Clean Air Act: Key Stakeholders' Views on Revisions to the New	 
Source Review Program (02-FEB-04, GAO-04-274).			 
                                                                 
Environmental Protection Agency (EPA) revisions to the New Source
Review (NSR) program to control industrial emissions have drawn  
attention from state and local agencies that implement the	 
program, as well as industry and environmental and health groups.
Under the revisions, companies may not have to install pollution 
controls when making some facility changes. GAO was asked to	 
obtain the opinions of state air quality officials and other	 
stakeholders on the impact of both the final and proposed	 
revisions EPA issued in December 2002. GAO obtained survey	 
responses from NSR program managers in 44 states and certain	 
localities and contacted six environmental and health groups, and
eight industry groups active in the NSR debate. 		 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-274 					        
    ACCNO:   A09201						        
  TITLE:     Clean Air Act: Key Stakeholders' Views on Revisions to   
the New Source Review Program					 
     DATE:   02/02/2004 
  SUBJECT:   Air pollution					 
	     Air pollution control				 
	     Environment evaluation				 
	     Environmental law					 
	     Environmental monitoring				 
	     Environmental policies				 
	     Public health legislation				 
	     Strategic planning 				 
	     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-04-274

United States General Accounting Office

GAO

                       Report to Congressional Requesters

February 2004

CLEAN AIR ACT

     Key Stakeholders' Views on Revisions to the New Source Review Program

GAO-04-274

Highlights of GAO-04-274, a report to congressional requesters

Environmental Protection Agency (EPA) revisions to the New Source Review
(NSR) program to control industrial emissions have drawn attention from
state and local agencies that implement the program, as well as industry
and environmental and health groups. Under the revisions, companies may
not have to install pollution controls when making some facility changes.
GAO was asked to obtain the opinions of state air quality officials and
other stakeholders on the impact of both the final and proposed revisions
EPA issued in December 2002. GAO obtained survey responses from NSR
program managers in 44 states and certain localities and contacted six
environmental and health groups, and eight industry groups active in the
NSR debate.

GAO recommends that EPA (1) help state air quality agencies implement the
revisions, (2) monitor the effects of the rule that excludes routine
equipment replacements from NSR, and (3) consider stakeholders' concerns
before excluding other activities from NSR. In commenting on the report,
EPA's Assistant Administrator for Air and Radiation said that the agency
has concerns about our methodology and certain of our findings. GAO
believes its approach and presentation are appropriate. Moreover, EPA said
that our recommendations make sense, and that the agency already plans to
take these actions.

www.gao.gov/cgi-bin/getrpt?GAO-04-274.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact John Stephenson at (202)
512-3841 or [email protected].

February 2004

CLEAN AIR ACT

Key Stakeholders' Views on Revisions to the New Source Review Program

A majority (29 of 44) of the state officials responding to GAO's survey
expected the rule EPA finalized in December 2002 to provide industry with
greater flexibility to make some facility changes without having to obtain
NSR permits or, in some cases, install pollution controls. However, in
their opinion, 27 officials expected the rule to increase emissions of
harmful air pollutants, thereby hindering areas' efforts to meet air
quality standards and potentially creating or exacerbating public health
risks. This concern contrasts with EPA's assessment that the rule will
decrease emissions and maintain the current level of environmental
protection. Furthermore, 30 of the officials expected their agency's
workload would increase as they adopt and implement the rule into their
own programs. Almost all of the 44 officials would like EPA assistance
with implementation.

Similarly, 28 of the 42 officials responding expected the two NSR
revisions as proposed in December 2002-intended to provide more certainty
about when facility changes are considered routine maintenance, repair,
and replacement activities and can be excluded from NSR requirements-to
decrease the number of permits companies would have to obtain, thereby
giving them the flexibility to make some changes without installing
controls. However, 21 and 26 officials, respectively, thought that the two
exclusions would increase emissions; only relatively few thought the
exclusions would decrease emissions as EPA's analysis had predicted. About
a third of the officials thought the exclusions would exacerbate air
quality problems in areas that do not meet standards, but fewer officials
thought the exclusions would cause problems in areas that currently meet
standards. Finally, 27 thought that implementing the two exclusions would
increase states' administrative burden.

The other stakeholder groups GAO contacted agreed that the final rule and
two exclusions would decrease the regulatory burden on companies that
modify their facilities, but disagreed about the impact on emissions and
air quality agencies' workload. The six environmental and public health
officials expected that because companies would not have to obtain as many
NSR permits or install as many controls when modifying facilities,
emissions would rise and state and local agencies' workloads increase as
agencies sought alternative ways to meet standards. In contrast, the eight
industry officials expected the revisions to encourage companies to invest
in energyefficient projects they had avoided under the prior program,
which the officials believed would lower fuel use and emissions. The
officials also expected that fewer permits would lead to decreases in
agencies' workloads.

Determining the revisions' likely impacts is difficult because, as
discussed in GAO's August 2003 report on EPA's analytical basis for the
final rule (GAO03-947), little data exist to confirm stakeholders'
opinions. In that report, GAO recommended that EPA work with state and
local agencies to obtain data to assess the rule's emissions impact and
correct any adverse effects.

Contents

  Letter

Results in Brief
Background
A Majority of the State Officials Expect the December 2002 Rule

Will Provide Industry with Greater Operating Flexibility but Also
Increase Emissions and Agencies' Workload

At Least Half of the State Officials Expected the Proposed
Revisions Defining NSR Exclusions to Provide Industry Greater
Flexibility but Also Increase Emissions and the Administrative
Workload for State Agencies

Other Stakeholders Expected the Proposed and Final NSR
Revisions to Benefit Industry but Disagreed on Their Effect on
Emissions and Air Quality Agencies' Workload

Conclusions
Recommendations for Executive Action
Agency Comments and Our Evaluation

                                       1

                                      6 9

13

19

23 27 28 29

  Appendix I Summary of Local Agency Officials' Responses 32

Impacts on Industry 32
Emissions Impacts 32
Impacts on Local Agencies 33
Impacts on Industry 33
Emissions Impacts 33
Impacts on Agencies 34

Appendix II Objectives, Scope, and Methodology

Appendix III 	Comments from the Environmental Protection
Agency 41

GAO Comments 46

  Appendix IV GAO Contacts and Staff Acknowledgments 51

GAO Contacts 51
Staff Acknowledgments 51

  Tables

Table 1: NSR Revisions Included in the December 2002 Final Rule 11 Table
2: Anticipated Emissions Effects of the December 2002 Final Rule
Provisions (number of state officials' responses) 15 Table 3: Ways in
Which Officials Expect Their States to Adopt the

Final Rule 18 Table 4: States with Local Air Agencies 35 Table 5: Survey
Respondents Listed by EPA Region 37

Abbreviations

ALAPCO Association of Local Air Pollution Control Officials
DOE Department of Energy
EPA Environmental Protection Agency
NSR New Source Review
PSD Prevention of Significant Deterioration
STAPPA State and Territorial Air Pollution Program Administrators

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.

United States General Accounting Office Washington, DC 20548

February 2, 2004

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

The Environmental Protection Agency's (EPA) revisions to the Clean Air
Act's New Source Review (NSR) Program-one of the act's mechanisms
for maintaining air quality to protect public health-have provoked
controversy. These revisions are contained in rules that the agency issued
in December 2002 and October 2003, respectively. In general, these rules
provide companies with regulatory flexibilities to modify their industrial
facilities without triggering NSR requirements to install potentially
costly
pollution controls, if certain conditions are met. According to EPA, the
December 2002 rule will provide incentives for facilities to reduce
emissions, remove barriers to energy efficiency and pollution control
projects, and offer facilities greater regulatory flexibility, while the
October 2003 rule will allow companies to modernize facility operations in
ways that will maintain and improve safety, reliability, and efficiency.
EPA
also anticipates that the rules will enhance the NSR program's
environmental benefits.

Reactions to the rules have differed considerably. A number of industry
groups agree with EPA's position, and some states have filed legal
documents in court expressing support for the rules. Other states and
some localities, including a coalition of states and various localities
primarily located in the mid-Atlantic and northeast regions of the
country,1
as well as certain environmental groups, have filed lawsuits challenging
the legality of the two rules in court.

1The District of Columbia is included in this coalition.

The NSR program-established in 1977-seeks to protect public health,
maintain compliance with air quality standards, and enhance air quality in
national parks and scenic areas. The NSR program applies to nearly 17,000
industrial facilities, including fossil-fueled power plants, petroleum
refineries, and facilities that manufacture automobiles, chemicals,
pharmaceuticals, and paper. The program requires companies, when they are
constructing facilities, to obtain NSR permits that limit the amount of
pollution that facilities may emit and to install pollution controls when
necessary. The program imposes similar requirements when a company makes a
physical or operational change to an existing facility-such as adding new
production equipment-if the change would result in a significant net
increase in emissions.2 The Congress allowed existing facilities to defer
installation of pollution controls until a major modification was made
with the expectation that, over time, all facilities would install such
equipment, and this would lead to lower overall emissions.

Responsibility for implementing NSR, as well as other air quality
regulations, generally rests with state and local air quality agencies.
However, the stringency of the air quality regulations they set varies.
The Clean Air Act generally requires more stringent control measures for
industries located in areas that fail to meet at least one of the air
quality standards than for those located in areas that meet the standards.

Because of the NSR program's complexity and the administrative burden it
imposes, EPA has long recognized a need to revise it. In 1992, EPA began a
reform process that resulted in proposed changes to the program in 1996
and 1998. But the agency did not finalize the proposals as rules during
the previous administration. The current administration acted on certain
of the prior reform proposals by issuing a final rule in December 2002.3
This rule contained five provisions-including a new method for determining
whether a facility change will significantly increase net emissions-that
reduced the likelihood that certain of these changes would require an NSR
permit or, in some cases, the installation of pollution controls. In
assessing

2Such changes are called "major modifications" and the level of emissions
that will trigger the NSR requirements, known as the threshold, varies by
pollutant and the air quality status of the area in which a facility is
located.

3Prevention of Significant Deterioration (PSD) and Nonattainment New
Source Review (NSR): Baseline Emissions Determination, Actual-to-Future
Actual Methodology, Plantwide Applicability Limitations, Clean Units,
Pollution Control Projects, 67 Fed. Reg. 80186 (2002) (to be codified at
40 C.F.R. pts, 51 and 52).

the potential costs and benefits of the rule to determine whether to
pursue it, EPA anticipated that this rule will provide incentives for
facilities to reduce emissions, remove barriers to investments in energy
efficiency and pollution control projects, and offer facilities greater
regulatory flexibility. In addition, at the time the agency issued the
rule, it released an analysis of the rule's anticipated environmental
effects that concluded the rule would lead to an overall environmental
benefit.4 EPA estimated it will cost state and local agencies about $6.5
million annually to incorporate the rule into their air pollution control
plans. However, the agency expects that the rule will ultimately decrease
agencies' NSR costs after the first 3 years of implementation.

Also in December 2002, EPA issued a proposed rule, with two provisions,
that would define certain activities as routine maintenance, repair, or
replacement and, therefore, exempt from NSR requirements.5 In its
assessment of the rule, the agency asserted that the two provisions would
provide greater certainty to industry about when facility changes can be
exempt from NSR and encourage facilities to perform energy-efficiency
projects that were being hindered by the existing program's requirements
for permits and costly controls. The cost of installing controls varies
but, in extreme cases, costs can reach hundreds of millions of dollars,
according to EPA.

One of the provisions that EPA proposed would exclude activities from NSR
requirements considered "routine equipment replacements"- replacements of
worn-out or broken machinery with identical parts or those that perform
the same function as the existing part. EPA proposed several thresholds
below which expenditures for such equipment replacement could be
considered routine and exempt from NSR requirements and solicited public
comment on them.6 After reviewing the comments it received, EPA issued
this exclusion as a final rule in October 2003. EPA established 20 percent
of the cost to replace the entire process

4See U.S. Environmental Protection Agency, Supplemental Analysis of the
Environmental Impact of the 2002 Final NSR Improvement Rules (Washington,
D.C.: Nov. 21, 2002).

5Prevention of Significant Deterioration (PSD) and Nonattainment New
Source Review: Routine Maintenance, Repair and Replacement, 67 Fed. Reg.
80290 (2002).

6The cost of the replacement equipment had to be below a certain
percentage of the cost to replace the "process unit." A process unit for
power plants is defined as the portion of a plant that directly
contributes to electricity production (power plants can have more than one
such unit). The replacement equipment also had to meet certain criteria,
such as maintaining the basic design of the original unit.

unit-for example, a steam-generating unit in a power plant-as the cost
threshold companies could use to replace parts within that unit without
being subject to NSR, and concluded that the rule would have insignificant
environmental effects.7 EPA also estimated that the rule would impose
one-time costs of $8.7 million on industry and $1.7 million on state and
local agencies that adopt the rule, while saving the electric utility
industry hundreds of millions of dollars.8

The other provision that EPA proposed would create an "annual maintenance
allowance" exclusion that would enable companies to avoid NSR if the cost
of all routine maintenance and repair activities did not exceed a certain
percentage of the cost to replace the entire facility. The agency has not
determined whether it will finalize this portion of the proposal or pursue
other options to address routine maintenance and repair activities.

You asked us to address a number of questions about the basis of the
revisions and their potential impacts. In two previous reports on the
revisions, we reviewed (1) EPA's assessment of the economic and
environmental impact resulting from the December 2002 final rule, and (2)
the potential impact of the NSR revisions on the enforcement actions that
EPA had filed against coal-fired power plants for allegedly violating NSR
requirements and on public access to information on emissions. We
presented our findings on these reviews in reports issued on August 22,
2003, and October 21, 2003, respectively.9 In the August report, we
determined that data limitations precluded EPA from performing a
quantitative analysis of the effects of the December 2002 final rule. In
the

7Prevention of Significant Deterioration (PSD) and Nonattainment New
Source Review (NSR): Equipment Replacement Provision of the Routine
Maintenance, Repair and Replacement Exclusion, 68 Fed. Reg. 61248 (Oct.
27, 2003) (to be codified at 40 C.F.R. pts, 51 and 52).

8EPA based this statement on an analysis which concluded that revising the
routine maintenance, repair, and replacement exemption would lead to the
increased availability of existing power plants, thereby reducing the need
for new plants. See U.S. Environmental Protection Agency, Regulatory
Impact Analysis for the Specification of Categories of Activities as
Routine Maintenance, Repair, and Replacement for the New Source Review
Program (Washington, D.C.: August 2003).

9See U.S. General Accounting Office, Clean Air Act: EPA Should Use
Available Data to Monitor the Effects of Its Revisions to the New Source
Review Program, GAO-03-947 (Washington, D.C.: Aug. 22, 2003) and U.S.
General Accounting Office, Clean Air Act: New Source Review Revisions
Could Affect Utility Enforcement Cases and Public Access to Emissions
Data, GAO-04-58 (Washington, D.C.: Oct. 21, 2003).

October report, we determined that the revisions could affect the ongoing
cases and the public's access to emissions information, although EPA
program managers did not agree that the rule would affect access to
emissions information.

You also asked us to obtain the views of a number of key stakeholders
about a broader range of the revisions' potential impacts. More
specifically, you asked us to obtain (1) state air quality agency
officials' views about the impacts of the December 2002 final NSR rule on
industry, emissions, and agencies' workloads; (2) state air quality agency
officials' views about the impacts of the two December 2002 proposed NSR
exclusions on industry, emissions, and agencies workloads; and (3)
environmental, health, and industry organizations' views on the impacts of
all the NSR revisions. In addition, we determined selected local air
quality agencies' views on the revisions' potential effects.

To address the first two objectives, we administered a detailed survey to
the NSR program managers in 50 states and the District of Columbia using
the Internet. We surveyed program managers to ensure that we obtained
information from those most involved in the day-to-day administration of
the NSR program. In addition, we sent this survey to 71 local agencies,
primarily those with their own authority to issue NSR permits. To identify
the NSR program manager for each state or local agency, we worked with the
10 EPA regional offices and obtained some information from the State and
Territorial Air Pollution Program Administrators (STAPPA) and the
Association of Local Air Pollution Control Officials (ALAPCO) Internet
site. We pretested the survey with selected state and local program
managers to ensure the questions were clear, understandable, accurate, and
comprehensive. In addition, to ensure that the questions were neutral and
objective, we pretested the survey with states that were supportive of the
revisions, as well as states that were not. We received complete responses
from 44 state program managers and 60 local agency officials (each state
or local agency could only provide one response). The managers in four
states said they declined to respond so as not to disclose information
related to their state's ongoing NSR-related litigation. We have provided
a copy of our survey and detailed tables showing the state and local
officials' responses to the questions in a separate report, Survey of
State and Local Air Quality Officials Opinions on the Impacts of the
Environmental Protection Agency's Revisions to the Clean Air Act's New
Source Review Program (GAO-04-337SP), available on the Internet at
http://www.gao.gov/special.pubs/gao-04-337sp. We have also summarized the
main results of our survey of select local air quality agencies in
appendix I.

It is important to note that we asked the program mangers for their
opinions about the potential impacts, both positive and negative, of the
NSR revisions. Based on our prior work, we had established, and EPA
program managers told us, that very little data existed-either on the
impact of the NSR program before the revisions, or on the number of
facilities that might use any or all of the revisions-to try to assess the
revisions' impacts. In addition, at the time we asked the state and local
officials for their opinions about the exclusion of certain activities as
routine equipment replacement, EPA had not defined a specific cost
threshold for this exclusion as it did in the October 2003 rule. As a
result, we wanted to confirm that the opinions the officials stated about
the exclusion in response to our survey were still accurate. Therefore, we
provided a summary of our survey results regarding this provision to
STAPPA/ALAPCO. Working with its members, the association confirmed that
states and localities continue to have the same views about this exclusion
as they did at the time of our survey. Furthermore, we confirmed with the
EPA NSR program manager that he did not think the opinions of the state
and local officials had changed as a result of EPA finalizing the
exclusion in its October 2003 rule.

To address the third objective, we identified key stakeholders involved in
NSR policy decisions at the national level-including representatives of
industry, environmental, and public health interests-and sent them a more
general list of questions via electronic mail that solicited their
responses about the revisions' potential effects. For a more detailed
discussion of our scope and methodology, see appendix II.

A majority of the 44 state air quality officials responding to our survey
believes that the December 2002 final rule will provide industry greater
flexibility to modify facilities without having to install pollution
controls in some cases; a majority of the officials also think, however,
this flexibility will come at the cost of increases in emissions and
agencies' workloads. Regarding the impact of the rule on industry, 29 of
the 44 officials said that, in their professional opinion, the rule allows
companies to make more modifications without having to obtain permits,
which can trigger requirements for controls. The permitting process and
its requirements to install controls, however, was one of the best
features of the NSR program prior to the revisions, according to 40 of the
officials. Regarding the rule's impact on emissions, 27 of the 44
officials believe the rule will increase emissions; only 5 believe it will
decrease them (the remaining officials thought they would remain the same
or were unsure). Many officials expect that these potential emissions
increases will affect their state's

  Results in Brief

ability to meet the national health-based air quality standards: 13
officials said their state would have difficulty meeting standards, and 14
said they would look to other pollution control programs to try to offset
the anticipated increases. Only 7 said their state would not have
difficulty meeting standards. (Nine said they could not judge the rule's
effects). Finally, 30 of the officials predicted that implementing the
rule, such as educating industry and their own staff on its provisions,
would increase their agency's workload at a time when many state agencies'
resources are constrained. To implement the rule, almost all of the
officials said they would like EPA's assistance and we are recommending
that EPA provide states with this help.

Overall, the state officials had similar opinions about the impact of the
two provisions excluding some facility activities from NSR requirements if
they are considered routine maintenance, repair, or replacement. Regarding
the exclusions' impact on industry, 28 of the 42 officials responding to
this question said that both provisions would allow companies to make more
facility changes without having to obtain permits. On the other hand, 21
of the 42 officials thought the exclusion for routine equipment
replacement would increase emissions, and 26 thought the proposed annual
maintenance allowance exclusion would have this impact. Relatively few
officials thought the exclusions would decrease emissions. In general, the
officials were less concerned about the impact of the exclusions on
states' ability to address air quality problems or meet standards than
they were about the final rule's impact. More specifically, about a third
of the officials thought the increased emissions under the exclusions
would worsen air quality problems in areas that already did not meet
standards. Moreover, only 5 officials thought the exclusion for routine
equipment replacement would cause problems in areas that currently meet
standards, and only 7 thought the proposed annual maintenance allowance
exclusion would do so. Regarding the impact on agency workloads, 27 of the
44 officials responding thought that implementing the exclusions would
increase their administrative burden. Overall, 21 officials opposed the
exclusion for routine replacement of equipment; 32 opposed the proposed
annual maintenance allowance exclusion; and, relatively few supported
either provision (the remaining officials said they neither supported nor
opposed it, or they had no opinion).

The other stakeholders we contacted-representatives of key industry,
environmental, and public health interests who have been most active in
the NSR debate at the national level-also believed the revisions in the
final rule and the two proposed exclusions would decrease the regulatory
burden on industry. But the stakeholders differed in their opinions about

these revisions' impacts on emissions and agencies' workloads.
Representatives from the six environmental and public health groups we
contacted believed the revisions allow companies to make more
modifications without having to obtain an NSR permit or install controls,
and that emissions and public health risks would increase as a result.
These representatives also believed the changes would create more work for
state and local air quality agencies, such as their having to find other
ways to reduce emissions to meet air quality standards. On the other hand,
the eight industry representatives we contacted expected the revisions to
decrease emissions and health risks because they believe, consistent with
EPA's analysis, that companies will be more likely to pursue energy
efficiency projects they had postponed because of the prior NSR
requirements. With these projects, facilities could operate more
efficiently, burning less fuel and creating fewer emissions, according to
the representatives. As we noted in our August 2003 report, however, other
industry and environmental officials believed that if facilities can
operate more efficiently, they will increase their production and,
therefore, emissions overall. Finally, the industry representatives we
contacted believed that the revisions would lighten air quality agencies'
workloads. The NSR program will be clearer and simpler to implement and
enforce, and agencies will have fewer permits to issue, according to some
of these representatives.

Determining the likely impact of the revisions, given the conflicting
opinions of state officials and stakeholders, is difficult primarily
because little data exist to substantiate opinions. For example, one of
the stakeholders we contacted cited an EPA analysis of the equipment
replacement rule as support for his position, and another cited a
Department of Energy (DOE) analysis. However, neither analysis was a
comprehensive assessment of the revisions' effects. Furthermore, as we
noted in our August 2003 report, the overall economic and environmental
effects of the December 2002 rule are uncertain because of data
limitations and difficulty determining how companies will respond to the
rule. Therefore, we recommended in that report that EPA work with state
and local air quality agencies to obtain the data needed to monitor the
rule's emissions impacts and address any adverse effects. In this light,
we are making additional recommendations in this report: that EPA (1)
identify available data, or ways to obtain it, to monitor the emissions
impact of the NSR exclusion for routine equipment replacement (which EPA
issued as a final rule in October 2003) and (2) consider the state
officials' and stakeholders' concerns about emissions and workload impacts
that we identified before issuing a final rule on the proposed annual
maintenance allowance.

In commenting on the report, EPA's Assistant Administrator for Air and
Radiation said that the agency has concerns about our methodology and
certain of our findings. Nevertheless, EPA said that our recommendations,
on their face, make sense and that the agency already has plans to take
these actions. Specifically, EPA asserted that GAO (1) used the opinions
expressed in the survey responses as fact, and to draw conclusions and
make recommendations about the NSR program, (2) did not assure balance and
objectivity, (3) used a skewed survey sample, and (4) should have
evaluated whether the survey results were consistent with the facts cited
in EPA's analyses of the revisions' effects. GAO disagrees with each of
these assertions. First, GAO solicited opinions and carefully presented
them as such because we found, and EPA acknowledged, the emissions data
available to analyze the NSR revisions' impacts are so limited. We also
found that the state program managers' informed opinions raised
substantial concerns about the revisions' impacts. Our recommendations are
intended to address these legitimate concerns. Second, in designing our
survey, we took numerous steps to minimize bias, including asking
respondents' about both the positive and negative effects of the
revisions, conducting several pretests of the survey, and having a survey
specialist independent of its design review the survey to ensure that the
questions were not biased. Third, GAO surveyed the universe of state
program managers. These officials are on the front lines of program
implementation and are in the most informed position to weigh in on the
implementation impact questions we asked. Because of the large number of
other affected stakeholders, it was not feasible to survey the universe.
Instead, we surveyed 30 representative organizations, chosen because they
are involved in national NSR policy decisions and represent diverse
environmental, health, and industry perspectives. Finally, we did not use
the results of EPA's analyses as a benchmark to evaluate the survey
responses because our previous and current work has identified numerous
limitations with those analyses. EPA's written comments and our detailed
response are included as appendix III.

Background 	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. These include industrial facilities and
mobile sources, such as automobiles and other transportation. EPA has
issued air quality standards for six primary pollutants-carbon monoxide,
lead,

nitrogen oxides, ozone,10 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 death. In 2002, the most recent
year for which data were available, 146 million Americans lived in areas
that failed to meet at least one air quality standard, according to EPA.11

Subject to EPA's oversight, state and local air quality agencies generally
administer the NSR program and operate under one of two arrangements.
First, some agencies located in areas that meet air quality standards have
"delegation" agreements with EPA under which they implement the NSR
program contained in EPA's regulations. Under the second arrangement,
agencies design their own programs by incorporating all of their air
quality regulations, including federal requirements, into overall air
quality plans, known as state implementation plans. They update these
plans periodically and submit them to EPA for approval.

In addition, the Clean Air Act requires those agencies that implement
their own air quality programs to ensure their requirements are at least
as stringent as EPA's regulations. State and local agencies may also
supplement the federal NSR program with additional requirements. However,
some jurisdictions have laws or policies that prevent agencies from
implementing more stringent regulations.

Throughout its history, the NSR program has been characterized by
complexity and controversy, involving disputes between EPA and industry
about, among other things, whether certain facility changes qualified for
the routine maintenance, repair, and replacement exclusion. In recent
years, EPA has taken enforcement action against companies in several
industries, including some electricity producers, forest product
manufacturers, and petroleum refineries, alleging noncompliance with the
program.

10Ozone forms when nitrogen oxides react with volatile organic compounds
in the presence of heat and sunlight.

11This information, provided by EPA, is used for background purposes only.
We did not, therefore, assess its reliability.

In addition to concerns about enforcement related issues, some industry
representatives have also raised concerns that the time required to obtain
a NSR permit and the cost of installing controls have prevented facilities
from making changes that enhance energy efficiency and reduce air
emissions, such as modifying a boiler so that it produces the same amount
of energy with less fuel. In May 2001, the Vice President's National
Energy Policy Development Group recommended, among other things, that the
Administrator of the EPA, in consultation with the Secretary of Energy and
other federal agencies, examine the impact of the NSR program on
investments in new utility and refinery generation capacity, on energy
efficiency, and on environmental protection. In its June 2002 NSR Report
to the President, EPA concluded, among other things, that the program 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. This report also
contained recommendations for revising the program.

Subsequently, EPA issued a final rule on December 31, 2002, which
contained five provisions, identified in table 1, which exempt certain
facility changes from requirements to obtain NSR permits.12

        Table 1: NSR Revisions Included in the December 2002 Final Rule

Provision Final rule requirements

Revised method for calculating "baseline" Changes the method for computing
a piece of equipment's baseline emissions from the

emissions 	most recent 24-month period-or any other period more
representative of normal operations-to any 24-month period in the past 10
years adjusted for any new emissions limits added since the baseline
period. No changes were made to the baseline period for electric
utilities.

Revised test for calculating emissions Allows a facility to calculate
expected emissions after a facility change based on its

changes 	projection of future operation, rather than at full capacity.
This provision extended to all other industries the same type of
methodology for calculating expected emissions that EPA had granted to the
utility sector in the early 1990s.

Clean unit 	Excludes production equipment with state-of-the-art pollution
controls from NSR requirements for up to 10 years after installation
provided the unit will still meet the physical or operational
characteristics that formed the basis for the clean unit designation.

12The final rule's provisions were based on proposals considered by the
previous administration but never finalized. EPA solicited public comment
on those proposals in 1996 and again in 1998.

                       Provision Final rule requirements

Pollution control projects 	Exempts pollution prevention and control
projects from NSR if they are on EPA's list of "environmentally
beneficial" projects or on a case-specific basis if a nonlisted project is
determined to be environmentally beneficial. It also must be shown that
the project will not cause or contribute to a violation of federal air
quality standards or adversely impact air quality-related values (such as
visibility) for a national park.

Plantwide emissions limits 	Allows facilities to set a single emissions
limit (per pollutant) for an entire plant and then make changes within the
facility without triggering NSR, provided they do not exceed the limit.

Source: EPA.

These revisions have been the subject of congressional debate. For
example, in 2002, the Congress held hearings during which members of the
Congress, EPA officials, and a number of stakeholders-including industry,
states, and environmental groups-presented their positions on the
revisions. Also, legislation has been introduced in the Congress that
seeks to further regulate emissions from industrial facilities.13

In addition, a number of environmental and public health groups, as well
as a group of states primarily from the Mid-Atlantic and Northeast,
claimed that the December 2002 final rule violated the Clean Air Act and
asked EPA to reconsider several aspects of the rule. In July 2003, EPA
agreed to do so and then solicited public comment on the areas under
reconsideration. Based on this input, EPA announced at the end of October
2003 that it would make several technical changes to the rule. State and
local agencies that operate under delegation agreements were required to
have implemented the December 2002 rule by March 2003 or return
responsibility for implementing the rule to EPA, while those operating
under state implementation plans have until January 2006 to revise their
regulations accordingly.14

As for the December 2002 proposed provisions-that would further specify
what facility changes are exempt from NSR requirements under the routine
maintenance, repair, or replacement exclusions-a coalition of primarily
Mid-Atlantic and Northeastern states and environmental and public health
groups challenged the legality of the equipment replacement rule in court
after it was finalized in October 2003. State and local agencies

13Clear Skies Act of 2003 (S. 485 and H.R. 999); Clean Power Act of 2003
(S. 366); Clean Air Planning Act of 2003 (S. 843); Omnibus Mercury
Emission Reduction Act of 2003 (S. 484).

14Instead of meeting the March 2003 deadline, state and local agencies
could elect to have EPA take over the administration of portions of their
NSR programs affected by the revisions. Several agencies have chosen this
option.

that operate under delegation agreements were required to implement this
rule by December 26, 2003 or have EPA implement it for them, while those
operating under state implementation plans have until October 2006 to
revise their regulations accordingly. However, on December 24, 2003, the
U.S. Court of Appeals for the District of Columbia Circuit stayed the
equipment replacement rule pending further review, preventing the rule
from going into effect while the court considers the legal challenges. EPA
has not determined what additional action, if any, it will take regarding
establishing an annual maintenance allowance below which facility changes
would be considered exempt from NSR requirements.

A majority of the state officials expect that the December 2002 final rule
will provide industry with greater flexibility to make facility changes
without triggering NSR requirements for permits. However, a majority of
the officials also expect that the rule will lead to an overall increase
in emissions of harmful air pollutants and hinder efforts to meet air
quality standards, potentially creating or exacerbating risk to public
health. Most of the officials also expect that the rule will increase
their agencies' workload.

  A Majority of the State Officials Expect the December 2002 Rule Will Provide
  Industry with Greater Operating Flexibility but Also Increase Emissions and
  Agencies' Workload

    More than Half of the State Officials Expect Most of the Rule's Provisions
    to Decrease the Number of NSR Permits Issued and Provide Flexibility for
    Companies to Modify Their Facilities

In pursuing the December 2002 final rule, EPA, among other things, sought
to offer facilities greater flexibility to improve and modernize their
operations. Similarly, the state air quality agency officials (29 of 44)
said that in their professional opinion, obtaining fewer permits and more
flexibility to modify facilities are the rule's two primary positive
effects for industry. For example, more than half of the state officials
believe that four of the five provisions of the rule15-including the
revised test for determining whether a facility modification significantly
increases net emissions and is, therefore, subject to NSR-will decrease
the number of permits state air quality agencies issue.

15Less than half of the officials expect the provision exempting pollution
control projects to lead to fewer permits.

For perspective, the state officials reported that their agencies had
issued a total of 600 NSR permits during the 3 years prior to the final
rule to companies that were modifying existing facilities.16 The officials
expect the number of such permits issued to decrease under the final rule
because it expands the range of activities that companies may pursue
without a permit and, in some cases, controls. Forty of the state
officials identified the requirements to install pollution controls as one
of the best features of the NSR program prior to the final rule. According
to EPA, however, several provisions of the rule require companies to make
certain commitments, such as accepting an overall limit on their
emissions, in exchange for avoiding permitting. Therefore, EPA believes
the rule will encourage investments that decrease emissions.

As we reported in our August 2003 report, EPA found that the December 2002
final rule would lead to overall benefits by encouraging energy efficiency
projects, reducing emissions and related health risks, and providing
economic benefits to companies affected by the program. For example, EPA's
analysis found that the rule would encourage companies to implement energy
efficiency projects that would reduce emissions, such as upgrades to
boilers used to generate power.17 However, only 9 of the 44 officials we
surveyed anticipated that the rule would provide the impetus for companies
to increase these projects.

16According to EPA, this number may include some permits for new units at
existing facilities. The officials reported issuing 528 permits to
companies that were constructing new facilities. EPA program managers said
the December 2002 rule did not affect new facilities and new units at
existing facilities.

17As we concluded in our August 2003 report, EPA relied primarily on
anecdotal information from industry in concluding that the NSR program,
prior to the final rule, discouraged some energy efficiency projects,
including some projects that would have reduced air emissions.
Twenty-three of the 44 state officials we surveyed said that the prior NSR
program discouraged companies' willingness to pursue energy efficiency
projects.

    A Majority of the State Officials Expect the Rule to Increase Emissions and
    Hinder Efforts to Meet Health-based Air Quality Standards

A majority of the state officials expect emissions to increase as a result
of the final rule-in contrast to EPA's conclusion, in the agency's
analysis of the rule's environmental effects, that it will reduce
emissions from industrial facilities.18 More specifically, 27 of the 44
officials we surveyed expect that overall, the December 2002 rule will
increase emissions; 8 officials believe emissions will decrease or remain
the same (the remaining 9 officials could not judge the emissions impact).
At least half of the officials thought the rule would increase emissions
of carbon monoxide, nitrogen dioxide, ozone, particulate matter, or sulfur
dioxide- all of which have been linked to health problems and are
controlled by a variety of Clean Air Act programs.

When asked about the emissions impact of each specific provision in the
final rule, a majority of the state officials identified two of the rule's
provisions as most likely to cause emissions increases, as table 2
illustrates. These include the revised methods for determining (1) a
facility's historical or "baseline" emissions and (2) whether a change
will result in a significant net emissions increase.

Table 2: Anticipated Emissions Effects of the December 2002 Final Rule
Provisions (number of state officials' responses)

                              Increase Decrease     No change Unable to 
     Final rule provision    emissions emissions in emissions     judge Total 
          Clean unit                20         6           12         6 
Plantwide emissions limit        24        10            4         6 
Pollution control project        14        10           12         8 
      Revised method for                                                
    calculating "baseline"          29         2            5         8 
           emissions                                                    
       Revised test for                                                 
     calculating emissions          29         1            5         9 
            changes                                                     
    Overall effects of rule         27         5            3         9 

Source: GAO analysis of survey responses.

For example, a majority of the officials believe the "baseline" provision
will increase emissions. This provision allows industrial facilities to
use any consecutive 24-month period in the previous decade as a
baseline.19 EPA changed this emissions calculation method to, among other
things, account for variations in business cycles. The agency concluded
that this

18See U.S. Environmental Protection Agency, Supplemental Analysis of the
Environmental Impact of the 2002 Final NSR Improvement Rules (Washington,
D.C.: Nov. 21, 2002).

19EPA historically used the 2 years immediately preceding the proposed
change to establish a facility's actual emissions. However, in some cases,
the agency allowed use of an earlier period.

provision would have negligible emissions consequences because it would
not alter the baseline for most facilities, including coal-fired power
plants (the largest emitting group of facilities). In addition, companies
must adjust their baselines downward to reflect any other emissions
limitations that have become effective since the period of time they
selected for establishing their baseline, according to EPA.20 EPA program
managers, therefore, maintain that emissions baselines will not
significantly increase as a result of this provision.

Nevertheless, some officials provided written responses to our survey
describing their concerns over this provision. Several such officials
asserted that it allows companies to select the 24-month period within the
previous 10 years in which their emissions were highest.

In addition, 24 officials thought that the provision for plantwide
emissions limits, whereby facilities accept a cap on their overall
emissions to avoid undergoing NSR, would nevertheless increase emissions.
For example, several officials said that the rule enables facilities to
establish their emissions cap based on their highest 2 years of emissions
in the previous 10 years, thereby enabling them to create a cap that
exceeds their current emissions. On the other hand, 10 officials said this
provision would decrease emissions, and several asserted that it creates
incentives for facilities to reduce or limit their emissions. EPA program
managers maintain that this provision will decrease emissions.

In addition to these overall effects, 24 of the state officials anticipate
that the rule will particularly allow facilities built prior to the
establishment of the NSR program in 1977 to increase their emissions. At
the time, the Congress decided to allow existing facilities to defer
installation of pollution controls until a major modification was made
with the expectation that, over time, all facilities would install such
equipment, and this would lead to lower overall emissions. However, as we
concluded in our June 2002 report on emissions from older power plants,21
taken as a whole, such plants still emit more air pollution for each unit
of electricity generated than newer plants. For example, we found that for
each

20According to EPA, such requirements include those limiting emissions of
toxic air pollutants, nitrogen oxides, as well as requirements in state
implementation plans for attainment of the air quality standard for ozone.

21See U.S. General Accounting Office, Air Pollution: Emissions from Older
Electricity Generating Units, GAO-02-709 (Washington, D.C.: June 2002).

megawatt of electricity produced, the older facilities emitted about 100
percent more sulfur dioxide and 25 percent more nitrogen oxides than newer
facilities.

State officials that believe emissions increases will occur under the rule
gave various opinions as to how they would manage such increases. For
example, 7 officials said that the rule would not impede their ability to
meet or maintain air quality standards. Another 14 expect they will offset
the anticipated increases using other air quality regulations, such as
those used to control emissions from mobile sources (automobiles and other
transportation). However, 13 others expect the rule to impede their
ability to meet or maintain standards-despite these other regulations.
(Nine said they could not judge the rule's effects.) This could create
challenges for agencies that expect the rule to interfere with efforts to
meet air quality standards, but that said they were prohibited from
adopting more stringent regulations, such as the District of Columbia,
Kentucky, New Jersey, New Mexico, Oklahoma, Pennsylvania, and Wisconsin.22
On the other hand, 28 state officials said that state law or policy does
not prohibit them from adopting more stringent rules than federal
requirements.

Most State Officials Believe the Rule Will Not Resolve Uncertainty About
When NSR Applies to a Modification

A majority of the state officials' responses contrasted with EPA's
statement that the final rule would provide greater certainty than in the
past for companies and regulators when determining when NSR requirements
apply. Officials identified this uncertainty as one of the program's main
problems before the rule, and 30 officials identified continued
uncertainty as the rule's greatest negative impact on state agencies and
industry.

More specifically, one official explained that the rule is too vague to be
implemented with certainty or enforced. Another state official said that
the rule's new method for determining whether a facility modification
would significantly increase its emissions is by far the most complicated
process yet devised for making such determinations. Furthermore, the
official stated that a company trying to do the right thing could easily
be confused when attempting to determine its future levels of emissions.
This confusion could increase both the burden that the rule imposes on
state

22The following jurisdictions said that they are restricted from adopting
more stringent regulations than federal requirements: the District of
Columbia, Iowa, Mississippi, Missouri, Montana, New Jersey, New Mexico,
Kentucky, Ohio, North Dakota, Oklahoma, Pennsylvania, South Dakota, Utah,
Virginia, and Wisconsin.

agencies to implement and enforce it and the costs for companies that want
to use its provisions.

In addition, 30 of the state officials said that the final rule did not
resolve any of the other significant problems with the program, including
difficulty in determining the stringency of pollution controls that
facilities should install when required to do so.

    A Majority of State Officials Expect the Rule Will Increase Agencies'
    Workloads and Would Like EPA's Assistance in Understanding and Implementing
    the Rule

The state officials' survey responses showed that many expect the final
rule to impose demands on their agencies, including increased workloads.
This comes at a time when many states face budget deficits. Nevertheless,
many of the state officials said their agencies plan to adopt all or most
of the rule's provisions as written (see table 3).

Table 3: Ways in Which Officials Expect Their States to Adopt the Final
Rule

For areas that meet For areas that do not standards meet standards

Adopt all or most provisions as written 20

Seek review of rule in court 8

Adopt or maintain more stringent
regulations 3

Other/do not know 13

Total 44

Source: GAO analysis of state survey responses.

Note: Some states have both areas that meet standards and areas that do
not meet them.

In revising state programs to incorporate the rule, 31 of the officials
said that it would take between one and four staff to adopt the rule's
provisions and obtain EPA's approval of their proposed implementation
plans. Seventeen of the 36 officials that were able to anticipate the
staff needed said that their agency had a plan for obtaining the necessary
staff time, but 15 others did not (4 said they did not know if their
agency had a plan).

In addition, 30 state officials expect that having to administer the rule
after it is adopted will increase their workload at least to some
extent-despite the fact that most expect a decrease in the number of
permits issued in the future. Another 6 expect a decrease in their
workload, and 1 expected no change. As noted above, most officials
expected continued uncertainty for state agencies as a negative impact of
the rule. One official explained that

the state agency was spending considerable time learning the regulations
and training agency staff and companies, while also developing record
keeping, tracking, and other administrative processes. Another official
expected a dramatic increase in the agency's administrative workload,
including time spent reviewing information associated with the rule's
provisions for plantwide limits, among other things. Similarly, another
official expected a high demand among companies for plantwide limits and
that developing them would be very resource intensive. However, another
official said that the workload would increase initially because of the
learning curve with the new program but then decrease over time. The
remaining 7 officials did not know or had not assessed the rule's workload
impact. EPA program managers maintain that, over time, the rule will
decrease the workload for agencies.

To better understand and implement the rule, all but one of the agency
officials said that they would benefit from some type of assistance from
EPA, including updated guidance or workshops.

Similar to their opinions on the final rule, a majority (28 of 42) of the
state officials expected EPA's two NSR revisions-as proposed in December
2002-to provide companies the flexibility to perform maintenance and
replacement activities without obtaining permits and installing pollution
controls. However, at least half of the officials also expected that, as a
result, emissions would increase, and a third expected the exclusions
would exacerbate existing air quality problems and health risks in areas
that already do not meet standards. A majority also expected a greater
administrative burden and uncertainty for agencies in determining when a
facility's activities can be excluded.

  At Least Half of the State Officials Expected the Proposed Revisions Defining
  NSR Exclusions to Provide Industry Greater Flexibility but Also Increase
  Emissions and the Administrative Workload for State Agencies

    A Majority of State Officials Expected the Two Exclusions to Provide
    Industry the Flexibility to Undertake More Facility Modifications without a
    NSR Permit and Pollution Controls

Twenty-eight state officials expected the two exclusions would exempt
facility changes from requirements for permits and controls, decreasing
the number of permits they issue over the next 5 years. This would provide
industry with greater flexibility to perform routine maintenance, repair,
and replacement activities without incurring the costs and delays of the
NSR program. EPA previously determined which activities were considered
routine maintenance, repair, and replacement, and thus excluded from NSR,
on a case-by-case basis. In December 2002, EPA proposed that, in addition
to the case-by-case determination, exclusions could also be determined
according to a cost threshold mechanism, below which activities could be
exempted from NSR, rather than an emissions threshold. Although, at the
time of our survey, the 20 percent cost threshold for replacing equipment
had not been established, one official said that this exclusion would
exempt most facility modifications from NSR.

The state officials identified fewer permits and increased flexibility as
the exclusions' most positive benefits for companies. In addition, 19 of
the officials expected that the exclusions would have a positive effect on
companies' efforts to pursue energy efficiency projects. These officials'
opinions are, therefore, consistent with EPA's finding that the exclusions
would remove barriers to energy efficiency investments.

    About Half of the State Officials Opposed the Equipment Replacement
    Exclusion and a Majority Opposed the Annual Maintenance Allowance

Overall, 21 of the 44 officials said they opposed the equipment
replacement exclusion. Another 12 said they supported this provision, and
the others said they neither supported nor opposed it, or had no opinion.
One of the officials who expressed concerns about the proposal said that
implementing the equipment replacement exclusion would reduce or eliminate
incentives for companies to install well-controlled equipment. Another
official expressed the concern that the exclusion did not include the
necessary provisions to ensure that a company does not replace an entire
emissions unit over a period of just a few years without installing
controls.

In addition, 32 of the 44 officials said they opposed the annual
maintenance allowance exclusion. Another 3 officials said they supported
this provision, and the others said they neither supported nor opposed it,
or had no opinion. Specifically, some states were concerned that the
financial analysis to evaluate the cost data to determine exclusions is
too complex. One official asserted that the annual maintenance allowance
would enable companies to conduct projects that are not routine, thereby
extending the life of equipment that should have been upgraded with more

efficient equipment. According to EPA, the agency received a mixture of
positive and negative comments on the annual maintenance allowance
approach from key stakeholders, including industry, state and local
agencies, and environmental groups. The agency has not determined whether
it will finalize this portion of the proposal or pursue other options to
address routine maintenance activities.

    At Least Half of the State Officials Expected the Proposals Would Increase
    Emissions, Potentially Worsening Air Quality

At least half of the state officials believed that the exclusions would
result in increased emissions of harmful air pollutants. For example, half
expected that the equipment replacement exclusion would increase
emissions, and several believe the cost threshold mechanism will allow
older facilities to avoid installing pollution controls. Only 2 officials
thought that this exclusion would decrease emissions, while the others
expected no change (7) or could not judge (12).

Similarly, 26 of 42 officials who responded said they expected the
proposed annual maintenance allowance exclusion would increase emissions.
For example, one official explained that because the exclusion is based
solely on the amount of money spent without regard to emissions increases,
facilities could make changes that increase emissions and be exempt from
NSR. Only 1 official expected this exclusion would decrease emissions,
while the others expected no change (1) or could not judge (14).

Overall, 21 of the 44 state officials believed the two exclusions would
enable older facilities, built prior to 1977, to increase emissions.
Another 8 expected emissions to decrease or remain the same, and 15 were
unable to judge. As discussed earlier, older power plants emit more pounds
of pollutants per unit of energy generated than newer plants. One official
said that older facilities would continue to be modified without going
through NSR and upgrading their pollution controls. Another official said
that enabling older power plants to avoid installing pollution controls
violated the intent of the Clean Air Act.

While at least half of the officials expected the exclusions to increase
emissions, fewer expected them to exacerbate existing air quality problems
or create new ones. For example, of the 30 officials located in states
with areas that currently do not meet air quality standards, about a third
expect the equipment replacement exclusion to interfere with areas'
efforts to meet standards, while another third did not expect it to
interfere, and the final third could not judge. In addition, 13 of these
officials expected the annual maintenance allowance exclusion to
interfere, while 5

did not, and 12 could not judge. In terms of creating new air quality
problems in areas that currently meet standards, only 5 of 44 officials
expected the equipment replacement exclusion to have this impact, while 20
did not, and 19 could not judge. Furthermore, only 7 of these officials
expected the annual maintenance allowance exclusion to have this impact,
while 16 did not, and 21 could not judge.

The opinions of officials that expect emissions increases and adverse air
quality effects contrast with EPA's conclusion that the exclusions would
enhance the environmental protection and benefit derived from the program.
In addition, EPA's economic analysis of the exclusions found that they
would lead to health benefits and did not account for any potential
health-related costs.23 However, to the extent that either exclusion would
cause or exacerbate violations of health-based air quality standards,
EPA's analysis would have underestimated the health effects and costs of
the exclusions.

    A Majority of the State Officials Expected the Exclusions Would Create a
    Greater Administrative Workload

A majority of the officials said that implementing the exclusions would
increase their administrative burden (27 of 44) and create uncertainty for
agencies in determining when a facility's activities can be excluded (28
of 44). These opinions contrast with EPA's conclusion in the analysis
noted above that they would provide greater regulatory certainty. Several
officials expressed concerns about the complex accounting procedures they
would need to use to determine compliance with the cost threshold
mechanisms and whether modifications could be excluded from NSR
permitting. For example, one official said that the accounting procedures
were well beyond the expertise of the state agency, and another official
described how the agency would need to hire certified public accountants
to determine compliance with the exclusions.

23See U.S. Environmental Protection Agency, Regulatory Impact Analysis for
the Specification of Categories of Activities as Routine Maintenance,
Repair and Replacement for the New Source Review Program (Washington,
D.C.: August 2003).

  Other Stakeholders Expected the Proposed and Final NSR Revisions to Benefit
  Industry but Disagreed on Their Effect on Emissions and Air Quality Agencies'
  Workload

According to key stakeholders we contacted, the proposed and final
revisions to the NSR program would benefit industry by decreasing the
regulatory burden on companies that modify their industrial facilities,
but these stakeholders disagreed on the revisions' impact on emissions and
other factors. Stakeholders representing environmental and public health
groups anticipated that the revisions would mean fewer modifications will
be subject to NSR's permit and control requirements, but more work for
regulators as they look for alternative ways to control emissions. In
contrast, stakeholders representing the industry groups asserted that the
proposed and final changes clarified the NSR program, thereby making
permitting easier, and encouraging investment in energy efficient projects
that lower fuel consumption and emissions. As we concluded in our August
2003 report, the overall economic and environmental effects of the
December 2002 rule are uncertain because of data limitations and
difficulty determining how individual companies will respond to the rule.

    Environmental and Public Health Group Stakeholders Expected the Proposed and
    Final Revisions to Decrease Industry's Regulatory Burden but Increase
    Emissions and Air Quality Agencies' Workload

According to the opinions of the six environmental and public health group
stakeholders we contacted, as well as an association representing all of
the state and local air quality agencies, the proposed and final revisions
would lessen the regulatory burden on companies because, as discussed
earlier, fewer modifications would trigger NSR. Under the prior rules, to
obtain a permit, a company would have to submit an application and go
through a public notice and comment period-a process that could take 3
months to more than 1 year. The company would also have to periodically
report on their compliance with the permit. Furthermore, in cases where
the modification would significantly increase emissions, the company would
have to go through the time and expense of installing emission controls.
As a result of the NSR revisions, however, environmental and public health
stakeholders anticipate that companies would forgo the emissions
reductions that would have been achieved by installing controls, thereby
increasing emissions and public health risks.

As with a majority of the state air quality officials responding to our
survey, nearly all of the environmental and public health group
stakeholders asserted that the proposed and final revisions would create
more work for state and local air quality agencies. Several of them
believe that, because the revisions would result in fewer permits, they
would also result in fewer recordkeeping and reporting requirements for
industry. This, in turn, would make it harder for the agencies to track
and monitor changes at facilities that could influence emissions. For
example, according to the association representing these agencies, the
revisions would make it difficult for them because they would now have to
identify

other sources of emissions information instead of relying on companies to
report this information, as companies were previously required to do under
the NSR program. We concluded in our October 2003 report that, overall, as
a result of the final rule, the public may have less assurance that they
will have notice of, and information about, company plans to modify
facilities in ways that affect emissions, as well as less opportunity to
provide input on these changes and verify they will not increase
emissions.

Some of the environmental and public health stakeholders also pointed out
that the agencies will be forced to find programs other than the federal
NSR program to control emissions so that local air quality meets the
national standards. For example, areas not meeting at least one of the
standards must develop a state plan showing how they will reduce emissions
to comply with the standard.24 But with fewer modifications and facilities
subject to emission controls through NSR, air quality agencies will have
to look for other ways to reduce or control emissions. However, according
to some environmental and public health groups, these alternative
regulations and programs can be more difficult to implement because, for
example, they focus on smaller sources of emissions compared with the
sources subject to the federal NSR program. Therefore, to achieve the same
emissions savings as they would have under NSR, the agencies will have to
track emissions and pursue reductions from a greater number of sources,
requiring more staff time and resources for permitting and enforcement.

24Otherwise, the areas may be subject to sanctions, such as the loss of
access to federal transportation funding.

    Industry Stakeholders Asserted That the Proposed and Final Revisions Clarify
    When NSR Applies to a Modification, Thereby Encouraging Energy Efficiency
    Projects and Reducing Emissions and Air Quality Agencies' Workload

Most industry stakeholders we contacted felt the proposed and final
revisions would lessen, or at least not increase, their regulatory burden,
similar to the opinions of the environmental and public health
stakeholders. Fewer modifications would be subject to the requirements to
obtain a permit and install controls. Furthermore, several industry
stakeholders said their regulatory burden would decrease because the
revisions clarified when NSR actually applied. Several industry
stakeholders explained that before the revisions, companies were uncertain
as to whether some of their modifications triggered NSR. For example, one
stakeholder said that the existing routine maintenance exclusion was
arbitrary and unclear. As a result, to avoid enforcement actions and
penalties, companies would opt not to make the modifications.

On the other hand, the industry stakeholders disagreed with the
environmental and public health stakeholders on a number of other
potential impacts. First, all of the industry stakeholders believed the
changes will encourage companies to invest in energy efficiency projects
they avoided in the past because of NSR requirements. For example, as we
discussed in our October 2003 report, under the prior program, to
determine if a modification would increase emissions enough to trigger
NSR, companies generally had to assume that facilities would run at the
maximum capacity or the highest capacity allowed by the existing NSR
permit after making the modification. A company had to make this
assumption even if the facility had not run at this level in the past or
was not expected to in the future. Industry stakeholders argued that
having to assume this potential increase in emissions biased the test and
overstated the true emissions impact of a project. One industry
representative gave the example of a proposed modification that had the
potential to save the company an estimated $300,000 per year and reduce
emissions, but that the company did not pursue because the emissions test
predicted it would have triggered costly NSR controls.

In the December 2002 final rule, EPA revised the method of calculating the
expected emissions so a company can project the actual activity level-as
opposed to the maximum potential activity level-after the facility change
and estimate the resulting emissions accordingly. Therefore, according to
some of these stakeholders, such energy efficiency projects most likely
will not trigger NSR requirements under the revised rule and will be less
costly for companies to pursue. The industry stakeholders believed that,
with the increased energy efficiency investments, facilities would use
less fuel for the same levels of production.

However, as we discussed in our August 2003 report, industrial facilities'
future production levels and air pollutant emissions may fluctuate in
response to changing economic conditions and other factors. In that
report, we also noted that 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." As a result, some environmental groups that disagreed with industry
were concerned that, if facilities become more efficient, they will
actually cause a net increase in overall emissions and health risks. On
the other hand, according to an EPA program manager, the agency expected
that, if a company increased production at its more efficient facilities,
it could decrease production at its less efficient facilities, more than
offsetting any emissions impact. However, the manager said that the agency
had not analyzed the air pollution impacts of shifts in production that
facilities make after implementing energy efficiency projects to support
the agency's viewpoint.

The industry stakeholders we contacted believed the increased projects and
lower emissions they anticipate will result more from the revisions
included in the December 2002 final rule rather than the October 2003
rule. This is because, according to some stakeholders, the latter rule
simply reinforces how companies had already been interpreting NSR in the
past to determine if a modification was a routine replacement of equipment
and, therefore, exempt from NSR requirements. However, the October 2003
rule specifies a 20 percent cost threshold, below which a company could
make certain changes as routine replacement and exempt from NSR.

Also, in contrast with the environmental and public health groups, some of
the industry stakeholders argued that even with the NSR exemptions,
companies will still have to monitor facility emissions and install
emission control technologies because of other clean air regulations. For
example, under the acid rain program, some utilities have had to control
their facilities' sulfur dioxide and nitrogen oxide emissions. Under the
air toxics program, some companies have had to install controls to reduce
facility emissions of hazardous air pollutants. In addition, the
stakeholders maintained that state and local air quality agencies will
still have to monitor any project that could increase emissions to ensure
compliance with these programs, and the agencies may have their own
requirements governing facility modifications. While this is true, we
noted in our October 2003 report that the scope of the state and local
program requirements varies widely.

Finally, most of the industry stakeholders, unlike the environmental and
health stakeholders, expected a decrease in the state and local air
quality agencies' workload as a result of the proposed and final
revisions. The stakeholders claim the revisions will streamline agencies'
monitoring, minimize the time they spend determining if companies have
properly complied with NSR, and ease the permitting process.

While the stakeholders based their views primarily on professional
opinion, one cited a DOE analysis and another cited an EPA analysis as
support for their views. The DOE analysis included an estimate of
emissions if all coal-fired power plants installed pollution controls
while the EPA analysis focused on the possible emissions consequences of
the equipment replacement exclusion. Neither analysis comprehensively
assessed the impacts of the NSR revisions.

One environmental representative compared the emissions levels in the DOE
analysis with those in the EPA analysis to support the assertion that the
exclusions would represent a rollback from the current program because the
levels in the DOE analysis were lower than EPA's. However, the DOE
analysis is not useful as a benchmark for assessing the effects of EPA's
revisions because, under the NSR program, facilities only have to install
the best available controls when making major modifications. In addition,
this analysis was not specifically related to EPA's NSR revisions.

An industry stakeholder cited the above-mentioned EPA analysis of the
equipment replacement rule to support the assertion that the exclusions
would decrease emissions. However, the EPA analysis was limited in
scope-it considered only power plants (the largest emitting category of
facilities) and only two pollutants, nitrogen oxides and sulfur dioxide.
Another related analysis performed by an EPA contractor included six
additional industries and was based on case studies.

Finally, as we concluded in our August 2003 report, the overall economic
and environmental effects of the December 2002 rule are uncertain because
of data limitations and difficulty determining how industrial companies
will respond to the rule.

Conclusions 	EPA's assessments of the December 2002 and October 2003 NSR
revisions concluded that the rules would provide industry with greater
flexibility to modify their facilities without having to obtain NSR
permits or, in some cases, install pollution controls, while enhancing the
program's environmental benefits. The survey responses indicate that most
state

program managers agreed with EPA's conclusion that the revisions would
enhance flexibility for industry. However, a majority of state program
managers did not agree with EPA's conclusion that the increased
flexibility would lead to less pollution, raising questions about the
final and proposed revisions' environmental effects. Specifically, most of
the state officials believed that the December 2002 rule and the not-yet
finalized annual maintenance allowance exclusion would increase emissions,
and half believed the equipment replacement provision would have this
effect. Furthermore, of those that believe emissions increases will occur,
a number of the officials thought that these anticipated increases would
cause violations of health-based air quality standards or delay the
attainment of the standards in areas that already have poor air quality,
potentially creating or exacerbating health risks. Environmental groups
agreed with the state program managers who expressed concerns, but other
state officials and industry stakeholders maintained the revisions would
have positive environmental effects. Little data currently exist to
resolve these competing viewpoints. We therefore recommended in our August
2003 report that EPA determine what data are available to monitor the
December 2002 rule's effects and use the monitoring results to determine
what effects the rule has created. For the same reason, if the equipment
replacement rule eventually takes effect-pending the resolution of legal
challenges-it will be necessary to monitor its implementation to determine
its environmental and other effects. In addition, more EPA assistance for
states would help them implement the new rules and lessen their
administrative burden.

To ensure that state and local air quality agencies are adequately
equipped to implement the new NSR rules, as required by EPA, and that the
rules do not have unintended effects on emissions and public health, we
recommend that the EPA Administrator (1) provide state and local air
quality agencies with assistance in implementing the December 2002 rule,
(2) pending the court's decision on the equipment replacement rule, work
with state and local air quality agencies to identify the data that the
agency would need to monitor the effects of this rule and use the
monitoring results to identify necessary changes,25 and (3) consider the
state and stakeholder concerns about emissions and workload impacts that
we

25EPA should coordinate this effort with its response to a similar
recommendation in our August 2003 report (GAO-03-947).

  Recommendations for Executive Action

identified before deciding whether to issue a final rule on the second
proposed exclusion, the annual maintenance allowance exclusion.

                                Agency Comments
                               and Our Evaluation

We provided EPA with a draft of this report for review. The Assistant
Administrator for Air and Radiation said that the agency has concerns
about our methodology and certain of our findings. Nevertheless, EPA said
that our recommendations, on their face, make sense, and that the agency
already has plans to take these actions.

Specifically, EPA asserted that GAO (1) in some instances, used the
opinions expressed in the survey responses-which EPA believes may not have
been grounded in a correct understanding of the revisions-as fact, and to
draw conclusions and make recommendations about the NSR program, (2) did
not carry out its work in a way that assured balance and objectivity, (3)
used a skewed survey sample, and (4) should have evaluated whether the
survey results were consistent with the facts cited in EPA's analyses of
the revisions' effects.

GAO disagrees with each of EPA's assertions. First, as we previously
reported and EPA acknowledged, there are limited data available to assess
the effects of the NSR revisions. Therefore, consistent with the review's
objectives, we solicited the opinions of key stakeholders on the
revisions' effects and clearly presented them as opinions in both the
title and body of the report. When, in this context of scarce data, many
state program managers responsible for program implementation express
concerns about the revisions' adverse effects, we believe it would be
prudent to take these concerns seriously. As such, GAO makes a number of
recommendations to (1) collect data on and monitor the revisions' actual
impacts, (2) consider stakeholders' opinions before further revising the
NSR program, and (3) provide state and local agencies assistance in
implementing the revisions. Taking this latter action will help to address
EPA's concerns that the respondents' may not have fully understood the
revisions.

Second, we developed the survey using standard survey research principles
and took steps to minimize question bias, including conducting several
pretests, asking respondents about both the positive and negative effects
of the revisions, and subjecting the survey to a thorough review by a GAO
survey specialist not involved in its development. To ensure the
independence of our efforts, we do not routinely seek the subject agency's
review of our survey instruments. Nonetheless, we worked with NSR program
managers within EPA to understand how the revisions would

work in practice as well as their potential effects and used this
information to design the survey questions.

Third, GAO surveyed the universe of state program managers because we
believe they are in the most informed position to determine the revisions'
impacts on their programs and workloads. Furthermore, in the survey's
instructions we asked the managers, when answering the questions, to
coordinate with the officials within their agencies as they deemed
necessary and appropriate. As such, we relied on each state agency's own
procedures for completing and reviewing the survey responses. In addition,
we surveyed select stakeholders representing environmental, health, and
industry interests. Because of the large number of other affected
stakeholders, it was not feasible to survey the universe. Instead, we
surveyed 30 organizations representing diverse perspectives and chose them
because they were involved in national NSR policy decisions. A number of
these groups represent the views of large numbers of industrial companies
or have a national membership base.

Finally, GAO believes EPA's assertion that we should have evaluated
whether the opinions of state officials responsible for program
implementation were consistent with "facts" cited in EPA's analyses is
disingenuous. As we point out in our previous and current work, these
"facts" are largely assertions based on EPA's limited analysis of the
revisions' effects. We therefore did not use the agency's analysis as a
benchmark to evaluate the survey responses. We further believe that the
state program managers provided plausible explanations for why their views
disagreed with those asserted by EPA.

Appendix III contains the text of EPA's letter along with our detailed
responses to the issues raised. EPA also provided a number of technical
comments, which we have incorporated 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 of the report to the
EPA Administrator and other interested parties. We also will make copies
available to others upon request. In addition, the report will be
available at no charge on the GAO Web site at http://www.gao.gov.

If you have any questions about this report, please contact me at
(202) 512-3841 or [email protected]. Key contributors to this report
are listed in appendix IV.

John B. Stephenson
Director, Natural Resources and Environment

Appendix I: Summary of Local Agency Officials' Responses

  Impacts on Industry

Emissions Impacts

This summary provides an overview of survey responses completed by 45
local air quality agencies-those with independent authority to adopt rules
and write, review, or issue New Source Review (NSR) permits. Seventeen of
the 45 local air quality agencies are in California. The remainder are
scattered across the remaining 14 states that have local air quality
agencies. (See appendix II, table 3.) Detailed local survey results are
available at: http://www.gao.gov/special.pub/gao-04-337sp.

Similar to the state officials, more than half of the local officials
expect that the December 2002 rule will provide industry with greater
flexibility to make facility changes, but also believe that the rule will
result in increased emissions. However, in contrast to the state
officials, fewer than half of the local officials anticipate their
workload to increase as a result of the rule.

For some questions, the number of officials that provided answers varied.

In terms of positive effects, 28 of 45 local officials believe the rule
will result in greater flexibility for industry to make facility changes,
similar to state officials. Also, 24 of the local officials believe that
the rule will benefit industry by enabling companies to avoid NSR
permitting. In addition, 10 of the officials identified greater
opportunities for industry to pursue energy efficiency projects as one of
the positive effects of the rule. On the other hand, only 2 of the local
officials believe the rule will positively affect industry by providing
companies with greater certainty as to when NSR applies to a facility
modification. Twenty of the officials believe that regulatory uncertainty
is one of the rule's primary negative effects.

As with the state officials, a majority (24 of 44) of local officials
expect the rule to increase emissions, while 10 expect no change and 9
were unable to judge. More than half of the officials believe the revised
methods for calculating a facility's historical "baseline" emissions (25
of 44) and estimating emission changes from a modification (23 of 44) will
lead to increased emissions. Fewer than half expect the remaining
provisions to increase emissions. Twenty-two of the officials anticipate
that the rule will allow facilities built prior to 1977-which did not have
to install controls until they made a modification that significantly
increased emissions-to increase total emissions because they can continue
to postpone installing controls, while 11 anticipate no change in
emissions from such facilities. Only 6 officials do not believe the rule
will affect their ability to meet or

Appendix I: Summary of Local Agency Officials' Responses

  Impacts on Local Agencies

Impacts on Industry

Emissions Impacts

continue to meet air quality standards. On the other hand, 16 expect that
they can use other clean air regulations to meet standards, and 11 believe
that taking into consideration the impacts of the final rule, these other
regulations will not help them meet or continue to meet the standards.

Although a greater percentage of local officials (21-24 percent) than
state officials (6-7 percent) anticipate adopting or maintaining more
stringent regulations than EPA, fewer local officials (22 of 45) than
state officials (30 of 44) expect the rule to increase their workload. In
addition, 5 of 43 officials do not anticipate the need for additional
staff to adopt the final rule and obtain EPA approval in contrast to state
officials. All of the local officials said they would like some type of
assistance from EPA, such as implementation workbooks and training
courses.

Similar to state officials, at least half of the local officials expected
the two exclusions for routine maintenance, repair, and replacement
activities to provide industry greater flexibility to make changes, but
unlike state officials, fewer than half expected the exclusions would
increase emissions or their administrative burden. Overall, 22 of 45
officials said that they opposed the equipment replacement exclusion and
16 supported it, while 28 opposed the annual maintenance allowance
exclusion and 5 supported it.

More than half (24 of 45) of the local officials believed the exclusions
would provide industry with greater flexibility to make facility changes,
as did half of the state officials. Twenty-seven believed that not having
to obtain a NSR permit would be one of the exclusions' most positive
benefits for industry. Thirteen of 43 local officials expected the
exclusions to positively affect a company's ability to pursue energy
efficiency projects, while 16 expected no change.

Eighteen of 45 officials expected the equipment replacement exclusion to
increase emissions, while 14 expect no change, and 10 were unable to
judge. Twenty-one of 45 officials expected the annual maintenance
allowance exclusion to increase emissions, while 8 expected no change, and
15 were unable to judge. In addition, 21 of the 45 officials anticipated
that facilities built prior to 1977 would increase emissions as a result
of the exclusions, while 11 expected no change and 10 were unable to
judge. These figures are similar to the state responses, however, compared
with state officials, fewer local officials expected the exclusions to
result in

Appendix I: Summary of Local Agency Officials' Responses

significant enough emissions changes to exacerbate air quality problems in
areas that do not meet standards or cause new problems in areas that
currently meet the standards.

Impacts on Agencies 	Unlike state officials, fewer than half (22 of 45) of
the local officials believed the exclusions would increase their
administrative burden.

Appendix II: Objectives, Scope, and Methodology

The Ranking Minority Member of the Senate Environment and Public Works
Committee and Senator Lieberman asked us to obtain the views of a number
of key stakeholders about the revisions' potential impacts. More
specifically, they asked us to obtain (1) state air quality agency
officials' views about the impacts of the December 2002 final NSR rule on
industry, emissions, and agencies' workloads; (2) state air quality agency
officials' views about the impacts of the two December 2002 proposed NSR
exclusions on industry, emissions, and agencies' workloads; and (3)
environmental, health, and industry organizations' views on the impacts of
all the NSR revisions. In addition, we determined selected local air
quality agencies' views on the revisions' potential effects.

To address the first two objectives and gather information from local
agencies, we conducted an Internet-based survey of 50 state air quality
agencies, the District of Columbia, and the 71 local air quality agencies
that have responsibility for implementing the New Source Review (NSR)
regulations and could potentially issue NSR permits. To ensure that we
obtained information from those that were most involved in the day-to-day
administration of the NSR program and therefore in the best position to
judge the revisions' potential impacts, we worked with the 10 EPA regional
offices and obtained information from the Internet Web site of the
Association of State and Territorial Air Pollution Program Administrators
(STAPPA) and the Association of Local Air Pollution Control Officials
(ALAPCO) to identify the NSR program manager for each agency. The 15
states with local air quality agencies that issue NSR permits are listed
in table 4 below.

     Table 4: States with Local Air Agencies State Number of local agencies

California

Ohio

Missouri, Tennessee

                       Arizona, Indiana, North Carolina 3

                   Alabama, Nebraska, Nevada, Pennsylvania 2

                     Kansas, Kentucky, New Mexico, Oregon 1

Source: GAO.

California is the only state with local agencies covering the entire
state. For the other states, the local agencies are typically located in
larger metropolitan areas with air quality problems. In order to present a
national perspective of the issues faced by air quality officials, we
focused on the

Appendix II: Objectives, Scope, and Methodology

responses from states and highlighted areas where local agencies had
differing points of view.

The state and local air quality officials survey was developed between
December 2002 and April 2003. It includes questions to determine
respondents' views on the NSR program prior to the revisions, as well as
the anticipated effects the proposed and final revisions would likely have
on their programs.

Because we administered the survey to all of the state air quality
agencies and local agencies that have responsibility for implementing the
NSR regulations and could potentially issue NSR permits, our results are
not subject to sampling error. However, the practical difficulties of
conducting any survey may introduce other types of errors, commonly
referred to as nonsampling errors. For example, differences in how a
particular question is interpreted, the sources of information available
to respondents in answering a question, or the types of people who do not
respond can introduce unwanted variability into the survey results. We
included steps in the development of the survey, the collection of data,
and the editing and analysis of data for the purpose of minimizing such
nonsampling errors.

To reduce nonsampling error, we had cognizant officials from STAPPA and
ALAPCO review the survey to make sure that they could clearly comprehend
the questions and estimate the burden it would place on them. We also
pretested the survey with three states and one local agency to ensure that
(1) the questions were clear and unambiguous, (2) terminology was used
correctly, (3) the survey did not place an undue burden on agency
officials, and (4) the survey was comprehensive and unbiased. In selecting
the pretest sites, we sought to include agencies in states that supported
the rules as well as those that did not. We also considered major
subgroups such as states with and without local permitting authorities and
locations across a wide geographical area. To determine what concerns, if
any, those states involved in litigation against the Environmental
Protection Agency (EPA) regarding the NSR reforms would have in completing
the survey, we had an official from the New York State Attorney General's
Office (who is involved in the litigation) review the survey. We asked the
official to identify those questions that states might refuse to answer
because of litigation concerns. In the end, four states involved in the
litigation did not respond to the survey. We made changes to the content
and format of the final questionnaire based on the pretests.

Appendix II: Objectives, Scope, and Methodology

We conducted the survey using self-administered electronic questionnaires
posted to GAO's Web site on the Internet. We sent e-mail notifications to
alert the appropriate officials of the forthcoming questionnaire. These
were followed by another e-mail containing unique passwords and usernames
that enabled the officials to access and complete the survey and notifying
officials that the survey was activated. The questionnaire was available
on the Web until July 7, 2003. We received responses from 44 states and 60
local agencies (each agency could only provide one response). In
summarizing the survey data, the District of Columbia was included in the
state responses. However, 15 of the local agencies that responded told us
that they do not have the authority to adopt their own NSR regulations, or
they do not write or issue NSR permits. Therefore, they were not eligible
respondents and did not provide responses to our more detailed questions.
Thus, 45 local agencies provided complete responses. The overall response
rate was 83 percent. We edited all completed surveys for consistency and,
if necessary, contacted respondents to clarify responses. Table 5 below
lists the states that responded, by EPA region, as well as those that did
not respond (listed in parentheses). It is important to note that four
states in EPA Region 1 declined to respond so as not to disclose
information about their ongoing NSR-related litigation.

                Table 5: Survey Respondents Listed by EPA Region

EPA Region 1 	Maine, Rhode Island (Connecticut, Massachusetts, New
Hampshire, Vermont)

                       EPA Region 2 New Jersey, New York

EPA Region 3 	Delaware, District of Columbia, Maryland, Pennsylvania,
Virginia, West Virginia

EPA Region 4 	Alabama, Georgia, Kentucky, Mississippi, North Carolina,
South Carolina, Tennessee (Florida)

      EPA Region 5 Illinois, Indiana, Michigan, Minnesota, Ohio, Wisconsin

EPA Region 6 Arkansas, Louisiana, New Mexico, Oklahoma, Texas

EPA Region 7 Iowa, Kansas, Missouri, Nebraska

EPA Region 8 	Colorado, Montana, North Dakota, South Dakota, Utah, Wyoming

EPA Region 9 California, Hawaii, Nevada (Arizona)

EPA Region 10 Alaska, Oregon, Washington (Idaho)

Source: GAO analysis of survey responses.

Note: States that did not respond are listed in parentheses.

Appendix II: Objectives, Scope, and Methodology

At the time we conducted our survey, we asked state and local officials
about the impacts of a proposed exclusion from NSR for equipment
replacement activities. Because EPA finalized this exclusion as a rule
after we completed our survey, we took steps to determine whether the
officials' views on the proposal were also true for the final rule. For
example, in December 2003, the national association representing state and
local air pollution control officials told us that, based on their ongoing
dialogue with state and local officials, the survey responses on the
proposed exclusion were consistent with state and local officials' views
on the final rule. In addition, an EPA manager for the NSR program said
that he does not anticipate that the officials who responded to our survey
would have changed their opinions on this exclusion in the time since they
responded to the survey, even though it was not yet in final form at the
time they commented.

To address the third objective, we identified key stakeholders involved in
national level NSR policy decisions and sent them a survey via e-mail
soliciting their responses to a number of questions about the proposed and
final NSR revisions' potential impacts on emissions, industry investments,
and air quality agencies' workloads. We distributed the survey to 30
organizations representing diverse industry and environmental interests.
We used several criteria to select stakeholders for comment. For example,
because of the large number of stakeholders involved in NSR issues at the
national, state, and local level, we focused exclusively on groups that
have a national perspective, including some law firms that represent
several large industries. The stakeholders we selected included the
following:

o  groups identified by knowledgeable EPA officials as key stakeholders;

o  	members of EPA's Permits/NSR/Toxics Subcommittee within its Clean Air
Act Advisory Council (CAAAC) that have a national scope (CAAAC is a senior
level policy committee consisting of approximately 60 senior managers and
experts representing state and local government, environmental and public
interest groups, academic institutions, unions, trade associations,
utilities, industry, and other experts);

o  	national level groups that have testified in Congress on NSR and Clean
Air Act issues over the last several years;

o  national level groups that commented on EPA's NSR proposals; and

o  	trade associations representing those industries identified by EPA as
those most affected by NSR.

Appendix II: Objectives, Scope, and Methodology

We again took steps in the design, data collection, and analysis phases of
the survey to minimize nonsampling and data processing errors, including
pretesting of the survey questions, follow-up with those that did not
respond promptly, and independent verification of all survey responses
entered into an analysis database. We conducted two pretests of the survey
and made changes to the content and format of the final questionnaire
based on the pretests.

The survey was sent to the key stakeholders on July 2, 2003, and was
available until July 18, 2003. Of the 30 stakeholders contacted, the
following 14 responded to this survey:

o  American Forest & Paper Association;

o  American Lung Association;

o  American Petroleum Institute;

o  Clean Air Task Force;

o  Clean Air Trust;

o  Council of Industrial Boiler Owners;

o  Edison Electric Institute;

o  Energy and Innovation Center, Environmental Law Institute;

o  Hogan & Hartson LLP;

o  Morgan, Lewis & Bockius LLP;

o  	National Environmental Development Association's Clean Air Regulatory
Project;

o  National Petrochemical & Refiners Association;

o  National Resources Defense Council; and

o  STAPPA/ALAPCO.

Appendix II: Objectives, Scope, and Methodology

We edited all completed surveys for consistency and, if necessary,
contacted respondents to clarify responses.

For all of these objectives, we worked with cognizant EPA officials,
including the agency's NSR program manager.

Detailed survey results are available at:
http://www.gao.gov/special.pubs/gao-04-337sp.

We conducted our review from September 2002 through January 2004 in
accordance with generally accepted government auditing standards.

                        Appendix III: Comments from the
                        Environmental Protection Agency

Note: GAO comments supplementing those in the report text appear at the
end of this appendix.

See comment 1.

        Appendix III: Comments from the Environmental Protection Agency

                         See comment 2. See comment 3.

        Appendix III: Comments from the Environmental Protection Agency

                                 See comment 1.

                                 See comment 4.

        Appendix III: Comments from the Environmental Protection Agency

                                 See comment 4.

                                 See comment 1.

                                 See comment 1.

                         See comment 1. See comment 2.

        Appendix III: Comments from the Environmental Protection Agency

                                 See comment 3.

                                 See comment 3.

                                 See comment 3.

Appendix III: Comments from the Environmental Protection Agency

  GAO Comments

The following are GAO's comments on the letter from the Environmental
Protection Agency dated January 23, 2004.

1. 	GAO does not agree with EPA's assertions that we, in some instances,
used the opinions expressed in the survey results as facts. Consistent
with the review's objectives, this report carefully characterizes the
survey results as opinions. To this end, the report's title clearly points
out that we are presenting stakeholders' views. In addition, based on our
two earlier reports on the revisions, we determined that, at best, limited
data exist on the effects of the prior NSR program or the potential
effects of the revisions. In fact, as we found in our August 2003 report
on the analytical basis for EPA's December 2002 rule,1 EPA itself relied
primarily on the professional judgment of agency staff and comments
received on earlier NSR reform proposals, rather than a comprehensive
quantitative analysis of the rule's possible effects, in initially
justifying the rule. That report also described limitations with the
agency's subsequent analysis of the rule's environmental effects. Because
of our earlier findings about data limitations, we believe it useful and
entirely appropriate to supplement the available information with the
informed opinions of those most involved in the day-to-day administration
of NSR programs.

GAO also disagrees with EPA's characterization that we improperly used
stakeholder opinions to draw conclusions and make recommendations about
the NSR revisions. While recognizing that the results were based on
opinion, it is important to point out that these were the opinions of
those on the front lines of program implementation. In this case, these
informed opinions raise important questions about the revisions' effects.
Because of these questions and, in light of the limited hard analytical
data on the revisions' effects, GAO recommends that EPA collect data on
the revisions' actual impacts. We made this recommendation in our August
2003 report regarding the December 2002 rule and again in this report
regarding the equipment replacement exclusion. We further recommend that
EPA consider these informed opinions before further revising the NSR
program. EPA also questioned whether the respondents' opinions were
grounded in a correct understanding of the rules' provisions. We believe
that our recommendation that EPA provide state and local air

1U.S. General Accounting Office, Clean Air Act: EPA Should Use Available
Data to Monitor the Effects of Its Revisions to the New Source Review
Program, GAO-03-947 (Washington, D.C.: Aug. 22, 2003).

Appendix III: Comments from the Environmental Protection Agency

quality agencies with assistance in implementing the revisions will help
to address this concern. Despite its concerns, EPA said these three
recommendations, on their face, makes sense and that the agency plans to
take these actions.

2. 	GAO disagrees with EPA's assertion that the way we carried out our
work did not assure balance and objectivity. We developed the survey using
standard survey research principles. This included taking steps to
minimize question bias, asking respondents about both the positive and
negative effects of the revisions, providing respondents with a range of
answers to each question (including "no change" or "no effect"), and
assessing each question for bias and problematic wording during an
extensive pretesting and review process. We also sought to eliminate bias
and problematic wording by subjecting the survey to a thorough review by a
GAO survey specialist who was not involved in its development. Regarding
the external review of our survey, we point out in the objectives, scope,
and methodology section that we asked the trade association that
represents state and local air quality control agencies (i.e., the
officials we surveyed) to help ensure that their members could clearly
comprehend the questions and estimate the burden it would place on them.
We also asked a representative of the New York Attorney General's Office
to review the survey specifically to gauge whether they thought those
states involved in lawsuits with EPA over the reforms would be concerned
about completing the survey. Finally, to ensure the independence of our
efforts, we do not routinely seek the subject agency's review of our
survey instruments. Nonetheless, we held discussions with staff in EPA's
Office of Air Quality Planning and Standards, Office of Enforcement, and
regional offices to make sure we understood the technical nature of the
revisions when developing the survey. In preparing our two prior NSR
reports, we also worked closely with EPA's managers of the NSR program to
understand the agency's assessment of how the revisions would work in
practice, as well as the potential effects. We used all of this
information to design the survey questions.

3. 	GAO disagrees with EPA's assertion that our survey sample was skewed.
GAO surveyed various stakeholders including state and local officials, as
well as industry and environmental groups, and this report presents a
range of views on the possible effects of the NSR revisions. We describe
our methodology for selecting these stakeholders in the report's
objectives, scope, and methodology section. More specifically, with
respect to our survey of state agency officials, we point out that we did
not survey a sample, but the universe of state NSR program managers. We
sent the survey to the manager of each state's NSR

Appendix III: Comments from the Environmental Protection Agency

program within the state environmental agency (instead of the agency head)
because these program managers are responsible for day-to-day program
implementation and hence are in the most informed position to determine
the revisions' impacts on their programs and workloads. Furthermore, in
the survey's instructions we asked the managers, when answering the
questions, to coordinate with the officials within their agencies as they
deemed necessary and appropriate. As such, we relied on each state
agency's own procedures for completing and reviewing the survey responses.
In several cases, in fact, the program managers told us the reason they
needed additional time to submit their responses to us was because the
responses were under review by others within their agencies.

EPA also questioned why we surveyed every state and local agency but only
a handful of environmental groups and industry trade associations, and no
individual industry officials. GAO gathered more detailed information from
state and local agencies than from other stakeholders because these
agencies generally implement the regulations and we were asked to obtain
information on how this implementation would affect agencies' programs and
workload. Because of the large number of other affected stakeholders, it
was not feasible to survey the universe. Instead, we surveyed key
stakeholders that had been involved in national NSR policy decisions,
which included 30 organizations, in order to obtain diverse industry and
environmental perspectives. Of the 30 organizations we surveyed, 14
responded, and 8 of the respondents represented industry. A number of the
organizations that responded represent large numbers of industrial
companies, including the American Forest & Paper Association and the
American Petroleum Institute. Likewise several of the environmental and
health groups represent a national membership base, including the American
Lung Association and the Natural Resources Defense Council.

GAO also disagrees with EPA's assertion that we focused disproportionately
on the state officials' unfavorable opinions of the rule. In presenting
the state survey results, we generally listed the total number of
officials responding to a question and information on the distribution of
their responses. We then provided more detailed information about the
majority's opinion for each question, consistent with standard survey
principles. In most cases, it turned out that the majority of respondents
to our questions held the view that the revisions would have an adverse
impact on emissions and their workload, contrary to EPA's conclusions
about the revisions' impacts.

Appendix III: Comments from the Environmental Protection Agency

We were very careful, however, to also discuss the number of respondents
who held the minority view on a particular topic.

4. 	EPA's letter suggests that GAO should have evaluated whether the
survey results were consistent with the "facts," asserts that many of the
survey responses are not, and cites its own analysis of the revisions'
emissions impacts as factual support for its position. EPA also cites two
examples of cases in which the agency believes respondents' opinions
conflict with fact. EPA's comment related to the "facts" however, largely
only represents references to its own assertions.

First, as discussed previously we have identified limitations with EPA's
analysis of the revisions' impacts. As we stated in our August 2003
report, a senior EPA economist said that uncertainty about the extent to
which companies might elect to use the NSR alternatives in the December
2002 rule limited the agency's ability to estimate the rule's impacts. For
these reasons, we did not use EPA's analysis as a benchmark to evaluate
the survey responses. Again, in this context, the opinions of key
stakeholders, especially those responsible for implementing the
regulations, provide an important perspective appropriately considered by
congressional decisionmakers.

Second, as to EPA's examples of opinions conflicting with facts, the
agency suggests that the opinions of those who expect the December 2002
rule to delay attainment of air quality standards are incorrect. In its
first example, EPA states that (1) even if emissions increased, the
increase would be small and dwarfed by decreases coming from other air
quality regulations, and (2) facilities affected by the revisions either
already have emissions controls or would have to have them to qualify for
many of the rule's exemptions, such as those for plantwide emissions
limits, clean units, and pollution control projects. Regarding the first
point, as we report, 7 officials agreed with EPA and said that the rule
would not impede their ability to meet or maintain air quality standards.
Another 14 expect they will offset the anticipated increases using other
air quality regulations. A minority of the respondents (13) said that,
despite these other regulations, they would still have difficulty meeting
or maintaining air quality standards. Therefore, these 13 officials
already took into account the other air quality regulations EPA cites.
Regarding the second point, EPA did not mention the rule's key
exemption-the revised method for determining facilities past
emissions-that does not require that facilities have emissions controls,
and was the provision cited most often by the state officials as likely to
lead to emissions increases. While EPA maintains

Appendix III: Comments from the Environmental Protection Agency

that this provision will not have a significant environmental impact,
agency managers for the NSR program acknowledged that EPA's analysis
justifying its position was not based on a statistically valid sample of
affected facilities. Ultimately, many stakeholders disagreed with EPA's
assertions.

Regarding the second example, GAO agrees that these opinions conflict with
EPA's information but also believes the state officials provided plausible
explanations for why they expect their burden to increase even though they
expect to issue fewer permits. As we point out in the report, some of the
officials said that they find the December 2002 rule confusing,
complicated, and leading to more uncertainty about the NSR program-all of
which can contribute to agencies' workloads. While EPA asserts that the
rule will lead to an overall reduction in workload for agencies based on
its experience with six states that have used flexible permitting systems,
our survey results found that four of these same states said the
opposite-they expect the rule to increase their workload (one had not
assessed such impacts, and the other did not know what the effect would
be). Officials from the four states said they would spend more time
drafting laws, regulations, and guidance, as well as processing permits,
explaining the rule to industry, and training staff. Furthermore,
officials from the four states said that they expected more work
associated with the rule's revised method for determining facilities' past
emissions. Therefore, GAO disagrees with EPA's assertion that the
experience of these states shows that the rule will reduce agencies'
workloads or that the survey results are contradictory.

Appendix IV: GAO Contacts and Staff Acknowledgments

GAO Contacts

  Staff Acknowledgments

(360404)

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

In addition to the individuals named above, Ulana Bihun, Michael Hix,
Jeffrey Larson, Lisa Turner, and Laura Yannayon made key contributions to
this report. Nancy Crothers, Bob DeRoy, Tim Guinane, Karen Keegan, Judy
Pagano, Minette Richardson, and Monica Wolford also made important
contributions.

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

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.

Order by Mail or Phone 	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:

To Report Fraud, Web site: www.gao.gov/fraudnet/fraudnet.htm

  Waste, and Abuse in E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Jeff Nelligan, Managing Director, [email protected] (202) 512-4800

Public Affairs 	U.S. General Accounting Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***