Clean Air Act: EPA Should Use Available Data to Monitor the	 
Effects of Its Revisions to the New Source Review Program	 
(22-AUG-03, GAO-03-947).					 
                                                                 
A recent Environmental Protection Agency (EPA) final rule	 
changing the Clean Air Act's New Source Review (NSR) program--a  
key means to protect public health and enhance air quality--has  
been under scrutiny by the Congress, industry, environmental	 
groups, state and local air quality agencies, and the courts. GAO
was asked to determine the basis of EPA's conclusions that (1)	 
the rule's economic impacts would not be significant enough to	 
merit a detailed analysis and (2) the NSR program, prior to the  
rule, discouraged some energy efficiency projects. GAO, among	 
other things, reviewed EPA's analysis of the rule and its	 
impacts, as well as guidance from EPA and the Office of 	 
Management and Budget (OMB) on analyzing such impacts. GAO also  
met with industry and environmental stakeholders.		 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-03-947 					        
    ACCNO:   A08063						        
  TITLE:     Clean Air Act: EPA Should Use Available Data to Monitor  
the Effects of Its Revisions to the New Source Review Program	 
     DATE:   08/22/2003 
  SUBJECT:   Air pollution control				 
	     Economic analysis					 
	     Energy efficiency					 
	     Environmental legislation				 
	     Program evaluation 				 
	     Data integrity					 
	     New Source Review Program				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-03-947

Report to Congressional Requesters

United States General Accounting Office

GAO

August 2003 CLEAN AIR ACT EPA Should Use Available Data to Monitor the
Effects of Its Revisions to the New Source Review Program

GAO- 03- 947

Consistent with agency guidance, EPA used a limited screening analysis
that relied on staff*s professional judgment and public comments from
earlier reform proposals to conclude that the final rule would decrease
emissions and health risks and not impose significant costs. EPA
determined that neither the rule*s benefits nor its costs would exceed a
$100 million threshold that triggers requirements to conduct a more
comprehensive assessment. EPA issued the rule to streamline the NSR
permitting process and provide flexibility to industry. For example, the
rule provides a

mechanism for companies to develop plantwide emissions limits, which would
allow them to make changes in one part of a facility*s operations as long
as they offset emissions increases with decreases elsewhere within the
facility. While OMB agreed with EPA*s conclusion that the rule would not
have significant economic effects, it determined that the rule was
significant for policy reasons. Therefore, OMB asked EPA if it could
better quantify the

rule*s potential impacts, but the agency lacked the necessary data to do
so. EPA lacked comprehensive data on the program*s economic impacts, and
could not predict how many facilities would use the rule*s optional
provisions. Several states and environmental groups disagree with EPA*s
conclusions, claiming that it will enable facilities to increase their
emissions. These parties have filed suit against EPA challenging the rule
and also have petitioned EPA to reconsider the rule. We did not identify
any

comprehensive assessments that contradicted or supported EPA*s conclusions
or the assertions of those who oppose the rule. Because of the data
limitations, it was not possible to verify EPA*s conclusions about the
rule*s effects. Because it lacked comprehensive data, EPA relied on
anecdotes from the

four industries it believes are most affected by NSR to conclude that the
NSR program (prior to the rule) discouraged some energy efficiency
projects, such as upgrades to industrial boilers, including some that
would have

decreased emissions. Because the information is anecdotal, EPA*s findings
do not necessarily represent the program*s effects across the industries
subject to the program. Several environmental groups disputed EPA*s
findings. One such group said that factors other than NSR, such as
economic downturns, discouraged the projects. Furthermore, EPA*s
conclusion that some projects would have decreased emissions assumed that
facilities would not increase production after performing the projects.
However, according to EPA and the executive director of an industry group,
companies often expand production after implementing energy efficiency
projects because it is advantageous to maximize production at the most
efficient facilities. Such expansions could increase emissions and related
health risks, although EPA asserts that this would be offset by decreased
production and emissions at less efficient facilities. A recent
Environmental Protection

Agency (EPA) final rule changing the Clean Air Act*s New Source Review
(NSR) program* a key

means to protect public health and enhance air quality* has been under
scrutiny by the Congress, industry, environmental groups, state and local
air quality agencies, and the courts. GAO was asked to determine the basis
of EPA*s conclusions that (1) the rule*s economic impacts would not be
significant enough to merit a detailed analysis and (2) the NSR program,
prior to the rule,

discouraged some energy efficiency projects. GAO, among other things,
reviewed EPA*s analysis of the rule and its impacts, as well as guidance
from EPA and

the Office of Management and Budget (OMB) on analyzing such impacts. GAO
also met with industry and environmental stakeholders.

Because of the lack of data and uncertainties about the rule*s impacts, we
recommend that EPA determine what data are available to monitor the rule*s
effects, identify additional data needs and

ways to fill them, and use the monitoring results to determine whether the
rule has created adverse effects that the agency needs to address. EPA
agreed with GAO*s conclusions and

recommendations.

www. gao. gov/ cgi- bin/ getrpt? GAO- 03- 947. To view the full report,
including the scope and methodology, click on the link above. For more
information, contact John Stephenson at stephensonj@ gao. gov. Highlights
of GAO- 03- 947, a report to the

Ranking Minority Member of the Committee on Environment and Public Works,
U. S. Senate, and another requester

August 2003

CLEAN AIR ACT

EPA Should Use Available Data to Monitor the Effects of Its Revisions to
the New Source Review Program

Page i GAO- 03- 947 Clean Air Act Letter 1 Results in Brief 3 Background 5
EPA*s Economic Analysis of the Final Rule Complied with EPA and

OMB Cost- Benefit Analysis Requirements, but Some Stakeholders Have Sought
to Have EPA Reconsider the Rule 10 EPA Relied Primarily on Anecdotes from
Industries Most Affected

by NSR to Conclude That it Discouraged Some Energy Efficiency Projects 16
Conclusions 24 Recommendations for Executive Action 25 Agency Comments 25
Appendix I Objectives, Scope, and Methodology 27

Appendix II GAO Contacts and Staff Acknowledgments 29 GAO Contacts 29
Acknowledgments 29 Table

Table 1: Chronology of the New Source Review Program 8 Figure

Figure 1: Percentage of Total U. S. Emissions Released by Industrial,
Transportation, and Other Sources in 2001 6 Contents

Page ii GAO- 03- 947 Clean Air Act Abbreviations

DOJ Department of Justice EPA Environmental Protection Agency NAPA
National Academy of Public Administration NSR New Source Review OMB Office
of Management and Budget WEPCO Wisconsin Electric Power Company v. Reilly

This is a work of the U. S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

Page 1 GAO- 03- 947 Clean Air Act

August 22, 2003 The Honorable James M. Jeffords Ranking Minority Member
Committee on Environment and Public Works United States Senate

The Honorable Joseph I. Lieberman United States Senate

Recent changes to the Clean Air Act*s New Source Review (NSR) program* one
of the act*s key mechanisms for maintaining air quality to protect public
health* have been the subject of congressional debate and

have drawn scrutiny from numerous stakeholders, including representatives
of industry, environmental groups, and state and local air pollution
control authorities. While some industry officials describe the existing
program as costly and characterized by uncertainty, environmental groups
and a coalition of state attorneys general assert that

it is, and has been, an important component of the Clean Air Act. In
recent years, the program has become increasingly controversial, as the
Environmental Protection Agency (EPA) has taken enforcement action against
companies in several industries, including some electricity producers,
forest products manufacturers, and petroleum refineries, alleging
noncompliance with the program. Some of the affected companies have agreed
to settlements that will cost hundreds of millions of dollars and require
emissions reductions, while others are in various stages of litigation.

The NSR program, which seeks to protect public health, maintain compliance
with air quality standards, and preserve and enhance air quality in
national parks and scenic areas, requires companies that are major sources
of air pollution to install pollution controls in their facilities when
constructed. The program also requires companies to install such controls
in existing facilities when making physical or operational changes* such
as the addition of new production equipment* that cause a significant
increase in air emissions. 1 Such changes are called *major

1 The thresholds for these so- called major modifications* physical or
operational changes that cause a significant increase in emissions* vary
by pollutant and the air quality status of the area in which a facility is
located.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 03- 947 Clean Air Act

modifications.* Congress believed that incorporating pollution controls
into the design and construction of new and modified air pollution sources
was generally an efficient way of controlling air pollution from large
industrial sources. Congress also excluded existing facilities from NSR
requirements until they made changes that increased their emissions.
Companies that want to make major modifications in existing facilities
must apply to state or local agencies for an NSR permit and then install
the controls. The cost of installing controls varies but can reach
hundreds of millions of dollars for some facilities, according to an EPA
program manager. However, companies can qualify for exemptions from these
requirements if, for example, (1) a modification is considered *routine
maintenance and repair,* (2) the company agrees not to significantly
increase its emissions after making a physical or operational change to
its facility, or (3) the company offsets any emissions increases resulting
from a change in a facility with emissions reductions achieved elsewhere
within that facility.

EPA has long recognized a need to revise the NSR program and began a
reform process in 1992 that resulted in proposed changes to the program in
1996 and 1998. The agency received wide- ranging comments from the public
on how the program should be revised and held meetings with the public and
other stakeholders, but did not develop final rules by the time the new
administration took office in 2001. In May 2001, as part of its proposed
national energy policy, the Vice President*s National Energy Policy
Development Group recommended that EPA report to the President on the NSR
program*s impact on energy efficiency investments, among other things. In
response to this recommendation, EPA concluded in its June 2002 NSR Report
to the President, that NSR had discouraged some energy efficiency
investments at existing industrial facilities.

After completing this report, EPA modified certain of the proposed 1996
NSR revisions and finalized them as a rule in December 2002 (hereafter
referred to as the NSR final rule). According to EPA, the rule will, among
other things, provide greater certainty for facilities regulated under the
program and streamline the NSR permitting process while ensuring the
current level of environmental protection. As part of this process, EPA
analyzed the rule*s anticipated economic effects, such as its impacts on
emissions, health risks, the costs of installing and maintaining pollution
control equipment, and administrative costs incurred by government
agencies. Under Executive Order 12866 and the Unfunded Mandates Reform Act
of 1995, agencies must perform detailed assessments of economically
significant rules* those rules that may have an annual effect on the
economy of $100 million or more. If a rule*s impacts are not

Page 3 GAO- 03- 947 Clean Air Act

expected to exceed this threshold, a more detailed economic analysis is
generally not required. According to the Office of Management and Budget
(OMB), the agency responsible for overseeing agency compliance with

Executive Order 12866, agencies may use their discretion when conducting a
screening analysis to determine whether a rule*s economic impacts may
reach the $100 million threshold and require a more detailed assessment.
EPA*s Office of Air Quality Planning and Standards has developed a
guidance document that the agency uses to analyze the impacts of air
quality rules, which discusses how to conduct a screening analysis. 2 You
asked us to determine the basis of (1) EPA*s analysis of the economic
impacts of the final rule and its conclusion that the rule would not
create

significant enough benefits or costs to require a more detailed analysis
and (2) EPA*s conclusions that the NSR program (prior to the final rule)
discouraged some energy efficiency projects. You also asked us to provide
information on several other aspects of the final rule and proposed
revisions to the definition of routine maintenance and repair under NSR,
which we will address in subsequent reports.

To respond to these objectives, among other things, we used OMB and EPA
guidance to review EPA*s screening analysis of the final rule*s economic
impacts. We also met with the NSR program manager within EPA, other senior
EPA officials within the agency*s Office of Air Quality Planning and
Standards, and senior OMB staff within the Office of Information and
Regulatory Affairs who were responsible for reviewing EPA*s analysis. In
addition, we reviewed the information that EPA relied on in preparing its
findings on the NSR program*s effects on energy efficiency projects. We
also met with representatives of industry and an environmental group.
Appendix I provides a more detailed description of our scope and
methodology.

EPA relied primarily on the professional judgment of agency staff and
comments it received on earlier NSR revision proposals to conclude in its
screening analysis that the final rule would not generate benefits or
costs

of more than $100 million and, therefore, that it could proceed with the 2
OAQPS Economic Analysis Resource Document, U. S. Environmental Protection
Agency, Office of Air Quality Planning and Standards Innovative Strategies
and Economics Group, April 1999. Results in Brief

Page 4 GAO- 03- 947 Clean Air Act

rule without a detailed economic analysis. In taking this approach, the
agency complied with its guidance for conducting economic analyses, which
states that, to focus resources on those rules that will have a large
impact, a screening analysis of benefits and costs may be qualitative or
rely on limited data. From this screening analysis, EPA concluded that the
rule would encourage energy efficiency projects while reducing emissions
and related health risks without imposing significant economic impacts.
Senior OMB staff responsible for reviewing the analysis said that while
OMB concurred with EPA*s conclusion that the rule was not economically
significant, it considered the rule significant for policy reasons. As a
result,

OMB sought to determine whether EPA could quantify the final rule*s
potential effects, but concluded that the agency lacked the necessary data
to do so. For example, EPA does not maintain comprehensive information

on the economic impacts of the NSR program, and the agency could not model
how often or when companies would decide to use any of the voluntary
provisions of the rule. EPA later conducted two additional analyses of
some of the rule*s impacts to provide the public with more information and
to satisfy requirements of the Paperwork Reduction Act, but these were not
comprehensive assessments of the final rule. Attorneys

general from 10 Northeastern states and several environmental
organizations have filed suit against EPA in a challenge to the final
rule, asserting, among other things, that the rule will allow companies to

increase their emissions. If these claims prove correct, EPA*s screening
analysis would have underestimated the rule*s impacts because it did not
account for costs associated with increased emissions, such as adverse
public health effects. Because of the data limitations, it was not
possible to verify these parties* or EPA*s conclusions. As a result, the
rule*s effects are uncertain. Therefore, we are recommending that EPA
identify the data it has available, as well as additional data it needs
and could obtain, to monitor the effects of the final rule and use the
monitoring results to determine whether the rule has created adverse
effects that the agency needs to address.

EPA relied primarily on anecdotal information from the industries most
affected by NSR in concluding that (prior to the final rule) the program
discouraged some energy efficiency projects, including some that would
have reduced air emissions. EPA staff responsible for this analysis said
they relied on anecdotal information from industry sources such as
electricity producers, chemical and forest products manufacturers, and
petroleum refiners because they lacked comprehensive data on the number of
projects that did not go forward as a result of NSR, such as upgrades to
industrial boilers. These anecdotes suggested that the NSR program posed
several barriers that discouraged some energy efficiency

Page 5 GAO- 03- 947 Clean Air Act

projects, such as the high costs of installing pollution controls and
delays in obtaining permits that in turn delayed project construction.
Several environmental groups, however, disagreed with industry*s claims.
Because EPA based its conclusion that NSR discouraged some energy
efficiency projects on anecdotal information rather than a comprehensive
survey or representative sample of industries subject to the program, its
findings are

not necessarily representative of the program*s effect on energy
efficiency projects throughout the industries subject to the program. In
addition, EPA*s finding that some forgone energy efficiency projects would
have reduced air emissions was based on the assumption that facilities
would not increase their production levels after performing the projects.
However, facilities* future levels of production and emissions are
uncertain because they may fluctuate in response to economic conditions,
and other factors. For example, according to EPA and the executive
director of an industry group, companies often expand production after
implementing energy efficiency projects because it is advantageous to
maximize production at the most efficient facilities. Such expansions
could increase emissions and related health risks, although EPA asserts
that this would be offset by decreased production and emissions at less
efficient facilities.

Under the Clean Air Act, EPA establishes health- based air quality
standards that the states must meet and regulates air pollutant emissions
from various sources, including industrial facilities and mobile sources
such as automobiles and other transportation. Figure 1 compares the
emissions of key pollutants from industrial facilities to those from
transportation and other sources. Background

Page 6 GAO- 03- 947 Clean Air Act

Figure 1: Percentage of Total U. S. Emissions Released by Industrial,
Transportation, and Other Sources in 2001

Note: Percentages for carbon monoxide, nitrogen oxides, and sulfur dioxide
do not total 100 due to rounding.

EPA has issued health- based air quality standards for six primary
pollutants* carbon monoxide, lead, nitrogen oxides, ozone, 3 particulate
matter, and sulfur dioxide* that have been linked to a variety of health
problems. For example, ozone can inflame lung tissue and increase

susceptibility to bronchitis and pneumonia. In addition, nitrogen oxides
and sulfur dioxide contribute to the formation of fine particles that have
been linked to aggravated asthma, chronic bronchitis, and premature

3 Ozone forms when nitrogen oxides react with volatile organic compounds
in the presence of heat and sunlight.

Page 7 GAO- 03- 947 Clean Air Act

death. In 2001 (the most recent year for which data were available), 133
million Americans lived in areas with air pollution levels above at least
one of the health- based air quality standards, according to EPA.

The New Source Review program was established in 1977 and is intended to
protect public health, as well as national parks and wilderness areas,
from additional air pollution when new industrial facilities are built and
existing ones expand. The fundamental logic of the program, according to
EPA, is that industrial facilities should install modern pollution
controls at

the time of construction or when making physical or operational changes,
such as adding new production equipment, that cause a significant increase
in air emissions. Subject to EPA*s oversight, state and local air quality
agencies generally administer air quality programs, including the NSR
program. In recent years, EPA has taken enforcement action against
companies in several industries, including some electricity producers,
forest products manufacturers, and petroleum refineries, alleging
noncompliance with the program. Some of these parties settled these cases
soon after the enforcement actions were filed, although electricity
producers assert that the actions are inconsistent with the Clean Air Act,

according to the U. S. Department of Justice (DOJ). In January 2002,
however, DOJ concluded that the enforcement actions are consistent with
the act.

Recognizing the need for revisions to the NSR program, EPA began a reform
effort in 1992 and 1993 when it held workshops with stakeholders and
established a federal advisory committee in 1993. Largely on the basis of
this committee*s recommendations, EPA issued proposed NSR revisions in
1996 that were intended to reduce costs imposed on companies that undergo
NSR permitting without interfering with efforts to attain air quality
goals. EPA solicited public comment on the proposals at that time and
again in 1998, when it sought additional information on an alternative
method for determining whether a facility modification should be subject

to NSR. The agency received numerous comments that provided wideranging
views on how the program should be revised. Despite additional public
meetings and discussions with stakeholders on NSR reforms, EPA

had not developed final rules by the time the new administration took
office in 2001.

In May 2001, when the Vice President*s National Energy Policy Development
Group issued its proposed national energy policy, it recommended that EPA
report to the President on the NSR program*s impact on investments in new
utility and refinery generation capacity, energy efficiency, and
environmental protection. In response to this

Page 8 GAO- 03- 947 Clean Air Act

recommendation, and given that EPA does not maintain such information, the
agency solicited public input on how the NSR program had affected the
ability of companies to undertake energy efficiency projects in their
existing facilities. EPA defined energy efficiency projects as those that
would have produced greater output per unit of fuel input (e. g., more
electricity per ton of coal burned), regardless of the effect on
emissions. In its June 2002 NSR Report to the President, EPA concluded,
among other things, that NSR had not affected investments in new power
plants and refineries but had discouraged some energy efficiency projects
at existing facilities, including some that would have reduced air
emissions.

After completing this report, EPA modified the 1996 proposed NSR revisions
to provide regulatory flexibility to industrial facilities so that they
could pursue energy efficiency projects, among other things. EPA assessed
the economic impacts of implementing these revisions, and finalized them
as a rulemaking* hereafter referred to as the *final rule** in December
2002. Table 1 provides a chronology of the NSR program.

Table 1: Chronology of the New Source Review Program Date Description

1970 Clean Air Act became law. 1972 EPA created the Prevention of
Significant Deterioration Program by

rulemaking. This program implemented NSR in areas that meet air quality
standards.

1977 Clean Air Act Amendments of 1977 became law. 1990 Clean Air Act
Amendments of 1990 became law. 1992- 1994 EPA issued notices of violation
to companies in the plywood and wood

products industry. 1993 EPA convened a federal advisory committee to
address policy and

technical issues associated with revising NSR. 1996 EPA issued a NSR
Simplification Proposal to streamline permitting,

relieve regulatory burden, and provide states with flexibility. EPA also
began investigating coal- fired electricity producers, petroleum refiners,
and the pulp and paper industry for violations of NSR rules. 1998 EPA
solicited further public comment on NSR revisions.

1999 DOJ filed lawsuits against seven electricity producers charging that
17 power plants made major modifications without installing required
pollution control equipment.

2000- 2003 EPA settled several NSR cases with electricity producers and
refiners. May 2001 The administration*s proposed energy policy called for
EPA and the

Department of Energy to review the implementation of NSR regulations, and
for DOJ to review existing NSR legal actions. DOJ later reported that the
actions were consistent with the Clean Air Act.

Page 9 GAO- 03- 947 Clean Air Act

Date Description

June 2001 EPA issued a NSR background paper as a partial response to
recommendations in the energy plan.

June 2002 EPA issued New Source Review: Report to the President and
recommendations for improving the NSR program.

December 2002 EPA issued the NSR final rule and nine northeast states
filed suit

challenging the final rule. January 2003 A tenth northeast state filed
suit challenging the rule, and these states,

California, and four California air quality agencies petitioned EPA to
reconsider the final rule.

July 2003 EPA announced that it would reconsider parts of the NSR final
rule. Source: EPA and National Academy of Public Administration.

Specific revisions in the final rule include the following:  a revised
method for determining a facility*s baseline emissions level that a
company would use as the starting point for determining whether any
changes in emissions resulting from a planned physical change or change in
the method of operation subjected the company to NSR;

 a revised test that a company would use after establishing a facility*s
baseline emissions level to determine if a physical or operational change
would increase emissions beyond the NSR threshold;

 exemptions from the program if companies demonstrate that (1) equipment
qualifies as a *clean unit* because they already use state- of- theart
pollution control equipment or (2) a proposed modification specifically
controls air pollution and achieves an environmental benefit; and

 a mechanism for companies to work with state or local permitting
authorities to develop plantwide emissions limits, which would allow
companies to make changes in one part of a facility*s operations as long
as they offset any emissions increases with decreases elsewhere within the

facility. In addition to the final rule, EPA has proposed further NSR
revisions that the agency believes will provide greater certainty about
activities that are considered routine maintenance, repair, and
replacement. According to a NSR program manager, the agency is reviewing
public comments on this proposal and expects to finalize the rule by
December 2003.

Page 10 GAO- 03- 947 Clean Air Act

EPA relied primarily on the professional judgment of its staff, as well as
public comments on the agency*s prior proposal to revise the NSR program,
in concluding from its screening analysis that the final rule would not
create benefits or costs beyond the $100 million threshold that triggers
requirements for a more detailed economic analysis. EPA*s approach, while
limited, is consistent with agency guidance for assessing the economic
impacts of proposed rules. In addition, EPA would have had difficulty
conducting a more quantitative analysis because of data

limitations. OMB agreed that the rule would not have a significant
economic impact but was significant for policy reasons. OMB asked EPA if
it could better quantify impacts and was convinced that the agency lacked
the necessary data to do so. EPA did later conduct two additional analyses

of some of the rule*s costs and benefits, but they also were not
comprehensive economic assessments. Some stakeholders have formally asked
EPA to reconsider the rule, arguing, among other things, that it will
enable facilities to increase their emissions. Because of the limited data
on the NSR program, it was not possible to verify agency or stakeholder
conclusions about the rule*s anticipated economic impacts.

EPA*s screening analysis of the final rule*s anticipated effects was
consistent with the agency*s guidance for conducting economic analyses.
According to senior OMB staff, the office does not have guidance for
agencies to use when conducting a screening analysis to determine whether
a rule will impose significant economic impacts and, thus, merit further
analysis. Therefore, agencies have latitude in determining how best to
conduct a screening analysis. EPA*s Office of Air Quality Planning and
Standards has developed a guidance document that describes the process
agency economists should use when analyzing air quality rules. Recognizing
the need to focus agency resources on rules that have a large impact, the
guidance states that a screening analysis of benefits and costs

may be qualitative in nature or rely on limited data. According to a NSR
program manager in EPA, agency staff relied primarily on their
professional judgment in estimating the rule*s economic impacts, such as
its effect on air pollutant emissions and the costs companies incur when
they install pollution controls, as well as public comments the agency
received on its 1996 and 1998 NSR revision proposals. For example, several
industry trade associations submitted information asserting that the
ability to use plantwide emissions limits would reduce EPA*s Economic

Analysis of the Final Rule Complied with EPA and OMB CostBenefit Analysis
Requirements, but Some Stakeholders Have Sought to Have EPA Reconsider the
Rule

EPA*s Reliance on Professional Judgment Was Consistent with Agency
Guidance for Screening the Economic Impacts of Rules

Page 11 GAO- 03- 947 Clean Air Act

costs for industry and provide other benefits without compromising air
quality. 4 On the basis of this information, EPA staff determined that the
rule would lead to overall economic and environmental benefits by
encouraging energy efficiency projects, reducing emissions and related
health risks, and providing economic benefits to companies affected by the
NSR program, according to the NSR program manager. For example, EPA
forecasted that the rule would encourage companies to implement energy
efficiency projects that would reduce emissions, such as upgrades to
boilers used to generate power.

In its screening analysis, EPA assumed that the final rule would not
impose significant economic costs on companies because the rule created
voluntary options and companies would most likely only elect to use them

if they thought the provisions would achieve an overall economic benefit.
Therefore, the agency assumed that any time a company opted to use one of
the provisions, the benefits to the company would outweigh any costs
incurred. In addition, because EPA concluded that the rule would decrease
emissions, it did not forecast any increases in public health costs
resulting from the final rule, such as increased incidence of asthma or
other respiratory problems. Consistent with its guidance, EPA then
concluded that, because the rule was not expected to create $100 million
in benefits or costs, the agency could proceed in finalizing the rule
without a more quantitative and comprehensive analysis.

Even if EPA had been required to conduct a more detailed economic
analysis, it would have had difficulty doing so because the agency is not
required to systematically collect comprehensive data on the economic
effects of the NSR program. 5 Regarding the benefits of the program, EPA
does not maintain comprehensive data on the number and type of facilities
that obtain NSR permits, or the reduced air emissions achieved after
facilities install pollution controls, according to a senior agency
economist. In 2001, EPA attempted to estimate the emissions reductions at
facilities that obtained NSR permits; however, senior agency officials

4 EPA also relied on an analysis of flexible permitting programs as part
of the basis for the agency*s findings regarding the benefits of plantwide
emissions limits. 5 Under section 312 of the Clean Air Act, EPA
periodically reports on the overall costs incurred and benefits achieved
under the act. However, in fulfilling this requirement, the agency
generally provides a comprehensive assessment of such impacts and does not
provide a breakout of the costs and benefits of individual programs under
the act. EPA Lacked Data to Conduct a More

Comprehensive Analysis of the Economic Impacts of the NSR Final Rule

Page 12 GAO- 03- 947 Clean Air Act

responsible for the analysis acknowledged that it had several limitations.
6 For example, these officials said the analysis included only facilities
that were located in areas that met federal air quality standards, thereby
excluding a large portion of the universe of affected facilities. In
addition, EPA had incomplete data on facilities located in EPA region 6,
which includes Arkansas, Louisiana, New Mexico, Oklahoma, and Texas.
Furthermore, the analysis did not distinguish between benefits that
resulted from the installation of pollution controls at new facilities and
those at existing facilities, which are the focus of the final rule. With
respect to costs, EPA is not required to maintain comprehensive
information on the costs of the NSR program, and would therefore have had
difficulty quantifying all of the costs of the rule, according to a NSR
program manager.

In addition to EPA*s lack of data on the NSR program*s benefits and costs,
a senior agency economist said that uncertainty about the extent to which
companies might elect to use the NSR alternatives provided in the final
rule also limited EPA*s ability to estimate the rule*s impacts. For
example, the economist said that the final rule allows companies to
develop a plantwide emissions limit as an alternative to NSR, but only
those companies that find this provision advantageous are likely to use it
and EPA could not accurately determine how many companies this might
include. According to EPA, companies* decisions about whether to pursue
voluntary options are case- specific and dependent on a number of factors.
Therefore, the agency was unable to model how often and when the final
rule*s options would be used. In contrast, most of the other rules EPA

develops generally impose new requirements on a known universe of
companies, according to a NSR program manager. In these cases, it is much
easier to determine the costs and benefits of a rule because the agency
can gather information from the affected companies. For example, if EPA
required all companies within a particular industry to install certain
pollution controls, it could gather information on the average costs of
such equipment and the anticipated reductions in air emissions. Because of
these data limitations we identified, it was not possible to conduct our
own assessment of the final rule*s possible effects and verify EPA*s
analyses and conclusions.

6 This analysis was summarized in an October 2001 EPA memorandum, Benefits
of the Prevention of Significant Deterioration Program.

Page 13 GAO- 03- 947 Clean Air Act

According to senior EPA and OMB staff, OMB agreed with EPA*s finding that
the final rule was not economically significant because it was not
expected to impose costs or provide benefits beyond the $100 million
threshold that triggers requirements to conduct a more thorough analysis.
Nevertheless, according to senior OMB staff responsible for reviewing the
analysis, while the office found EPA*s analysis* which was presented in an
oral briefing but not documented* persuasive, OMB determined that the
final rule was significant for policy reasons. Under Executive Order
12866, rules that *raise novel legal or policy issues arising out of legal
mandates, the President*s priorities* or other criteria can be categorized
as

significant. According to the Executive Order, when rules fall into this
category, the agency issuing the rule must provide OMB with an assessment
of the potential costs and benefits of the regulatory action, but the
order does not elaborate on the form of the assessment. The senior OMB
staff said that EPA*s screening analysis satisfied this requirement.
Nevertheless, OMB staff asked EPA if it would be possible to conduct an
analysis that quantified the rule*s effects. However, as we previously

discussed, EPA identified numerous data limitations that it claimed
prevented its staff from conducting such an analysis, and OMB acknowledged
these limitations and concurred with EPA.

In November 2002, EPA issued a supplemental analysis intended to provide
the public with additional information on the rule*s potential
environmental effects. According to EPA, this analysis was not intended as
a comprehensive economic analysis of the rule*s benefits and costs and was
not used to make decisions about the rule. Like the screening analysis, it
relied primarily on qualitative information and arrived at similar
conclusions. For example, EPA asserted that the exemption for companies
that use state- of- the- art pollution controls would save companies NSR
permitting costs. EPA also asserted that this provision would induce
facilities to voluntarily install controls to avoid NSR, thereby reducing
emissions.

While the supplemental analysis was qualitative, it used some data from a
limited number of facilities to estimate the effects of some of the rule*s
provisions on a wider universe of facilities. Specifically, the analysis
considered the experiences of six companies that had used regulatory
options similar to those provided for in the final rule to determine that
such limits would lead to emissions reductions. For example, on the basis
of these six case studies, EPA stated that if 75 percent of facilities in
three industry sectors opted to use plantwide emissions limits, emissions
of volatile organic compounds could be cut by up to 17,000 tons annually
OMB Concurred with

EPA*s Analysis and Conclusions and Determined That Data Limitations
Precluded More Quantitative Analysis

EPA Conducted Two Additional Analyses of the Rule*s Effects, but They Did
Not Comprehensively Assess Economic Impacts

Page 14 GAO- 03- 947 Clean Air Act

(less than 1 percent of the total volatile organic compounds emitted in
2001, the most recent year for which data were available). EPA also said
that the emissions reductions would be greater if the analysis was
extended to other industry sectors and pollutants.

However, EPA did not use statistically valid methods to identify the six
companies on which it based this portion of its analysis. Therefore, the
experiences of these companies may not be representative of how plantwide
limits will affect emissions at other industrial companies that may opt to
use this provision. In addition, the Secretary of the Delaware Department
of Natural Resources and Environmental Control* the state in which one of
the six companies was located* wrote the EPA

Administrator cautioning against using the experience of the Delaware
company to support the final rule. The Secretary noted that the regulatory
option used by that company provided for emissions reductions as a

prerequisite for participation, while EPA*s plantwide emission limit does
not.

As noted above, EPA could not determine with any certainty the number of
facilities that would opt to use the final rule*s voluntary provisions, or
the changes in the number of NSR permits, amount of emissions, or other
effects that would result. However, EPA was required under the Paperwork
Reduction Act to assess some of the costs and benefits that would accrue
to companies and government agencies under the final rule. Specifically,
the act requires agencies to estimate the record keeping

burden associated with a rule and report this information to OMB.
Therefore, EPA relied on limited available data and its professional
judgment to make estimates necessary to satisfy this requirement. In
February 2003, after issuing the final rule, EPA estimated that it would
impose about $6.5 million in annual burden on state and local air quality
agencies, which include legal and other costs associated with
incorporating the final rule into the state*s air pollution control plan,
collecting public comment on the changes, and obtaining state
legislatures* approval of the changes. 7 This analysis also estimated that
14 facilities would use the final rule*s provisions during each of the
first 3

7 During the first 3 years of implementation, the final rule will only
affect regulatory agencies and companies in jurisdictions that meet the
federal air quality standards. According to EPA, about 10 to 12 percent of
all affected companies are located in such areas. Other jurisdictions are
not required to revise their NSR programs to accommodate the final rule
until 2006. Therefore, the final rule is not expected to impose costs on

regulatory agencies and companies in these areas until 2006.

Page 15 GAO- 03- 947 Clean Air Act

years of implementation, reducing the previous annual burden by $650,000.
This analysis, like the screening analysis, found that the rule would
impose less than $100 million in annual costs on companies and government
agencies. However, a senior EPA economist said this was a limited analysis
intended to identify information collection and record keeping
requirements that the final rule would impose. In addition, this analysis
does not comprehensively address all of the costs that are likely to
result from implementation of the rule.

Nine northeast states filed a petition in the U. S. Court of Appeals for
the District of Columbia Circuit on December 31, 2002* the day the rule
was finalized* disagreeing with EPA*s conclusions about the rule*s
effects. 8 They asserted, among other things, that the final rule violated
the Clean

Air Act and would enable companies to increase their emissions because, by
using the provisions to opt out of NSR, they will no longer be required to
install pollution control equipment. 9 According to New York*s Attorney
General, the final rule will lead to more smog, asthma, and respiratory
disease. In addition, Earthjustice* acting on behalf of a coalition of
environmental and public health advocacy groups* filed similar petitions
for review, also in the U. S. Court of Appeals for the District of
Columbia Circuit. According to the American Lung Association, one of the
groups represented by Earthjustice, the final rule creates loopholes that
allow companies to increase their emissions without installing pollution
controls. Similarly, Environmental Defense, another group represented by
Earthjustice, claims that the final rule will enable thousands of
factories,

power plants, and other industrial companies to pollute more. 10 On
February 6, 2002, the nine Northeast states, along with Pennsylvania,
filed a motion, which the court denied, seeking to halt implementation of
the

final rule pending a ruling on the earlier petitions. 8 According to DOJ,
EPA has also received eight formal petitions seeking to have EPA
reconsider the NSR final rule. In addition to the parties identified
above, four air quality agencies within the state of California, and the
state itself have formally requested that EPA reconsider the final rule.

9 Connecticut, Maine, Maryland, Massachusetts, New Hampshire, New Jersey,
New York, Rhode Island, and Vermont were the original petitioners. A tenth
state, Pennsylvania, filed a petition for review on January 28, 2003. 10
This report does not address the merits of the claims made by the
litigants in these cases. Some State Attorneys General and

Environmental Organizations Question EPA*s Conclusions and Assert That the
Final Rule Will Increase Emissions, Harming Public Health

Page 16 GAO- 03- 947 Clean Air Act

While these parties do not support the final rule, several states,
including Indiana, Kansas, Nebraska, North Dakota, South Carolina, South
Dakota, and Utah, as well as the American Petroleum Institute and other
industry groups, have filed petitions with the court in support of the
final rule. In July 2003, EPA responded in part to the petitions for
reconsideration by requesting public comment on six limited issues in the
final rule. The agency said that the decision to reconsider these issues
did not mean that EPA had decided to change any aspect of the rule and
that the agency would make that decision after the comment period closed.
Our review did not identify any comprehensive assessments of the final

rule*s effects that contradicted or supported the results of EPA*s
analysis or the assertions of those who oppose the final rule. As a
result, the economic impacts of the final rule are uncertain. Two studies
commissioned by the Environmental Integrity Project of the Rockefeller
Family Fund and performed by Abt Associates, an EPA contractor, focused on
facilities that obtained NSR permits and installed emissions controls
prior to the final rule. The studies found that the facilities would not
have been required to install pollution controls if they had made their
modifications after implementation of the final rule, and could have

increased their emissions. However, because these analyses focus on just
two facilities and only one of the four provisions of the final rule,
their results may not be representative of the rule*s overall
environmental effects. In addition, EPA asserts that these studies were
based on an incorrect interpretation and application of the final rule*s
provisions.

EPA relied primarily on anecdotal information from industry in concluding
that the NSR program, prior to the final rule, discouraged some energy
efficiency projects* such as upgrades to industrial boilers* including
some projects that would have reduced air emissions. The anecdotes, which
were provided primarily by four of the industries most affected by the NSR
program, suggested that the program imposed several barriers that deterred
energy efficiency projects, including delays in obtaining permits and the
high costs of installing pollution controls. Several environmental groups
disputed EPA*s findings, and a representative of one group cited other
factors, such as poor economic conditions and a lack of willingness to
control air pollution, as the barriers to energy efficiency

projects, although EPA program managers said that the agency found
industry*s claims to be more persuasive. Nonetheless, because EPA relied
on anecdotal information rather than a statistically valid sample or
industrywide survey, the agency*s findings do not necessarily represent
EPA Relied Primarily

on Anecdotes from Industries Most Affected by NSR to Conclude That It
Discouraged Some Energy Efficiency Projects

Page 17 GAO- 03- 947 Clean Air Act

NSR*s effect on energy efficiency projects throughout the industries
subject to the program.

According to EPA officials responsible for the NSR Report to the
President, during development of the proposed NSR revisions, the agency
received information regarding the program*s effect on energy efficiency
projects from numerous stakeholders. This input included written

responses to EPA*s request for information, as well as case- specific
anecdotes and supplemental documents. In assessing this information, the
agency defined any project that included a facility modification that
would directly result in greater output per fuel input (such as more
electricity generated from each ton of coal burned) as an energy
efficiency project. Although EPA based its conclusion about NSR*s impact
on such projects on the total available information, it relied heavily on
anecdotes describing cases in which chemical manufacturers, electric
utilities, forest products manufacturers, and petroleum refineries* four
of the industries EPA identified as most affected by the program* had
decided not to pursue energy efficiency projects because of NSR. While EPA
also received comments from several environmental organizations and state
and local air quality agencies that disputed industry*s claims, an EPA
program manager said that the specific examples provided by industry were

convincing. After obtaining the anecdotes submitted to EPA by the parties
the agency identified as its primary data sources* including trade
associations, individual firms, and other business interests* and removing
those that did not address energy efficiency or contain complete
information, we reviewed 69 anecdotes that described how NSR deterred
companies from making energy efficiency investments. We determined that
these anecdotes generally went through two rounds of review. First, trade
associations said that their members reviewed and discussed the anecdotes
before submitting them to EPA. After receiving the submissions, EPA
program managers said they used their expertise and

judgment to evaluate the credibility and reliability of the anecdotes.
According to these managers, they paid specific attention to whether (1)
the project described in an anecdote was technically feasible; (2) an
anecdote*s conclusions about how NSR discouraged a project were consistent
with EPA*s understanding of how the NSR provisions would apply to that
project; and (3) the conclusions were based on *real- life,* rather than
hypothetical, situations. EPA program managers found the anecdotes to be
generally credible. While some may have been more relevant than others (e.
g., *real life* examples were more relevant than EPA*s Conclusion That

NSR Discouraged Energy Efficiency Projects Was Based Primarily on
Anecdotes from Four of the Industries It Believes Are Most Affected by the
Program

Page 18 GAO- 03- 947 Clean Air Act

hypothetical ones), none was dismissed for failing to meet these criteria,
according to an EPA program manager. Agency staff familiar with the NSR
program then compiled, reviewed, and synthesized this information and
concluded that complying with NSR may have deterred some investments in
energy efficiency projects, including some that may have reduced air

emissions. It is important to note that the energy efficiency findings in
the

Report to the President were not the basis of the analysis of the economic
impacts of the final rule, which found that the final rule would encourage
energy efficiency projects, according to an EPA manager of the NSR
program.

The anecdotes generally cited three ways in which these industries believe
complying with NSR discouraged energy efficiency projects, including
concerns that (1) EPA would subsequently determine that projects

companies initiated as routine maintenance, repair, and replacement
without an NSR permit were actually major modifications subject to NSR and
enforcement action; (2) the test used to measure the emissions

impacts of company modifications was not fair; and (3) the NSR permitting
process caused unanticipated project delays and increased costs.

The first barrier, concerns about possible enforcement actions, stemmed
from recent EPA enforcement litigation against certain electricity
producers in which EPA contested industry*s claims that certain projects
were exempt from NSR because they qualified as routine maintenance. EPA
maintained that these projects were in fact major modifications that
should have triggered NSR. According to senior industry representatives,

this meant that EPA could consider potentially thousands of projects that
industry had previously completed without an NSR permit under this
exemption as NSR violations, including some that state and local air
quality agencies had approved and confirmed did not trigger NSR. This
prompted some industry officials to allege that EPA was reinterpreting
what could be considered routine maintenance exempt from NSR. EPA*s Office
of Enforcement, however, maintains that the agency is correctly
interpreting and enforcing the program.

According to senior representatives of the electric utility and refining
industries, this litigation has produced substantial uncertainty for
companies pursuing facility modifications. Part of this uncertainty may
stem from what industry officials describe as a lack of clear policy
guidance on what qualifies as routine maintenance. EPA has never
explicitly defined routine maintenance since the exclusion was
established, although some clarification has grown out of EPA

Page 19 GAO- 03- 947 Clean Air Act

enforcement, most notably in the case of Wisconsin Electric Power Co. v.
Reilly (WEPCO). This court decision upheld EPA*s consideration of the
nature, extent, purpose, frequency, and cost of facility modifications, as
well as other relevant factors, when determining whether a project
qualifies for the routine maintenance and repair exemption. EPA maintains
that it takes a case- by- case approach to determining whether a
modification constitutes routine maintenance, and has cited WEPCO as
support for its recent enforcement actions.

However, industry comments submitted to EPA claim that the agency*s
enforcement actions, combined with what they regard as a lack of clear
guidance, make it difficult for them to reliably predict when their
projects will trigger NSR, especially when EPA may later disagree with
state and local NSR determinations. Companies are therefore reluctant to
perform projects that could trigger NSR or possible enforcement
litigation,

including, they assert, energy efficiency projects. According to EPA, the
agency*s proposed rule on the routine maintenance exemption would address
some of these concerns by changing its definition. EPA expects to issue
the rule by the end of 2003.

Several environmental groups disputed EPA*s findings. For example, an
environmental advocacy group involved in NSR issues claimed that the
proposed rule would simply broaden the exemption in violation of the Clean
Air Act. One alternative in the proposed rule would allow companies to
claim as routine maintenance any modifications as long as they cost less
than a certain percentage* depending on the industry* of the total cost of

the polluting unit, such as a boiler, or the entire facility. Another
alternative would allow companies to invoke the routine maintenance
exemption if they are replacing equipment that performs the same function
as its predecessor and does not alter the basic design of a facility.

Environmentalists assert that these exemptions will allow companies to
falsely treat major plant modifications as routine maintenance, avoid NSR
requirements, and increase emissions.

The second barrier cited in the anecdotes was the test used to determine
whether a modification would increase emissions beyond the NSR threshold.
Under the program prior to the final rule, companies making physical or
operational changes that did not qualify for exemptions, such as the
routine maintenance exemption, had to undergo this test to gauge the
changes* effects on emissions. Before the final rule, a company was to
compare a facility*s emissions during the previous 24 months to its future
potential emissions if its facility was run at maximum capacity or the
highest capacity allowed by the existing NSR permit after making the

Page 20 GAO- 03- 947 Clean Air Act

change, even if the facility had not run at this level before, or did not
plan to in the future. If the expected future emissions resulting from the
change were more than 40 tons per year higher after making the change, the
project qualified as a significant emission increase and triggered NSR. 11
Companies (except electric utilities) could request that the permitting

agency allow the use of any different two- year period based on a
demonstration that it is more representative of normal operation. Electric
utilities may use any 2- year period in the previous 5 years as their
actual emissions baseline.

Industry submissions to EPA on the NSR program asserted that having to
assume maximum capacity biased the test and significantly overstated the
true emissions impact of a project. They cited cases where they expected a
project to reduce emissions, but the test showed that it would increase
them. For example, a refinery planned to implement a project that it
expected would improve the energy efficiency of a furnace by 5 percent.
While the furnace was permitted to emit 45 tons of air pollution per year,
it was not operating at full capacity and was therefore only emitting 35
tons per year. According to the facility*s submission, the project would

decrease annual emissions to 32 tons per year through more efficient fuel
combustion while running at the same capacity. However, the emissions test
required the refinery to compare its historical emissions* 35 tons per
year* with its future potential emissions running at full capacity* 45
tons per year. Because the facility was located in an area where a 10 ton
per year increase triggers NSR requirements, the facility would have had
to obtain an NSR permit for the modification.

Likewise, a pulp and paper mill planned to install an air flow system that
would allow its boiler to more efficiently burn natural gas, creating
annual savings of $1 million. By using less fuel, the project was also
expected to reduce future emissions of carbon monoxide, nitrogen oxides,
and volatile organic compounds. However, the facility had been operating
below its

maximum capacity during the 24 months preceding the planned installation.
Therefore, when managers compared the facility*s actual emissions during
this period to its future potential emissions, assuming it would operate
at maximum capacity after the modification, the projected emissions
increase qualified as a major modification, even though the

11 Forty tons per year is the threshold for emissions of nitrogen oxides,
sulfur dioxide, and volatile organic compounds in areas with good air
quality, but the level can be lower in areas with poorer air quality.

Page 21 GAO- 03- 947 Clean Air Act

project was expected to reduce emissions. If the company proceeded with
the project under NSR and installed the best available pollution controls,
it would incur $17 million in total costs, making the project cost-
prohibitive, according to the industry submission.

While these anecdotes assert that having to assume maximum capacity under
this test was unfair, the NSR program required facilities to assume that
the changed equipment would operate at the maximum level allowed in its
operating permit unless the owners or operators of the facility made a
legally binding commitment to operate at lower production levels,
according to EPA. Otherwise, EPA and state and local air quality agencies
would have no way of ensuring that companies would not have a

significant increase in emissions under the NSR rules after making a
physical change, according to EPA. In addition, EPA acknowledged in its

Report to the President that performing an energy efficiency project can
provide an economic incentive to increase production levels at more
efficient facilities, potentially resulting in increased emissions.
However, EPA*s new final rule now gives a company the option to compare a
facility*s previous emissions to its projected actual emissions, instead
of its maximum potential emissions. EPA believes that this new test will
remove disincentives that discourage facilities from making the types of
changes that improve operating efficiency, implement pollution prevention
projects, and result in other environmentally beneficial changes. In

addition, EPA asserts that the new record keeping and reporting measures
required with this option will provide the information necessary for
reviewing authorities to ensure that such changes are made consistent with
the Clean Air Act.

The anecdotes also identified a third barrier. According to industry,
unpredictable delays associated with the NSR process disrupt a project*s
planning and construction timetables, increasing project costs. Program
managers and technical directors representing forest product companies,
chemical manufacturers, refineries, and utilities commented on this issue,
stating that the NSR permitting process can last anywhere from 6 to 24
months, thereby delaying or disrupting project planning and construction.
Petroleum refinery representatives claimed to be uniquely affected by
this, asserting that NSR permitting delays made it difficult for them to
meet federal mandates and deadlines for producing cleaner- burning
gasoline, potentially subjecting them to fines and enforcement actions.
Industry officials also claimed that project delays they attributed to NSR
permitting resulted in equipment wearing out, increasing replacement
costs, and raising safety issues. They also stated that certain repair
projects required quick decisions and turnaround, both of which are not
compatible with the

Page 22 GAO- 03- 947 Clean Air Act

amount of time it takes to obtain an NSR permit and install the
appropriate pollution controls. An EPA NSR program manager confirmed that
permitting delays could hinder energy efficiency upgrades in emergency
situations where a company needs to quickly replace broken or worn out
machinery and would like to install more efficient equipment. However, the
program manager said such situations are rare and that the timelines for
planned facility upgrades are generally compatible with the NSR permitting
schedule. In addition to permitting delays, some of the anecdotes also
asserted that, if a project triggered NSR requirements, the costs of
installing pollution controls would have outweighed the anticipated
benefits of the project.

In an April 2003 report to the Congress on the NSR program, the National
Academy of Public Administration (NAPA) stated that while a lack of data
prevented the organization from determining the extent to which NSR has
impeded energy efficiency improvements, NSR might indeed have discouraged
some industrial sources from undertaking economically and environmentally
sound maintenance and energy efficiency projects. However, according to
NAPA, some facilities continue to operate without modern pollution
controls because of widespread noncompliance with NSR or flaws in its
implementation. NAPA said that these facilities have an advantage over
their competitors because they have not incurred the costs of controlling
emissions, and that such facilities have little basis to complain that NSR
has adversely affected the efficiency of their operations. An EPA manager
for the NSR program took exception to NAPA's findings, stating that EPA
and the states have enforced the program over the years. The official also
said that facilities only trigger NSR when they make physical changes in,
or changes in the method of operation of, their facilities that
significantly increase emissions. Therefore, if the facilities NAPA refers
to have not undertaken such modifications, they have complied with the
program, according to the official.

Several environmental groups experienced in NSR issues disagreed with
industry*s claims about the NSR program*s effects on energy efficiency. A
representative of one such group, who has testified before the Congress on
NSR, pointed out that if companies truly wanted to avoid NSR, they could
accept an emissions limit in their operating permit and make a formal
commitment to not emit above the NSR threshold. Companies that agree to
such limits may make any modifications they want without triggering NSR,
provided they do not exceed the limit. With this option available, the
representative asserted that the NSR process could not deter The National
Academy of

Public Administration and Environmental Stakeholders Reached Different
Conclusions about NSR*s Effects on Energy Efficiency

Page 23 GAO- 03- 947 Clean Air Act

energy efficiency projects that were expected to reduce emissions because
they would not trigger NSR, since they would not increase emissions above
the program threshold. For example, a source with an operating permit
limit could make a physical or operational change without an NSR permit
and increase emissions by 39.9 tons per year (where the threshold is 40
tons per year) and not trigger NSR. The environmental representative cited
other factors, such as poor economic conditions and a lack of

willingness to control air emissions, as the primary factors hindering
companies from pursuing these projects, rather than compliance with NSR.

Representatives of the chemical, forest products, and electric utility
industries disagreed with this position and said that companies are
reluctant to accept an operating permit limit because they would be giving
up their flexibility to increase production in response to changing
economic and market conditions, just for the sake of one energy efficiency
project. A trade association representing the chemical industry also told
us that the lengthy and protracted process for obtaining an operating
permit makes it impractical to renegotiate a permit limit every time a
company wants to undertake an energy efficiency project. As EPA notes in
its Report to the President, its conclusions about the

effect of NSR on energy efficiency projects are based on anecdotal
information because the agency lacked comprehensive data on the number of
projects that did not go forward as a result of NSR, according to EPA
program managers. Because EPA based its conclusions on anecdotes, the
agency*s findings do not necessarily represent NSR*s effect on energy

efficiency projects within the industries that provided the anecdotes or
across all industries subject to the program. Reaching such conclusions
about the program*s broader effects on energy efficiency projects would
have required gathering information from either a statistically valid
sample of companies subject to the program or a comprehensive survey of

affected industries. Conducting such an analysis, however, would have
required substantial resources.

In addition, EPA*s conclusion that the NSR program had discouraged some
energy efficiency investments that would have reduced emissions was based
on the assumption that the companies would not increase their production
after completing the energy efficiency project, according to an EPA
official responsible for the analysis. Under this assumption, 23 of the 69
anecdotes predicted that the proposed projects would decrease emissions,
11 predicted an increase, 2 predicted no change, and 33 (or 48 The
Anecdotes Do Not Necessarily Represent

NSR*s Effect on Energy Efficiency Projects Industrywide or Their Overall
Impact on Emissions

Page 24 GAO- 03- 947 Clean Air Act

percent) did not include sufficient data to determine the emissions
impact. 12 However, facilities* future production levels and air pollutant
emissions may fluctuate in response to changing economic conditions and
other factors. In addition, performing an energy efficiency project can
provide an economic incentive to increase production levels at more
efficient facilities, potentially increasing emissions and related health
risks.

The executive director of one industry trade association stated that it
would make economic sense to increase production at more efficient
facilities. The representative *could not imagine a utility spending money
on extra capacity, and then not utilizing it.* In addition, according to
EPA, production at more efficient facilities could supplant that at less
efficient and higher- emitting facilities. Therefore, even in cases where
an energy efficiency project was expected to reduce emissions, future
increases in

production after implementing a project could possibly increase emissions,
as well as related health risks, past the NSR threshold.

On the other hand, according to an EPA official responsible for the
agency*s energy efficiency analysis, the agency expected that if a company
increased production at its more efficient facilities, it could decrease
production at its less efficient facilities. Therefore, any emissions
increases due to higher production levels at more efficient facilities
would be offset by decreased production elsewhere. In addition, a
facility*s emissions could decrease more than expected after implementing
an energy efficiency project if the facility decreased production, for
example, due to poor economic conditions. An EPA program manager said that
the

agency has not analyzed the air pollution impacts of shifts in production
that facilities make after implementing energy efficiency projects.

While EPA determined that the final rule would lead to overall economic
and environmental benefits, these effects are uncertain because of limited
data and difficulty in determining how industrial companies will respond
to the rule. Consistent with the relevant executive order and guidance,
EPA conducted a limited analysis that determined that the rule was not
economically significant, and OMB concurred. In addition, EPA identified
data limitations that prevented the agency from conducting a more

12 The anecdotes did not always include sufficient information to
determine the magnitude of anticipated emissions changes after completing
an energy efficiency project. Conclusions

Page 25 GAO- 03- 947 Clean Air Act

quantitative analysis. Some stakeholders disagree with EPA about the
rule*s effects, contending that the rule will allow companies to increase
air pollutant emissions, resulting in adverse public health effects. If
these stakeholders are correct, the final rule could exacerbate existing
air pollution problems* 133 million Americans already live in areas with
air pollution levels above at least one of the health- based air quality
standards* and impose negative economic impacts that EPA did not account
for in its analysis of the rule. Because of the lack of data and
uncertainties about the NSR final rule*s

impacts, we recommend that the EPA Administrator:  determine what data
are available that the agency could use to monitor

the emissions impacts of the rule,  work with state and local air quality
agencies to identify any additional

data needs and possible ways to fill them, and  use the monitoring
results to determine whether the rule has created

adverse effects that the agency needs to address. We provided EPA and OMB
with a draft of this report for review and comment. We subsequently
received comments from both agencies. EPA said that the report was
accurate and that the agency agreed with our conclusions and
recommendation. OMB said that the report was accurate but raised several
questions about EPA*s potential implementation of the

recommendation. Specifically, OMB said it would be difficult for EPA to
gather data to monitor the rule*s effects and that attributing emissions
changes to the final rule would pose challenges. For example, OMB said
that it would be difficult to determine the level of emissions facilities
would have released in the absence of the final rule. EPA, however, said
that the recommendation acknowledges these challenges and provides
flexibility for the agency to work with state and local agencies to
identify data collection strategies. EPA and OMB also recommended a number
of technical changes to the report, which we have incorporated into this
report as appropriate.

As agreed with your offices, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 10 days from
the report date. At that time, we will send copies to the EPA
Administrator, Recommendations for

Executive Action Agency Comments

Page 26 GAO- 03- 947 Clean Air Act

the Director, Office of Management and Budget, interested congressional
committees, and other interested parties. We will also make copies
available to others upon request. In addition, this report will be
available at no charge on GAO*s Web site at http:// www. gao. gov.

If you or your staffs have any questions, please call me at (202) 512-
3841. I can also be reached at StephensonJ@ gao. gov. Key contributors to
this report are listed in appendix II.

John B. Stephenson Director, Natural Resources and Environment

Page 27 GAO- 03- 947 Clean Air Act

The Ranking Minority Member of the Senate Environment and Public Works
Committee and Senator Lieberman asked us to determine the basis of (1)
EPA*s analysis of the economic impacts of the final rule and its
conclusion that the rule would not create significant enough benefits or
costs to merit a more detailed analysis and (2) EPA*s conclusions that the
NSR program (prior to the final rule) discouraged some energy efficiency
projects.

To respond to the first objective, we used OMB and EPA guidance to review
EPA*s screening analysis of the final rule*s economic impacts. EPA issued
its guidance to ensure that its economic analyses comply with the Unfunded
Mandates Reform Act of 1995 and Executive Order 12866. These directives
require federal agencies to analyze the economic effects of rules that may
impose annual costs on the government or private entities of more than
$100 million. In addition to reviewing EPA, OMB, and other federal
requirements and guidance on conducting economic analyses of proposed
rules, we reviewed two additional EPA analyses of the final rule*s
effects, as well as documents from environmental groups, industry trade
associations, and other interested parties related to the costs and
benefits of the final rule. Furthermore, we met with the NSR program
manager within EPA and other senior officials within the agency*s Office
of Air Quality Planning and Standards. We also met with senior OMB staff
responsible for reviewing EPA*s analysis within the Office of Information
and Regulatory Affairs.

To respond to the second objective, we reviewed information provided to
EPA by outside organizations that the agency relied on in preparing its
findings on the NSR program*s effects on energy efficiency projects at
industrial companies. This information consisted of written comments,
case- specific examples of foregone energy efficiency projects, and
supplemental documentation submitted to EPA by individual firms, industry
trade associations, and environmental advocacy groups during EPA*s public
comment period. We obtained documents through the EPA public docket A-
2001- 19, and directly from those companies and

organizations who submitted information to EPA. We also obtained documents
or conducted interviews with parties identified by EPA as the primary data
sources underlying its energy efficiency findings, including
representatives of the American Chemistry Council, the American Forest and
Paper Association, the American Petroleum Institute, the American Public
Power Association, British Petroleum, Detroit Edison, Duke Energy, the
Edison Electric Institute, Excel Energy, Exxon Mobil, the National Coal
Council, and the National Petrochemical & Refiners Association. Once we
obtained anecdotes provided by the parties EPA Appendix I: Objectives,
Scope, and Appendix I: Objectives, Scope, and

Methodology

Appendix I: Objectives, Scope, and Methodology

Page 28 GAO- 03- 947 Clean Air Act

identified as its primary data sources, we removed those that did not
pertain to the NSR program*s effects on energy efficiency as well as those
that did not provide sufficient information for analysis. We then reviewed
the 69 remaining anecdotes to assess their anticipated effects on
emissions. We did not independently verify the factual basis of the
anecdotes. We also spoke with the Natural Resources Defense Council, a
national environmental advocacy organization that disputed EPA findings.
Finally, we spoke with the NSR program manager within EPA and another
official within the agency*s Office of Air Quality Planning and Standards
who were responsible for the Report to the President.

We conducted our work between August 2002 and August 2003 in accordance
with generally accepted government auditing standards.

Appendix II: GAO Contacts and Staff Acknowledgments

Page 29 GAO- 03- 947 Clean Air Act

John B. Stephenson (202) 512- 3841 Eileen R. Larence (202) 512- 6510

In addition to the individuals named above Tim Guinane, David Hancock, and
Michael Hix made key contributions to this report. Nancy Crothers, Karen
Keegan, Jeffrey Larson, Judy Pagano, Lisa Turner, and Laura Yannayon also
made important contributions. Appendix II: GAO Contacts and Staff

Acknowledgments GAO Contacts Acknowledgments

(360258)

The General Accounting Office, the audit, evaluation and investigative arm
of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of
the federal government for the American people. GAO examines the use of
public funds; evaluates federal programs and policies; and provides
analyses, recommendations, and other assistance to help Congress make
informed oversight, policy, and funding decisions. GAO*s commitment to
good government is reflected in its core values of accountability,
integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO*s Web site (www. gao. gov) contains abstracts
and fulltext files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as *Today*s Reports,* on its
Web site daily. The list contains links to the full- text document files.
To have GAO e- mail

this list to you every afternoon, go to www. gao. gov and select
*Subscribe to e- mail alerts* under the *Order GAO Products* heading.

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to: U. S. General Accounting Office 441 G Street NW, Room LM
Washington, D. C. 20548 To order by Phone: Voice: (202) 512- 6000

TDD: (202) 512- 2537 Fax: (202) 512- 6061

Contact: Web site: www. gao. gov/ fraudnet/ fraudnet. htm E- mail:
fraudnet@ gao. gov Automated answering system: (800) 424- 5454 or (202)
512- 7470 Jeff Nelligan, Managing Director, NelliganJ@ gao. gov (202) 512-
4800

U. S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.
C. 20548 GAO*s Mission Obtaining Copies of

GAO Reports and Testimony

Order by Mail or Phone To Report Fraud, Waste, and Abuse in Federal
Programs Public Affairs
*** End of document. ***