

             Table III--Parameters for the Passenger Automobile Fuel Economy Targets, MYs 2012-2026
----------------------------------------------------------------------------------------------------------------
                                                                            Parameters
                                                 ---------------------------------------------------------------
                   Model year                                                       c (gal/mi/
                                                      a (mpg)         b (mpg)         ft\2\)        d (gal/mi)
----------------------------------------------------------------------------------------------------------------
2012............................................           35.95           27.95       0.0005308        0.006057
2013............................................           36.80           28.46       0.0005308        0.005410
2014............................................           37.75           29.03       0.0005308        0.004725
2015............................................           39.24           29.90       0.0005308        0.003719
2016............................................           41.09           30.96       0.0005308        0.002573

[[Page 25273]]

 
2017............................................           43.61           32.65       0.0005131        0.001896
2018............................................           45.21           33.84       0.0004954        0.001811
2019............................................           46.87           35.07       0.0004783        0.001729
2020............................................           48.74           36.47       0.0004603        0.001643
2021............................................           49.48           37.02        0.000453         0.00162
2022............................................           50.24           37.59        0.000447         0.00159
2023............................................           51.00           38.16        0.000440         0.00157
2024............................................           51.78           38.74        0.000433         0.00155
2025............................................           52.57           39.33        0.000427         0.00152
2026............................................           53.37           39.93        0.000420         0.00150
----------------------------------------------------------------------------------------------------------------

    (d) In addition to the requirements of paragraphs (b) and (c) of 
this section, each manufacturer shall also meet the minimum fleet 
standard for domestically manufactured passenger automobiles expressed 
in Table IV:

 Table IV--Minimum Fuel Economy Standards for Domestically Manufactured
                  Passenger Automobiles, MYs 2011-2026
------------------------------------------------------------------------
                                                              Minimum
                       Model year                            standard
------------------------------------------------------------------------
2011....................................................            27.8
2012....................................................            30.7
2013....................................................            31.4
2014....................................................            32.1
2015....................................................            33.3
2016....................................................            34.7
2017....................................................            36.7
2018....................................................            38.0
2019....................................................            39.4
2020....................................................            40.9
2021....................................................            39.9
2022....................................................            40.6
2023....................................................            41.1
2024....................................................            41.8
2025....................................................            42.4
2026....................................................            43.1
------------------------------------------------------------------------

* * * * *


0
14. Amend Sec.  531.6 by revising paragraphs (a) and (b) to read as 
follows:


Sec.  531.6  Measurement and calculation procedures.

    (a) The fleet average fuel economy performance of all passenger 
automobiles that are manufactured by a manufacturer in a model year 
shall be determined in accordance with procedures established by the 
Administrator of the Environmental Protection Agency under 49 U.S.C. 
32904 and set forth in 40 CFR part 600. For model years 2017 to 2026, a 
manufacturer is eligible to increase the fuel economy performance of 
passenger cars in accordance with procedures established by the EPA set 
forth in 40 CFR part 600, subpart F, including any adjustments to fuel 
economy the EPA allows, such as for fuel consumption improvements 
related to air conditioning efficiency and off-cycle technologies.
    (1) A manufacturer that seeks to increase its fleet average fuel 
economy performance through the use of technologies that improve the 
efficiency of air conditioning systems must follow the requirements in 
40 CFR 86.1868-12. Fuel consumption improvement values resulting from 
the use of those air conditioning systems must be determined in 
accordance with 40 CFR 600.510-12(c)(3)(i).
    (2) A manufacturer that seeks to increase its fleet average fuel 
economy performance through the use of off-cycle technologies must 
follow the requirements in 40 CFR 86.1869-12. A manufacturer is 
eligible to gain fuel consumption improvements for predefined off-cycle 
technologies in accordance with 40 CFR 86.1869-12(b) or for 
technologies tested using the EPA's 5-cycle methodology in accordance 
with 40 CFR 86.1869-12(c). The fuel consumption improvement is 
determined in accordance with 40 CFR 600.510-12(c)(3)(ii).
    (b) A manufacturer is eligible to increase its fuel economy 
performance through use of an off-cycle technology requiring an 
application request made to the EPA in accordance with 40 CFR 86.1869-
12(d). The request must be approved by the EPA in consultation with 
NHTSA. To expedite NHTSA's consultation with the EPA, a manufacturer 
shall concurrently submit its application to NHTSA if the manufacturer 
is seeking off-cycle fuel economy improvement values under the CAFE 
program for those technologies. For off-cycle technologies that are 
covered under 40 CFR 86.1869-12(d), NHTSA will consult with the EPA 
regarding NHTSA's evaluation of the specific off-cycle technology to 
ensure its impact on fuel economy and the suitability of using the off-
cycle technology to adjust the fuel economy performance. NHTSA will 
provide its views on the suitability of the technology for that purpose 
to the EPA. NHTSA's evaluation and review will consider:
    (1) Whether the technology has a direct impact upon improving fuel 
economy performance;
    (2) Whether the technology is related to crash-avoidance 
technologies, safety critical systems or systems affecting safety-
critical functions, or technologies designed for the purpose of 
reducing the frequency of vehicle crashes;
    (3) Information from any assessments conducted by the EPA related 
to the application, the technology and/or related technologies; and
    (4) Any other relevant factors.

PART 533--LIGHT TRUCK FUEL ECONOMY STANDARDS

0
15. The authority citation for part 533 is revised to read as follows:

    Authority:  49 U.S.C. 32902; delegation of authority at 49 CFR 
1.95.

0
16. In Sec.  533.5, amend paragraph (a) by revising Table VII and 
removing paragraph (k) to read as follows:


Sec.  533.5  Requirements.

    (a) * * *

[[Page 25274]]



                                    Table VII--Parameters for the Light Truck Fuel Economy Targets for MYs 2017-2026
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                Parameters
                                                 -------------------------------------------------------------------------------------------------------
                   Model year                                                c (gal/mi/                                          g (gal/mi/
                                                    a (mpg)      b (mpg)       ft\2\)     d (gal/mi)    e (mpg)      f (mpg)       ft\2\)     h (gal/mi)
--------------------------------------------------------------------------------------------------------------------------------------------------------
2017............................................        36.26        25.09    0.0005484     0.005097        35.10        25.09    0.0004546     0.009851
2018............................................        37.36        25.20    0.0005358     0.004797        35.31        25.20    0.0004546     0.009682
2019............................................        38.16        25.25    0.0005265     0.004623        35.41        25.25    0.0004546     0.009603
2020............................................        39.11        25.25    0.0005140     0.004494        35.41        25.25    0.0004546     0.009603
2021............................................        39.71        25.63     0.000506      0.00443           NA           NA           NA           NA
2022............................................        40.31        26.02     0.000499      0.00436           NA           NA           NA           NA
2023............................................        40.93        26.42     0.000491      0.00429           NA           NA           NA           NA
2024............................................        41.55        26.82     0.000484      0.00423           NA           NA           NA           NA
2025............................................        42.18        27.23     0.000477      0.00417           NA           NA           NA           NA
2026............................................        42.82        27.64     0.000469      0.00410           NA           NA           NA           NA
--------------------------------------------------------------------------------------------------------------------------------------------------------

* * * * *


0
17. Amend Sec.  533.6 by revising paragraphs (b) and (c) to read as 
follows:


Sec.  533.6  Measurement and calculation procedures.

* * * * *
    (b) The fleet average fuel economy performance of all light trucks 
that are manufactured by a manufacturer in a model year shall be 
determined in accordance with procedures established by the 
Administrator of the Environmental Protection Agency under 49 U.S.C. 
32904 and set forth in 40 CFR part 600. For model years 2017 to 2026, a 
manufacturer is eligible to increase the fuel economy performance of 
light trucks in accordance with procedures established by the EPA set 
forth in 40 CFR part 600, subpart F, including any adjustments to fuel 
economy the EPA allows, such as for fuel consumption improvements 
related to air conditioning efficiency, off-cycle technologies, and 
hybridization and other performance-based technologies for full-size 
pickup trucks that meet the requirements specified in 40 CFR 86.1803.
    (1) A manufacturer that seeks to increase its fleet average fuel 
economy performance through the use of technologies that improve the 
efficiency of air conditioning systems must follow the requirements in 
40 CFR 86.1868-12. Fuel consumption improvement values resulting from 
the use of those air conditioning systems must be determined in 
accordance with 40 CFR 600.510-12(c)(3)(i).
    (2) A manufacturer that seeks to increase its fleet average fuel 
economy performance through the use of off-cycle technologies must 
follow the requirements in 40 CFR 86.1869-12. A manufacturer is 
eligible to gain fuel consumption improvements for predefined off-cycle 
technologies in accordance with 40 CFR 86.1869-12(b) or for 
technologies tested using the EPA's 5-cycle methodology in accordance 
with 40 CFR 86.1869-12(c). The fuel consumption improvement is 
determined in accordance with 40 CFR 600.510-12(c)(3)(ii).
    (3) The eligibility of a manufacturer to increase its fuel economy 
using hybridized and other performance-based technologies for full-size 
pickup trucks must follow 40 CFR 86.1870-12 and the fuel consumption 
improvement of these full-size pickup truck technologies must be 
determined in accordance with 40 CFR 600.510-12(c)(3)(iii).
    (c) A manufacturer is eligible to increase its fuel economy 
performance through use of an off-cycle technology requiring an 
application request made to the EPA in accordance with 40 CFR 86.1869-
12(d). The request must be approved by the EPA in consultation with 
NHTSA. To expedite NHTSA's consultation with the EPA, a manufacturer 
shall concurrently submit its application to NHTSA if the manufacturer 
is seeking off-cycle fuel economy improvement values under the CAFE 
program for those technologies. For off-cycle technologies that are 
covered under 40 CFR 86.1869-12(d), NHTSA will consult with the EPA 
regarding NHTSA's evaluation of the specific off-cycle technology to 
ensure its impact on fuel economy and the suitability of using the off-
cycle technology to adjust the fuel economy performance. NHTSA will 
provide its views on the suitability of the technology for that purpose 
to the EPA. NHTSA's evaluation and review will consider:
    (1) Whether the technology has a direct impact upon improving fuel 
economy performance;
    (2) Whether the technology is related to crash-avoidance 
technologies, safety critical systems or systems affecting safety-
critical functions, or technologies designed for the purpose of 
reducing the frequency of vehicle crashes;
    (3) Information from any assessments conducted by the EPA related 
to the application, the technology and/or related technologies; and
    (4) Any other relevant factors.

PART 535--MEDIUM- AND HEAVY-DUTY VEHICLE FUEL EFFICIENCY PROGRAM

0
18. The authority citation for part 535 continues to read as follows:

    Authority:  49 U.S.C. 32902 and 30101; delegation of authority 
at 49 CFR 1.95.

0
19. Amend Sec.  535.6 by revising paragraphs (a)(4)(ii) and (d)(5)(ii) 
to read as follows:


Sec.  535.6  Measurement and calculation procedures.

* * * * *
    (a) * * *
    (4) * * *
    (ii) Calculate the equivalent fuel consumption test group results 
as follows for spark-ignition vehicles and alternative fuel spark-
ignition vehicles. CO2 emissions test group result (grams 
per mile)/((8,887 grams per gallon of gasoline fuel) x 
(10-2)) = Fuel consumption test group result (gallons per 
100 mile).
* * * * *
    (d) * * *
    (5) * * *
    (ii) Calculate equivalent fuel consumption FCL values for spark-
ignition engines and alternative fuel spark-ignition engines. 
CO2 FCL value (grams per hp-hr)/((8,887 grams per gallon of 
gasoline fuel) x (10-2)) = Fuel consumption FCL value 
(gallons per 100 hp-hr).
* * * * *

0
20. Amend Sec.  535.7 by revising the equations in paragraphs (b)(1), 
(c)(1), (d)(1), (e)(2), and (f)(2)(iii)(E) to read as follows:

[[Page 25275]]

Sec.  535.7  Averaging, banking, and trading (ABT) credit program.

* * * * *
    (b) * * *
    (1) * * *

Total MY Fleet FCC (gallons) = (Std - Act) x (Volume) x (UL) x 
(10-2)
Where:

Std = Fleet average fuel consumption standard (gal/100 mile).
Act = Fleet average actual fuel consumption value (gal/100 mile).
Volume = the total U.S.-directed production of vehicles in the 
regulatory subcategory.
UL = the useful life for the regulatory subcategory. The useful life 
value for heavy-pickup trucks and vans manufactured for model years 
2013 through 2020 is equal to the 120,000 miles. The useful life for 
model years 2021 and later is equal to 150,000 miles.
* * * * *
    (c) * * *
    (1) * * *

Vehicle Family FCC (gallons) = (Std - FEL) x (Payload) x (Volume) x 
(UL) x (10-\3\)

Where:

Std = the standard for the respective vehicle family regulatory 
subcategory (gal/1000 ton-mile).
FEL = family emissions limit for the vehicle family (gal/1000 ton-
mile).
Payload = the prescribed payload in tons for each regulatory 
subcategory as shown in the following table:

------------------------------------------------------------------------
                                                              Payload
                 Regulatory subcategory                       (tons)
------------------------------------------------------------------------
Vocational LHD Vehicles.................................            2.85
Vocational MHD Vehicles.................................            5.60
Vocational HHD Vehicles.................................             7.5
MDH Tractors............................................           12.50
HHD Tractors, other than heavy-haul Tractors............           19.00
Heavy-haul Tractors.....................................           43.00
------------------------------------------------------------------------


Volume = the number of U.S.-directed production volume of vehicles 
in the corresponding vehicle family.
UL = the useful life for the regulatory subcategory (miles) as shown 
in the following table:

------------------------------------------------------------------------
          Regulatory subcategory                     UL (miles)
------------------------------------------------------------------------
LHD Vehicles..............................  110,000 (Phase 1).
                                            150,000 (Phase 2).
Vocational MHD Vehicles and tractors at or  185,000.
 below 33,000 pounds GVWR.
Vocation HHD Vehicles and tractors at or    435,000.
 above 33,000 pounds GVWR.
------------------------------------------------------------------------

* * * * *
    (d) * * *
    (1) * * *
Engine Family FCC (gallons) = (Std - FCL) x (CF) x (Volume) x (UL) x 
(10-2)

Where:
Std = the standard for the respective engine regulatory subcategory 
(gal/100 hp-hr).
FCL = family certification level for the engine family (gal/100 hp-
hr).
CF= a transient cycle conversion factor in hp-hr/mile which is the 
integrated total cycle horsepower-hour divided by the equivalent 
mileage of the applicable test cycle. For engines subject to spark-
ignition heavy-duty standards, the equivalent mileage is 6.3 miles. 
For engines subject to compression-ignition heavy-duty standards, 
the equivalent mileage is 6.5 miles.
Volume = the number of engines in the corresponding engine family.
UL = the useful life of the given engine family (miles) as shown in 
the following table:

------------------------------------------------------------------------
          Regulatory subcategory                     UL (miles)
------------------------------------------------------------------------
SI and CI LHD Engines.....................  120,000 (Phase 1).
                                            150,000 (Phase 2).
CI MHD Engines............................  185,000.
CI HHD Engines............................  435,000.
------------------------------------------------------------------------

* * * * *
    (e) * * *
    (2) * * *

Vehicle Family FCC (gallons) = (Std - FEL) x (Payload) x (Volume) x 
(UL) x (10-3)

Where:

Std = the standard for the respective vehicle family regulatory 
subcategory (gal/1000 ton-mile).
FEL = family emissions limit for the vehicle family (gal/1000 ton-
mile).
Payload = 10 tons for short box vans and 19 tons for other trailers.
Volume = the number of U.S.-directed production volume of vehicles 
in the corresponding vehicle family.
UL = the useful life for the regulatory subcategory. The useful life 
value for heavy-duty trailers is equal to 250,000 miles.
* * * * *
    (f) * * *
    (2) * * *
    (iii) * * *
    (E) * * *

Off-cycle FC credits = (CO2 Credit/CF) x Production x VLM

Where:

CO2 Credits = the credit value in grams per mile 
determined in 40 CFR 86.1869-12(c)(3), (d)(1), (d)(2) or (d)(3).
CF = conversion factor, which for spark-ignition engines is 8,887 
and for compression-ignition engines is 10,180.
Production = the total production volume for the applicable category 
of vehicles
VLM = vehicle lifetime miles, which for 2b-3 vehicles shall be 
150,000 for the Phase 2 program.
The term (CO2 Credit/CF) should be rounded to the nearest 
0.0001
* * * * *

PART 536--TRANSFER AND TRADING OF FUEL ECONOMY CREDITS

0
21. The authority citation for part 536 is revised to read as follows:

    Authority:  49 U.S.C. 32903; delegation of authority at 49 CFR 
1.95.


0
22. Amend Sec.  536.4 by revising paragraph (c) to read as follows:


Sec.  536.4  Credits.

* * * * *
    (c) Adjustment factor. When traded or transferred and used, fuel 
economy credits are adjusted to ensure fuel oil savings is preserved. 
For traded credits, the user (or buyer) must multiply the calculated 
adjustment factor by the number of shortfall credits it plans to offset 
in order to determine the number of equivalent credits to acquire from 
the earner (or seller). For transferred credits, the user of credits 
must multiply the calculated adjustment factor by the number of 
shortfall credits it plans to offset in order to determine the number 
of equivalent credits to transfer from the compliance category holding 
the available credits. The adjustment factor is calculated according to 
the following formula:
[GRAPHIC] [TIFF OMITTED] TR30AP20.764

Where:

A = Adjustment factor applied to traded and transferred credits. The 
quotient shall be rounded to 4 decimal places;
* * * * *

0
23. Amend Sec.  536.5 by revising paragraphs (c) and (d)(6) to read as 
follows:

[[Page 25276]]

Sec.  536.5  Trading infrastructure.

* * * * *
    (c) Automatic debits and credits of accounts.
    (1) To carry credits forward, backward, transfer credits, or trade 
credits into other credit accounts, a manufacturer or credit holder 
must submit a credit instruction to NHTSA. A credit instruction must 
detail and include:
    (i) The credit holder(s) involved in the transaction.
    (ii) The originating credits described by the amount of the 
credits, compliance category and the vintage of the credits.
    (iii) The recipient credit account(s) for banking or applying the 
originating credits described by the compliance category(ies), model 
year(s), and if applicable the adjusted credit amount(s) and adjustment 
factor(s).
    (iv) For trades, a contract authorizing the trade signed by the 
manufacturers or credit holders or by managers legally authorized to 
obligate the sale and purchase of the traded credits.
    (2) Upon receipt of a credit instruction from an existing credit 
holder, NHTSA verifies the presence of sufficient credits in the 
account(s) of the credit holder(s) involved as applicable and notifies 
the credit holder(s) that the credits will be debited from and/or 
credited to the accounts involved, as specified in the credit 
instruction. NHTSA determines if the credits can be debited or credited 
based upon the amount of available credits, accurate application of any 
adjustment factors and the credit requirements prescribed by this part 
that are applicable at the time the transaction is requested.
    (3) After notifying the credit holder(s), all accounts involved are 
either credited or debited, as appropriate, in line with the credit 
instruction. Traded credits identified by a specific compliance 
category are deposited into the recipient's account in that same 
compliance category and model year. If a recipient of credits as 
identified in a credit instruction is not a current account holder, 
NHTSA establishes the credit recipient's account, subject to the 
conditions described in Sec.  536.5(b), and adds the credits to the 
newly-opened account.
    (4) NHTSA will automatically delete unused credits from holders' 
accounts when those credits reach their expiry date.
    (5) Starting in model year 2021, manufacturers or credit holders 
issuing credit instructions or providing credit allocation plans as 
specified in Sec.  536.5(d), must use the NHTSA Credit Template 
fillable form (OMB Control No. 2127-0019, NHTSA Form 1475). The NHTSA 
Credit Template is available for download on NHTSA's website. If a 
credit instruction includes a trade, the NHTSA Credit Template must be 
signed by managers legally authorized to obligate the sale and/or 
purchase of the traded credits from both parties to the trade. The 
NHTSA Credit Template signed by both parties to the trade serves as an 
acknowledgement that the parties have agreed to trade credits, and does 
not dictate terms, conditions, or other business obligations of the 
parties. All parties trading credits must also provide NHTSA the price 
paid for the credits including a description of any other monetary or 
non-monetary terms affecting the price of the traded credits, such as 
any technology exchanged or shared for the credits, any other non-
monetary payment for the credits, or any other agreements related to 
the trade. Manufacturers must submit this information to NHTSA in a PDF 
document along with the Credit Template through the CAFE email, 
[email protected]. NHTSA reserves the right to request additional 
information from the parties regarding the terms of the trade.
    (6) NHTSA will consider claims that information submitted to the 
agency under this section is entitled to confidential treatment under 5 
U.S.C. 552(b) and under the provisions of part 512 of this chapter if 
the information is submitted in accordance with the procedures of that 
part.
* * * * *
    (d) * * *
    (6) Credit allocation plans received from a manufacturer will be 
reviewed and approved by NHTSA. Starting in model year 2021, use the 
NHTSA Credit Template (OMB Control No. 2127-0019, NHTSA Form 1475) to 
record the credit transactions requested in the credit allocation plan. 
The template is a fillable form that has an option for recording and 
calculating credit transactions for credit allocation plans. The 
template calculates the required adjustments to the credits. The credit 
allocation plan and the completed transaction template must be 
submitted to NHTSA. NHTSA will approve the credit allocation plan 
unless it finds that the proposed credits are unavailable or that it is 
unlikely that the plan will result in the manufacturer earning 
sufficient credits to offset the subject credit shortfall. If the plan 
is approved, NHTSA will revise the respective manufacturer's credit 
account accordingly. If the plan is rejected, NHTSA will notify the 
respective manufacturer and request a revised plan or payment of the 
appropriate fine.

PART 537--AUTOMOTIVE FUEL ECONOMY REPORTS

0
24. The authority citation for part 537 is revised to read as follows:

    Authority:  49 U.S.C. 32907; delegation of authority at 49 CFR 
1.95.

0
25. Amend Sec.  537.5 by redesignating paragraph (d) as paragraph (e) 
and adding a new paragraph (d) to read as follows:


Sec.  537.5  General requirements for reports.

* * * * *
    (d) Beginning with model year 2023, each manufacturer shall 
generate reports required by this part using the NHTSA CAFE Projections 
Reporting Template (OMB Control No. 2127-0019, NHTSA Form 1474). The 
template is a fillable form.
    (1) Select the option to identify the report as a pre-model year 
report, mid-model year report, or supplementary report as appropriate;
    (2) Complete all required information for the manufacturer and for 
all vehicles produced for the current model year required to comply 
with CAFE standards. Identify the manufacturer submitting the report, 
including the full name, title, and address of the official responsible 
for preparing the report and a point of contact to answer questions 
concerning the report.
    (3) Use the template to generate confidential and non-confidential 
reports for all the domestic and import passenger cars and light truck 
fleet produced by the manufacturer for the current model year. 
Manufacturers must submit a request for confidentiality in accordance 
with part 512 of this chapter to withhold projected production sales 
volume estimates from public disclosure. If the request is granted, 
NHTSA will withhold the projected production sales volume estimates 
from public disclose until all the vehicles produced by the 
manufacturer have been made available for sale (usually one year after 
the current model year).
    (4) Submit confidential reports and requests for confidentiality to 
NHTSA on CD-ROM in accordance with Part 537.12. Email copies of non-
confidential (i.e., redacted) reports to NHTSA's secure email address: 
[email protected]. Requests for confidentiality must be submitted in a PDF 
or MS Word format. Submit 2 copies of the CD-ROM to: Administrator, 
National Highway Traffic Administration, 1200 New Jersey Avenue SE, 
Washington, DC 20590, and submit emailed reports electronically to

[[Page 25277]]

the following secure email address: [email protected];
    (5) Confidentiality Requests. Manufacturers can withhold 
information on projected production sales volumes under 5 U.S.C. 
552(b)(4) and 15 U.S.C. 2005(d)(1). In accordance, the manufacturer 
must:
    (i) Show that the item is within the scope of sections 552(b)(4) 
and 2005(d)(1);
    (ii) Show that disclosure of the item would result in significant 
competitive damage;
    (iii) Specify the period during which the item must be withheld to 
avoid that damage; and
    (iv) Show that earlier disclosure would result in that damage.
* * * * *

0
26. Amend Sec.  537.6 by revising paragraphs (b) and (c) to read as 
follows:


Sec.  537.6  General content of reports.

* * * * *
    (b) Supplementary report. Except as provided in paragraph (c) of 
this section, each supplementary report for each model year must 
contain the information required by Sec.  537.7(a)(1) and (a)(2), as 
appropriate for the vehicle fleets produced by the manufacturer, in 
accordance with Sec.  537.8(b)(1), (2), (3), and (4) as appropriate.
    (c) Exceptions. The pre-model year report, mid-model year report, 
and supplementary report(s) submitted by an incomplete automobile 
manufacturer for any model year are not required to contain the 
information specified in Sec.  537.7 (c)(4) (xv) through (xviii) and 
(c)(5). The information provided by the incomplete automobile 
manufacturer under Sec.  537.7(c) shall be according to base level 
instead of model type or carline.

0
27. Amend Sec.  537.7 by revising paragraph (a) to read as follows:


Sec.  537.7  Pre-model year and mid-model year reports.

    (a)(1) Provide a report with the information required by paragraphs 
(b) and (c) of this section for each domestic and import passenger 
automobile fleet, as specified in part 531 of this chapter, for the 
current model year.
    (2) Provide a report with the information required by paragraphs 
(b) and (c) of this section for each light truck fleet, as specified in 
part 533 of this chapter, for the current model year.
    (3) For model year 2023 and later, provide the information required 
by paragraphs (a)(1) and (2) of this section for pre-model and mid-
model year reports in accordance with the NHTSA CAFE Projections 
Reporting Template (OMB Control No. 2127-0019, NHTSA Form 1474). The 
required reporting template can be downloaded from NHTSA's website.
* * * * *

0
28. Amend Sec.  537.7 by revising paragraphs (b)(3), (c)(1), (c)(3), 
(c)(7)(i), (c)(7)(ii), and (c)(7)(iii) to read as follows:


Sec.  537.7  Pre-model year and mid-model year reports.

* * * * *
    (b) * * *
    (3) State the projected required fuel economy for the 
manufacturer's passenger automobiles and light trucks determined in 
accordance with Sec. Sec.  531.5(c) and 533.5 of this chapter and based 
upon the projected sales figures provided under paragraph (c)(2) of 
this section. For each unique model type and footprint combination of 
the manufacturer's automobiles, provide the information specified in 
paragraph (b)(3)(i) and (ii) of this section in tabular form. List the 
model types in order of increasing average inertia weight from top to 
bottom down the left side of the table and list the information 
categories in the order specified in paragraphs (b)(3)(i) and (ii) of 
this section from left to right across the top of the table. Other 
formats, such as those accepted by the EPA, which contain all the 
information in a readily identifiable format are also acceptable. For 
model year 2023 and later, for each unique model type and footprint 
combination of the manufacturer's automobiles, provide the information 
specified in paragraph (b)(3)(i) and (ii) of this section in accordance 
with the CAFE Projections Reporting Template (OMB Control No. 2127-
0019, NHTSA Form 1474).
    (i) In the case of passenger automobiles:
    (A) Beginning model year 2013, base tire as defined in Sec.  523.2 
of this chapter,
    (B) Beginning model year 2013, front axle, rear axle, and average 
track width as defined in Sec.  CFR 523.2 of this chapter,
    (C) Beginning model year 2013, wheelbase as defined in Sec.  523.2 
of this chapter, and
    (D) Beginning model year 2013, footprint as defined in Sec.  523.2 
of this chapter.
    (E) The fuel economy target value for each unique model type and 
footprint entry listed in accordance with the equation provided in part 
531 of this chapter.
    (ii) In the case of light trucks:
    (A) Beginning model year 2013, base tire as defined in Sec.  523.2 
of this chapter,
    (B) Beginning model year 2013, front axle, rear axle, and average 
track width as defined in Sec.  523.2 of this chapter,
    (C) Beginning model year 2013, wheelbase as defined in Sec.  523.2 
of this chapter, and
    (D) Beginning model year 2013, footprint as defined in Sec.  523.2 
of this chapter.
    (E) The fuel economy target value for each unique model type and 
footprint entry listed in accordance with the equation provided in part 
533 of this chapter.
* * * * *
    (c) * * *
    (1) For each model type of the manufacturer's automobiles, provide 
the information specified in paragraph (c)(2) of this section in 
tabular form. List the model types in order of increasing average 
inertia weight from top to bottom down the left side of the table and 
list the information categories in the order specified in paragraph 
(c)(2) of this section from left to right across the top of the table. 
For model year 2023 and later, CAFE reports required by part 537 of 
this chapter, shall for each model type of the manufacturer's 
automobiles, provide the information in specified in paragraph (c)(2) 
of this section in accordance with the NHTSA CAFE Projections Reporting 
Template (OMB Control No. 2127-0019, NHTSA Form 1474) and list the 
model types in order of increasing average inertia weight from top to 
bottom.
* * * * *
    (3) (Pre-model year reports only through model year 2022.) For each 
vehicle configuration whose fuel economy was used to calculate the fuel 
economy values for a model type under paragraph (c)(2) of this section, 
provide the information specified in paragraph (c)(4) of this section 
in accordance with the NHTSA CAFE Projections Reporting Template (OMB 
Control No. 2127-0019, NHTSA Form 1474).
* * * * *
    (7) * * *
    (i) Provide a list of each air conditioning efficiency improvement 
technology utilized in your fleet(s) of vehicles for each model year. 
For each technology identify vehicles by make and model types that have 
the technology, which compliance category those vehicles belong to and 
the number of vehicles for each model equipped with the technology. For 
each compliance category (domestic passenger car, import passenger car, 
and light truck), report the air conditioning fuel consumption 
improvement value in gallons/mile in accordance with the equation 
specified in 40 CFR 600.510-12(c)(3)(i).
    (ii) Provide a list of off-cycle efficiency improvement 
technologies

[[Page 25278]]

utilized in your fleet(s) of vehicles for each model year that is 
pending or approved by the EPA. For each technology identify vehicles 
by make and model types that have the technology, which compliance 
category those vehicles belong to, the number of vehicles for each 
model equipped with the technology, and the associated off-cycle 
credits (grams/mile) available for each technology. For each compliance 
category (domestic passenger car, import passenger car, and light 
truck), calculate the fleet off-cycle fuel consumption improvement 
value in gallons/mile in accordance with the equation specified in 40 
CFR 600.510-12(c)(3)(ii).
    (iii) Provide a list of full-size pickup trucks in your fleet that 
meet the mild and strong hybrid vehicle definitions. For each mild and 
strong hybrid type, identify vehicles by make and model types that have 
the technology, the number of vehicles produced for each model equipped 
with the technology, the total number of full-size pickup trucks 
produced with and without the technology, the calculated percentage of 
hybrid vehicles relative to the total number of vehicles produced, and 
the associated full-size pickup truck credits (grams/mile) available 
for each technology. For the light truck compliance category, calculate 
the fleet pickup truck fuel consumption improvement value in gallons/
mile in accordance with the equation specified in 40 CFR 600.510-
12(c)(3)(iii).
* * * * *

0
29. Amend Sec.  537.8 by revising paragraph (a)(3), adding paragraphs 
(a)(4) and (b)(4), and revising paragraph (c)(1) to read as follows:


Sec.  537.8  Supplementary reports.

    (a) * * *
    (3) For model years through 2022, each manufacturer whose pre-model 
or mid-model year report omits any of the information specified in 
Sec.  537.7(b) or (c) shall file a supplementary report containing the 
information specified in paragraph (b)(3) of this section. Starting 
model year 2023, each manufacturer whose pre-model or mid-model year 
report omits any of the information shall resubmit the information with 
other information required in accordance with the NHTSA CAFE 
Projections Reporting Template (OMB Control No. 2127-0019, NHTSA Form 
1474).
    (b) * * *
    (4) The supplementary report required by paragraph (a)(4) of this 
section must contain:
    (i) All information omitted from the pre-model or mid-model year 
reports under Sec.  537.6(c)(2); and
    (ii) Such revisions of and additions to the information submitted 
by the manufacturer in its pre-model or mid-model year reports 
regarding the automobiles produced during the current model year as are 
necessary to reflect the information provided under paragraph (b)(4)(i) 
of this section.
    (c)(1) Each report required by paragraphs (a)(1), (2), (3), or (4) 
of this section must be submitted in accordance with Sec.  537.5(c) not 
more than 45 days after the date on which the manufacturer determined, 
or could have determined with reasonable diligence, that the report was 
required.
* * * * *

    Dated: March 30, 2020.
Andrew Wheeler,
Administrator, Environmental Protection Agency.
    Issued on March 30, 2020 in Washington, DC, under authority 
delegated in 49 CFR 1.95 and 501.5
James Clayton Owens,
Acting Administrator, National Highway Traffic Safety Administration.
[FR Doc. 2020-06967 Filed 4-20-20; 4:15 pm]
BILLING CODE 4910-59-P