Order Code RL33776
Clean Air Act Issues in the 110th Congress:
Implementation and Oversight
January 3, 2007
James E. McCarthy
Specialist in Environmental Policy
Resources, Science, and Industry Division

Clean Air Act Issues in the 110th Congress:
Implementation and Oversight
Summary
Broad questions regarding the science underlying new ambient air quality
standards, the role of federal versus state governments in controlling air pollution,
and the appropriateness of economic versus regulatory approaches to controlling
emissions are underlying issues as the 110th Congress begins oversight of the Clean
Air Act. Oversight hearings are expected early in the new Congress. Specific issues
include whether the EPA’s new standards for ambient concentrations of fine
particulates and its soon-to-be-proposed standards for ozone adequately reflect the
state of the science; whether the EPA’s recently announced changes in the process
for setting ambient air quality standards politicize what traditionally have been
scientific judgments; and how best to control emissions of mercury and other
pollutants from electric power plants. To some extent, the issues overlap the debate
regarding control of greenhouse gases, which is also expected to be the subject of
hearings early in the new Congress. Amendments to the Clean Air Act or directives
to the EPA through its appropriations may be considered. In addition, the executive
and judicial branches and the states are expected to take action on air issues in the
coming year.
On October 17, 2006, EPA Administrator Stephen Johnson promulgated
revisions to the National Ambient Air Quality Standards (NAAQS) for particulate
matter (PM). The new standards will cut the allowable concentration of fine particles
(known as PM ) in the air averaged over 24-hour periods almost in half, avoiding
2.5
several thousand premature deaths annually. The Administrator left unchanged a
separate annual standard for fine particles, despite the recommendation of his
independent scientific advisory committee (CASAC) that the annual standard also
be reduced. The EPA has recently announced changes to the process by which
NAAQS are set that appear to downgrade the role of CASAC in any future NAAQS-
setting. The Administrator is not required by statute to follow CASAC’s
recommendations. Thus, whether CASAC’s role should be changed — either as the
EPA recently announced or, by contrast, to strengthen its statutory role — might be
at issue in the new Congress.
The regulation of mercury and other pollutants from coal-fired electric power
plants is another issue of congressional concern. In 2005, the EPA promulgated new
regional cap-and-trade requirements for mercury, sulfur dioxide, and nitrogen oxides
from power plants, after the Senate Environment and Public Works Committee did
not approve legislation (the Clear Skies Act, S. 131) that would have established
similar requirements nationally and modified existing clean air regulations. About
20 states have chosen to proceed with mercury standards stronger than those EPA
promulgated, and many health and environmental groups would like to see EPA’s
rules made more stringent. The costs and benefits of various levels of control, the
availability of control technology, and legal issues related to the promulgated
standards are among the issues faced by Congress, the courts, and the EPA.

Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Revision of the Particulate Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
CASAC’s Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Impacts of the New Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Implementation of the NAAQS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
CASAC’s Role in the NAAQS-Setting Process . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Multi-pollutant Legislation for Power Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Mercury From Power Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
New Source Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
List of Figures
Figure 1. Counties Exceeding Revised PM Standards, Based on
2.5
2003-2005 Monitoring Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
List of Tables
Table 1. Pre-existing, Recommended, and New NAAQS for PM
. . . . . . . . . . . 3
2.5
Table 2. Estimated Costs and Benefits of the EPA’s New PM Standards . . . . . 6
2.5

Clean Air Act Issues in the 110th Congress:
Implementation and Oversight
Introduction
Despite steady improvements in air quality in many of the United States’ most
polluted cities, the goal of clean air continues to elude many areas. The most
widespread problems involve ozone and fine particles. As of March 2006, 158
million people lived in areas classified “nonattainment” for the ozone National
Ambient Air Quality Standard1; 88 million lived in areas that were nonattainment for
fine particles (PM ).2
2.5
Air quality has improved substantially since the passage of the Clean Air Act
in 1970: annual emissions of the six most widespread (“criteria”) air pollutants have
declined 160 million tons (53%), despite major increases in population, motor
vehicle miles traveled, and economic activity.3
Meanwhile, however, scientific understanding of the health effects of air
pollution has caused the EPA to tighten standards for ozone and fine particles. (Fine
particles, as defined by the EPA, consist of particulate matter 2.5 micrometers or less
in diameter, abbreviated as PM .) The agency attributes at least 33,000 premature
2.5
deaths and millions of lost work days annually to exceedances of the PM standard.
2.5
Recent research has begun to tie ozone pollution to premature mortality as well.
Thus, there is continuing pressure to tighten air quality standards: a tightening of the
standard for fine particles was promulgated October 17, 2006. Ozone standards are
scheduled for review in 2007, with new standards to be proposed by May and a final
decision due by February 2008. In addition to the standards themselves, attention has
focused on the major sources of ozone and particulate pollution, such as coal-fired
power plants and mobile sources.
With this background in mind, this report provides a discussion of several
interrelated air issues of interest in the 110th Congress, including revision of the
particulate standards, the role of independent scientific review in the setting of air
quality standards, multi-pollutant legislation for electric power plants, mercury from
power plants, and New Source Review. This report provides an overview of these
1 Data for ozone nonattainment areas are from the U.S. EPA “Green Book,” at
[http://www.epa.gov/oar/oaqps/greenbk/gntc.html].
2 Data for PM nonattainment areas are also from the U.S. EPA “Green Book,” at
2.5
[http://www.epa.gov/oar/oaqps/greenbk/qntc.html].
3 See U.S. EPA, “Air Emission Trends — Continued Progress Through 2005,” at
[http://www.epa.gov/airtrends/econ-emissions.html].

CRS-2
issues; CRS reports that contain additional information and detailed sources are
referenced in the appropriate sections.
Revision of the Particulate Standards
On September 21, 2006, EPA Administrator Stephen Johnson signed revisions
to the National Ambient Air Quality Standards (NAAQS) for particulate matter.
(The standards appeared in the Federal Register on October 17.4) In developing the
revisions, the EPA reviewed 2,000 scientific studies on particulates and found
associations between particulates and numerous significant health problems,
including aggravated asthma, chronic bronchitis, reduced lung function, irregular
heart beat, heart attacks, and premature death in people with heart or lung disease.
The revisions will strengthen the pre-existing standard for particulate matter 2.5
micrometers or less in diameter (known as fine particles, or PM ), but the standard
2.5
has not been strengthened to the degree recommended by the agency’s staff or
scientific advisors. As shown in Table 1, the new standard cuts the allowable
concentration of PM in the air averaged over 24-hour periods from 65 micrograms
2.5
per cubic meter (µg/m3) to 35 µg/m3; the annual standard, set at 15 µg/m3, does not
change.
EPA’s professional staff and the Clean Air Scientific Advisory Committee
(CASAC), a group established by the Clean Air Act to provide independent scientific
advice to the Administrator, had recommended more stringent standards. CASAC
endorsed a 24-hour standard in the range of 30 to 35 µg/m3 and an annual standard
in the range of 13 to 14 µg/m3. Of the 22 CASAC panel members, 20 concurred in
the recommendation.5
4 71 Federal Register 61144. Extensive information related to the standards, including an
eight-page fact sheet explaining the standards, and maps and charts with background
material, is available at [http://epa.gov/pm/actions.html].
5 By statute, CASAC consists of seven members chosen by the EPA Administrator. To
review the NAAQS for a specific pollutant, CASAC forms a panel that includes as many
subject experts as CASAC deems appropriate, in addition to the seven statutory CASAC
members. Thus, the PM panel had 22 members.

CRS-3
Table 1. Pre-existing, Recommended, and New NAAQS for PM2.5
Annual Standard
24-Hour Standard
Pre-existing Standardsa
15 µg/m3
65 µg/m3
15 µg/m3 and mid to lower end of 25-35 µg/m3
EPA Staff Recommendation
OR
12-14 µg/m3 and mid to lower end of 30-40 µg/m3
CASAC Recommendation
13 to 14 µg/m3
30 to 35 µg/m3
Administrator’s Decision
15 µg/m3
35 µg/m3
a. Although these standards were promulgated in 1997, they are only now coming into effect, because
of legal challenges, the need to establish a monitoring network, and various administrative
factors. For additional information on implementation of the current standard, see CRS Report
RL32431, Particulate Matter (PM2.5): National Ambient Air Quality Standards (NAAQS)
Implementation
, by Robert Esworthy.
In the Administrator’s judgment, the science underlying this recommendation
was not sufficient, relying primarily on two studies, neither of which “provide[s] a
clear basis for selecting a level lower than the current standard....”6 The
Administrator agrees with CASAC that the science shows a relationship between
higher levels of PM and an array of adverse health effects, but he believes there is
2.5
too much uncertainty in the analysis to justify lowering the annual standard.7 He also
noted that the EPA is undertaking substantial research to clarify which aspects of
PM-related pollution are responsible for elevated risks of mortality and morbidity,
including a multi-million-dollar research program whose timeline should permit the
results to inform the Agency’s next periodic reevaluation of the PM standard,
2.5
required by statute within five years. Thus, he concluded, “...it would be wiser to
consider modification of the annual standard with a fuller body of information in
hand than initiate a change in the annual standard at this time.”8
The PM NAAQS also addresses slightly larger, but still inhalable, particles in
the range of 10 to 2.5 micrometers. These are referred to as thoracic coarse
particles
, or PM
. In its last review of the particulate standards (in 1997), the EPA
10-2.5
had regulated these as particles 10 microns or smaller (PM ), a category that
10
overlapped the PM category. Challenged in the D.C. Circuit Court of Appeals, the
2.5
PM standard was remanded to the EPA, the court having concluded that PM is a
10
10
“poorly matched indicator” for thoracic coarse particles because it includes the
smaller PM category as well as the larger particles. In response, in January 2006,
2.5
the EPA proposed a 24-hour standard for PM
. The standard would have been set
10-2.5
at a level of 70 µg/m3, compared with the current 24-hour PM standard of 150
10
µg/m3. The final standards promulgated in October reversed course, leaving in place
6 U.S. EPA, National Ambient Air Quality Standards for Particulate Matter, Proposed Rule,
Preamble, 71 Federal Register 2651, Jan. 17, 2006.
7 See discussion beginning at 71 Federal Register 61172, Oct. 17, 2006.
8 71 Federal Register 2652, Jan. 17, 2006.

CRS-4
the current form of the standard (PM ) and the current level (150 µg/m3). The only
10
change to the PM standard was the revocation of its annual component. The
10
agency argues that it has provided more thorough reasoning in support of the use of
PM as its coarse particle indicator, and believes that its explanation will satisfy the
10
court.
CASAC’s Views. The Administrator’s decisions represent the first time in
CASAC’s nearly 30-year history that the promulgated standards fall outside of the
range of the scientific panel’s recommendations. In a letter dated September 29,
2006, the seven members of CASAC objected to the Administrator’s actions, both
as regards PM and PM . With regard to PM , the letter stated: “CASAC is
10
2.5
2.5
concerned that EPA did not accept our finding that the annual PM standard was not
2.5
protective of human health and did not follow our recommendation for a change in
that standard.”9 The letter noted that “there is clear and convincing scientific
evidence that significant adverse human-health effects occur in response to short-
term and chronic particulate matter exposures at and below 15 µg/m3
,” and noted
that 20 of the 22 Particulate Matter Review Panel members, including all 7 members
of the statutory committee, were in “complete agreement” regarding the
recommended reduction: “It is the CASAC’s consensus scientific opinion that the
decision to retain without change the annual PM standard does not provide an

2.5
‘adequate margin of safety ... requisite to protect the public health’ (as required by
the Clean Air Act) ....
”10
With regard to PM , the letter stated that CASAC was “completely surprised”
10
at the decision to revert to the use of PM as the indicator for coarse particles, noting
10
that the option of retaining the existing daily PM standard was not discussed during
10
the advisory process and that CASAC views this decision as “highly problematic.”
The Administrator is not required by statute to follow CASAC’s
recommendations; the Act (Section 307(d)(3)) requires only that the Administrator
set forth any pertinent findings, recommendations, and comments by CASAC (and
the National Academy of Sciences) and, if his proposal differs in an important
respect from any of their recommendations, provide an explanation of the reasons for
such differences. Courts, in reviewing EPA regulations, generally defer to the
Administrator’s judgment on scientific matters, focusing more on issues of
procedure, jurisdiction, and standing. Nevertheless, CASAC’s detailed objections
to the Administrator’s decisions and its description of the process as having failed to
meet statutory and procedural requirements could play a role if the standards are
challenged in court.
Impacts of the New Standard. The EPA is prohibited from taking cost into
account in setting NAAQS, but to comply with an executive order, the agency has
produced a Regulatory Impact Analysis (RIA) analyzing in detail the costs and
9 Letter of Rogene Henderson et al. to Hon. Stephen L. Johnson, EPA Administrator, Sept.
29, 2006, available at [http://www.epa.gov/sab/pdf/casac-ltr-06-003.pdf].
10 Ibid. Italics in original.

CRS-5
benefits of the new PM standards.11 The agency estimates that compliance with the
new PM standard will prevent 1,200 to 13,000 premature deaths annually, as well
2.5
as substantial numbers of hospital admissions and missed work or school days due
to illness.12 The agency actually produced three sets of benefit numbers, based on
three different studies. The study on which the agency seems to have placed the
greatest emphasis, conducted for the American Cancer Society, was used to estimate
that 2,500 premature deaths would be avoided. The other two studies would have
produced higher benefit numbers. The Harvard Six-City Study, for example, was
used to estimate a reduction of 5,700 premature deaths annually, and an “Expert
Elicitation”13 produced a mean estimate of 7,000 premature deaths reduced. Critics
of the rule argue that as many as 30,000 premature deaths could be avoided annually
if the Administrator had chosen the more stringent standards endorsed by CASAC.14
The higher estimate is based on the agency’s Expert Elicitation.
The agency’s RIA estimates the cost of meeting the new standards at $5.4
billion annually in 2020 and, as shown in Table 2, provides a range of benefit
estimates (from $8 billion to $76 billion annually, depending on the number of
avoided deaths, the choice of discount rate, and other factors). A more stringent
alternative (reducing the annual standard to 14 µg/m3) would increase the cost by
about 50%, to $7.9 billion annually, according to the agency, but would nearly double
the estimated benefits.15 Thus, the benefit-cost ratio would be more favorable,
according to the agency’s analysis, had the Administrator chosen the more stringent
standard.
11 [http://epa.gov/pm/actions.html].
12 See “Regulatory Impact Analysis of EPA’s Final Revisions to the National Ambient Air
Quality Standards for Particle Pollution (Particulate Matter),” Fact Sheet, p. 2, at Ibid.
13 In response to recommendations made in a 2002 National Academy of Sciences (NAS)
report, “Estimating the Public Health Benefits of Proposed Air Pollution Regulations,”the
EPA has been exploring ways to improve the characterization of uncertainty in its analyses
of the health benefits of regulations affecting air quality. One suggested method for doing
so was through the use of expert judgment. To solicit such judgment, the EPA used a wide
range of nomination methods to assemble a group of 12 leading experts (8 epidemiologists,
3 toxicologists/health scientists, and 1 clinician) to respond to a question regarding the
change in mortality associated with a defined change in PM concentration. For additional
2.5
information, see Industrial Economics, Incorporated (for U.S. EPA, Office of Air Quality
Planning and Standards), Expanded Expert Judgment Assessment of the Concentration-
Response Relationship Between PM Exposure and Mortality
, Sept. 21, 2006.
2.5
14 “Stronger Soot Rule Could Avert 30,000 Premature Deaths — EPA Report,” E&E News
PM
, Sept. 22, 2006.
15 U.S. EPA, Regulatory Impact Analysis of the 2006 National Ambient Air Quality
Standards for Fine Particle Pollution (PM2.5)
, Table ES-1, available at [http://epa.gov/pm/
actions.html].

CRS-6
Table 2. Estimated Costs and Benefits of the
EPA’s New PM Standards
2.5
($ billion)
Basis of Benefit Estimate
Cost
Benefits
American Cancer Society Study
$5.4
$15 - $17
EPA Expert Elicitation
$5.4
$8 - $76
Source: EPA Regulatory Impact Analysis.
Using the most recent available monitoring data, the agency identified 143
counties where air quality is worse than allowed under the new standards. Observed
on a map (Figure 1), these areas can seem small compared with the approximately
3,000 counties in the United States, but two factors make the impact of the standards
far larger. First, the number of counties where emissions will need to be controlled
may be two or three times the number of those exceeding the standard, because
“nonattainment areas” include both counties where pollutant concentrations exceed
the standard and those that contribute to exceedance of the standard in adjoining
counties. Entire metropolitan areas tend to be designated nonattainment, even if only
one county in the area has readings worse than the standard. Second, the
nonattainment counties tend to have larger populations than those in attainment: 88
million people (about 30% of the U.S. population) live in the 208 counties designated
nonattainment for the current standard. The new standard may affect an even larger
percentage of the population.


CRS-7
Figure 1. Counties Exceeding Revised PM Standards,
2.5
Based on 2003-2005 Monitoring Data
Source: U.S. EPA.

CRS-8
Implementation of the NAAQS. A NAAQS does not directly limit
emissions; rather, it represents the EPA Administrator’s formal judgment regarding
the level of ambient pollution that will protect public health with an adequate margin
of safety. Promulgation of NAAQS sets in motion a process under which states and
the EPA first identify nonattainment areas. After these areas are formally designated
(a process the EPA estimates will take until April 2010 for the revised PM2.5
standard), the states have three years to submit State Implementation Plans (SIPs)
that identify specific regulations and emission control requirements that will bring
the area into attainment. Attainment of the revised standard is to be achieved by
2015, according to the EPA, with a possible extension to 2020.
Issues. A number of issues were raised during consideration of the proposed
standards, and most remain in the wake of the Administrator’s decision. Those who
would like to see stronger standards (including a number of states and environment
and health groups) have focused on the agency’s disregard of CASAC’s
recommendation that the annual PM standard be strengthened. Some industrial and
2.5
agricultural interests, on the other hand, are questioning the agency’s strengthening
of the standard for all fine particles, without distinguishing their source or chemical
composition. The agency’s response to this is that “... studies suggest that many
different chemical components of fine particles and a variety of different types of
source categories are all associated with, and probably contribute to, mortality, either
independently or in combinations.”16 These and other issues may be raised in court
challenges or in congressional oversight. The Clean Air Subcommittee of the Senate
Environment and Public Works Committee held oversight hearings on the PM
proposal on July 13 and July 19, 2006. Thirteen states, the District of Columbia,
electric utilities and other industry groups, groups representing farmers and ranchers,
and several environmental groups have challenged the standards in court.
(For a more detailed discussion of the new NAAQS, see CRS Report RL33254,
Air Quality: EPA’s 2006 Changes to the Particulate Matter (PM) Standard, by
Robert Esworthy and James McCarthy.)
CASAC’s Role in the NAAQS-Setting Process
The completion of the PM NAAQS review was followed in short order by an
EPA announcement, on December 7, 2006, that it will modify the process for setting
and reviewing NAAQS. Sections 108 and 109 of the Clean Air Act establish
statutory requirements for the identification of NAAQS (or “criteria”) air pollutants17
and the setting and periodic review of the NAAQS standards. However, the process
used by the agency is as much the result of 36 years of agency practice as it is of
statutory requirements. In Section 109, for example, the statute establishes a Clean
16 EPA, Office of Research and Development, Air Quality Criteria for Particulate Matter,
p. 9-31, as cited in Section II.C. of the Preamble to the Final Particulate Rule. See 71
Federal Register 61162 et seq., Oct. 17, 2006, for additional discussion.
17 Criteria pollutants are pollutants that endanger public health or welfare, in the
Administrator’s judgment, and whose presence in ambient air results from numerous or
diverse sources.

CRS-9
Air Scientific Advisory Committee to make recommendations to the Administrator
regarding new NAAQS and, at five-year intervals, to make reviews of existing
NAAQS with recommendations for revisions. In practice, EPA staff, not CASAC,
have prepared these reviews, drafting “criteria documents,” which review the science
and health effects of criteria air pollutants, and “staff papers,” which make policy
recommendations. CASAC’s role has been to review and approve these EPA
documents before they went to the agency’s political appointees and the
Administrator for final decisions.
Under the new procedures, the EPA’s political appointees will have a role early
in the process, helping to choose the scientific studies to be reviewed, and CASAC
will no longer have a role in approving the policy staff paper with its
recommendations to the Administrator. CASAC will be relegated to commenting on
the policy paper after it appears in the Federal Register, during a public comment
period. The goal, according to agency officials, is to speed up the review process,
which has consistently taken longer than the five years allowed by statute. “These
improvements, will help the agency meet the goal of reviewing each NAAQS on a
5-year cycle as required by the Clean Air Act, without compromising the scientific
integrity of the process,”18 according to the memorandum that finalized the changes.
The changes concern environmental groups and some in the scientific community,
however, because they appear to give a larger role to the agency’s political appointees
and a smaller role to EPA staff and CASAC.
As noted earlier, the Administrator is not required by the Clean Air Act to
follow CASAC’s recommendations; the Act ( Section 307(d)(3)) requires only that
the Administrator set forth in the Federal Register (in the NAAQS proposal) any
pertinent findings, recommendations, and comments by CASAC (and the National
Academy of Sciences) and, if his proposal differs in an important respect from any
of their recommendations, provide an explanation of the reasons for such differences.
However, the new procedures change the role that CASAC has historically played.
CASAC itself appears less concerned with the changes than some who are
advocating on its behalf. The committee does not plan to issue a formal response to
the December 7 memo. In response to a draft of the changes, the committee made
a number of suggestions, some of which, such as the convening of a science
workshop at the outset of the process to better focus the review, were incorporated
in the final memorandum. This appears to have addressed one of CASAC’s major
concerns, that the old process spent too much time compiling an encyclopedic review
of the literature, much of which had little relevance to the policy questions that
needed to be addressed. With respect to the EPA taking comments from CASAC at
the same time that it considers comments from the public, CASAC’s Chair is
18 “Process for Reviewing National Ambient Air Quality Standards,” Memorandum of
Marcus Peacock, Deputy EPA Administrator, to Dr. George Gray, Assistant Administrator,
Office of Research and Development, and Bill Wehrum, Acting Assistant Administrator,
Office of Air and Radiation, Dec. 7, 2006, p. 3, at [http://www.epa.gov/ttn/naaqs/memo_
process_for_reviewing_naaqs.pdf].

CRS-10
reported to have said, “[S]ome of the members were concerned but most are not,
because it doesn’t change CASAC’s ability to comment.”19
Reaction elsewhere has been stronger. Responding to the announced changes,
the incoming Chair of the Environment and Public Works Committee, Senator
Barbara Boxer, called them “unacceptable,” and said the committee plans to make
them a top priority for oversight in the 110th Congress.20 Seven Democratic members
of the committee, including Senator Boxer, wrote EPA Administrator Johnson to
express their strong opposition to the changes and to ask him to “abandon” them.21
Multi-pollutant Legislation for Power Plants
Besides air quality standards, the major focus of interest among members of
Congress and other policy makers concerned with air quality has been the regulation
of electric power plants. Coal-fired power plants are among the largest sources of air
pollution in the United States; however, under the Clean Air Act, they are not
necessarily subject to stringent requirements. Emissions and the required control
equipment can vary depending on the location of the plant, when it was constructed,
whether it has undergone major modifications, the specific type of fuel it burns, and,
to some extent, the vagaries of EPA enforcement policies. More than half a dozen
separate Clean Air Act programs could potentially be used to control emissions,
which makes compliance strategy complicated for utilities and difficult for
regulators. Because the cost of the most stringent available controls, for the entire
industry, could range into the tens of billions of dollars, utilities have fought hard and
rather successfully to limit or delay regulations affecting them, particularly with
respect to plants constructed before the Clean Air Act of 1970 was passed.
As a result, emissions from power plants have not been reduced as much as
those from some other sources. Many plants built in the 1950s and 1960s (generally
referred to as “grandfathered” plants) have little emission control equipment.
Collectively, these plants are large sources of pollution. In 2003, power plants
accounted for 10.2 million tons of sulfur dioxide (SO ) emissions (70% of the U.S.
2
total), about 45 tons of mercury emissions (more than 40% of the U.S. total), and 3.6
million tons of nitrogen oxides (19% of the U.S. total). Power plants are also
considered major sources of fine particles (PM ), many of which form in the
2.5
atmosphere from emissions from a wide range of stationary and mobile sources. In
addition, power plants account for about 40% of U.S. anthropogenic emissions of the
19 Comment of Dr. Rogene Henderson, CASAC Chair, in “EPA Adviser Plays Down
Democrats’ Criticism over New NAAQS Changes,” Inside EPA Clean Air Report, Dec. 14,
2006.
20 Office of Senator Barbara Boxer, “Boxer Statement on EPA’s Politicization of Clean Air
Health Standards,” Press Release, Dec. 8, 2006, at [http://boxer.senate.gov/news/releases/
record.cfm?id=266781].
21 Office of Senator Barbara Boxer, “Democratic Members of Senate EPW Committee Warn
EPA on Air Rollbacks,” Press Release, Dec. 21, 2006, at [http://boxer.senate.gov/news/
releases/record.cfm?id=267092].

CRS-11
greenhouse gas carbon dioxide; these emissions are not subject to federal regulation
but have been the focus of much debate in recent years.
With new ambient air quality standards for ozone and fine particles taking
effect, emissions of NOx (which contributes to the formation of ozone) and SO2
(which is among the sources of fine particles) would necessarily have to be reduced
to meet standards. Mercury emissions have also been a focus of concern: 44 states
have issued fish consumption advisories due to mercury pollution, covering 13
million acres of lakes, 765,000 river miles, and the coastal waters of 12 entire states.
The continuing controversy over the interpretation of New Source Review
requirements for existing power plants (discussed below) is also exerting pressure for
a more predictable regulatory structure.
Thus, many in industry, environmental groups, Congress, and the
Administration have said, for several years now, that the time is ripe for legislation
that addresses power plant pollution in a comprehensive (multi-pollutant) fashion.
Such legislation (the Administration version of which is entitled the Clear Skies
Act22) would address the major pollutants on a coordinated schedule and would rely,
to a large extent, on a system such as the one used in the acid rain program, where
national or regional caps on emissions are implemented through a system of tradeable
allowances. The key questions have been how stringent the caps should be and
whether carbon dioxide (CO ), the major gas of concern with regard to climate
2
change, will be among the emissions subject to a cap.
It is unclear what direction, if any, the new Congress will take regarding multi-
pollutant legislation, but bills introduced in previous Congresses have generally
fallen into three groups: (1) the Administration’s Clear Skies bill, which would
regulate three pollutants (SO , NOx, and mercury), give electric generators until 2018
2
to meet the bill’s final emission caps, allow trading of allowances for all three
pollutants, and remove or restrict numerous existing Clean Air Act requirements; (2)
Representative Waxman’s, Senator Leahy’s, and former Senator Jeffords’s bills,
which, although different from each other in many details, regulated four pollutants
(CO in addition to the other three), gave utilities less time (until 2009 or 2010) to
2
make reductions, set more stringent emission caps, did not allow trading of mercury
emission allowances, and generally left existing Clean Air Act requirements in place;
and (3) Senator Carper’s and former Representative Bass’s bills, which essentially
split the difference between the first two groups on the stringency and timing of SO ,
2
NOx, and mercury controls; established a CO control program (but less stringent
2
than the Waxman, Leahy, or Jeffords bills); and generally left existing Clean Air Act
requirements in place. (For additional information and a detailed comparison of
legislative proposals in the last Congress, see CRS Report RL32755, Air Quality:
Multi-Pollutant Legislation in the 109th Congress
, by Larry Parker and John
Blodgett.)
22 The Administration first proposed the Clear Skies Act on Feb. 14, 2002, and the bill was
introduced by request in the 107th Congress as H.R. 5266/S. 2815. In the 109th Congress,
a somewhat modified Clear Skies bill was introduced as S. 131.

CRS-12
Because the deadlines are far in the future, the Administration’s analysis of
Clear Skies has shown that utilities would be likely to “overcomply” in the early
years of the program. The Administration uses this as a selling point for its approach,
arguing that it will achieve reductions sooner than would a traditional regulatory
approach with the same deadlines. However, overcompliance in the early years
would lead to “banked” emission allowances; these could be used in later years to
delay achievement of required reductions. In its analysis of the bill, the EPA does
not expect to see the full 70% emission reductions that it requires until 2026 or later,
a point seized upon by its opponents to support a more aggressive approach.
As noted, the Clear Skies bill includes no cap on CO emissions. The
2
Administration has rejected mandatory controls on CO , in keeping with its
2
opposition to the Kyoto Protocol to the United Nations Framework Convention on
Climate Change. It opposes Kyoto for a variety of reasons, principally the potential
economic impacts on U.S. industries.
The absence of CO from the mix leads to different strategies for achieving
2
compliance, preserving more of a market for coal, and lessening the degree to which
power producers might switch to natural gas or renewable fuels as a compliance
strategy. In its opposition to CO controls, the Administration is supported by most
2
in the utility and coal industries. Others, mostly outside these industries but
including some utilities, view CO controls as inevitable, and perhaps desirable, and
2
support simultaneous implementation of cap-and-trade programs for CO and the
2
other pollutants.
The Senate Environment and Public Works Committee has voted twice on a
multi-pollutant bill, but none of the bills has progressed to the House or Senate floor.
On March 10, 2005, however, EPA announced that it would use existing Clean Air
Act authority to promulgate final regulations similar to the Clear Skies bill for utility
emissions of SO and NOx in 28 eastern states and the District of Columbia.23 The
2
Clean Air Interstate Rule (CAIR) established cap-and-trade provisions that mimic
those of Clear Skies, but the regulations cover only the eastern half of the country,
and, as a regulation, CAIR has no authority to allow the EPA to remove existing
Clean Air Act requirements, as Clear Skies would. Under CAIR, the EPA projects
that nationwide emissions of SO will decline 53% by 2015 and NOx emissions will
2
decline 48%. The agency also projects that the rule will result in $85-$100 billion
in health benefits annually by 2015, including the prevention of 17,000 premature
deaths annually.24 CAIR’s health and environmental benefits are more than 25 times
greater than its costs, according to the EPA. Similarly, any of the multi-pollutant
bills are expected to have benefits far outweighing their costs. (For additional
information on the CAIR rule, see CRS Report RL32927, Clean Air Interstate Rule:
Review and Analysis
, by Larry Parker. For a discussion of the costs and benefits of
the principal multi-pollutant approaches, see CRS Report RL33165, Costs and
Benefits of Clear Skies: EPA’s Analysis of Multi-Pollutant Clean Air Bills
, by James
E. McCarthy and Larry B. Parker.)
23 The rule appeared in the Federal Register on May 12, 2005 (70 FR 25162).
24 U.S. EPA, Office of Air and Radiation, “Clean Air Interstate Rule — Basic Information,”
available at [http://www.epa.gov/interstateairquality/basic.html].

CRS-13
Mercury From Power Plants
On March 15, 2005, the EPA also finalized through regulation a cap-and-trade
program for mercury emissions from electric utilities.25 The mercury regulations
(which, like CAIR, are based on the Clear Skies approach) rely almost entirely on co-
benefits of regulating SO and NOx. The agency’s analysis of the mercury rule finds
2
that less than 1% of coal-fired power plant capacity would install pollution control
equipment specifically designed to control mercury within 10 years as a result of the
mercury rule. By 2020, only 4% of capacity would have such equipment.
The EPA was required by the terms of the 1990 Clean Air Act Amendments and
a 1998 consent agreement to determine whether regulation of mercury from power
plants under Section 112 of the Clean Air Act was appropriate and necessary. It
concluded that it was, in a December 2000 regulatory finding. The finding triggered
other provisions of the consent agreement: that the agency propose Maximum
Achievable Control Technology (MACT) standards for electric power plants by
December 15, 2003, and finalize them by March 15, 2005.
The December 2003 proposal offered two alternatives. The first met the
agency’s requirement under the consent agreement by proposing MACT standards.
The standards would have applied on a facility-by-facility basis and would have
resulted in emissions of 34 tons of mercury annually, a reduction of about 30% from
the 1999 level. The standards would have taken effect in 2008, three years after
promulgation, with possible one-year extensions.
The second alternative uses Section 111(d) of the act. To avoid having to
promulgate MACT standards, which would set limits for each individual facility, the
agency proposed reversing its December 2000 regulatory finding, arguing that
although MACT standards were “appropriate,” they were not “necessary” because
the emissions could be controlled under Section 111(d) instead, a section that allows
a more flexible approach, such as a cap-and-trade program. Section 111(d) has rarely
been used before — and never for hazardous air pollutants.
The final regulations, promulgated March 15, 2005, chose the second approach,
establishing a national cap-and-trade system for power plant emissions of mercury.
The cap will be 15 tons of emissions nationwide in 2018 (about a 70% reduction
from 1999 levels, when achieved). There will also be an intermediate cap of 38 tons
in 2010. The caps will be implemented through an allowance system similar to that
used in the acid rain and CAIR programs, through which utilities can either control
the pollutant directly or purchase excess allowances from other plants that have
instituted controls more stringently or sooner than required. As with the acid rain and
CAIR programs, early reductions can be banked for later use, which the agency says
25 The mercury rule appeared in the Federal Register in two parts: in the first part, on
March 29, 2005 (as explained further in the text below), the agency revised its determination
that mercury emissions from electric generating units should be regulated as hazardous air
pollutants under Section 112 of the Clean Air Act (70 FR 15994); in the second part, on
May 18, 2005, the agency promulgated a cap-and-trade program under Section 111 of the
act (70 FR 28606).

CRS-14
will result in utilities delaying compliance with the full 70% reduction until well
beyond 2018, as they use up banked allowances rather than installing further controls.
The agency’s analysis projects actual emissions to be 24.3 tons (less than a 50%
reduction) as late as 2020. Full compliance with the 70% reduction would be delayed
until after 2025.26 (For additional information on the mercury rule, see CRS Report
RL32868, Mercury Emissions from Electric Power Plants: An Analysis of EPA’s
Cap-and-Trade Regulations
, by James E. McCarthy.)
Besides the stretched out implementation schedule, one of the main criticisms
of the cap-and-trade proposal is that it would not address “hot spots,” areas where
mercury emissions and/or concentrations in water bodies are greater than elsewhere.
It would allow a facility to purchase allowances and avoid any emission controls, if
that compliance approach makes the most sense to the plant’s owners and operators.
If plants near hot spots do so, the cap-and-trade system may not reduce mercury
concentrations in the most contaminated areas. By contrast, a MACT standard would
have required reductions at all plants, and would therefore be expected to improve
conditions at hot spots.
Many argue that the mercury regulations should be more stringent or
implemented more quickly. To a large extent, these arguments, and EPA’s counter-
arguments, rest on assumptions concerning the availability of control technologies.
Controlling SO , NOx, and mercury simultaneously, as the agency prefers, would
2
allow utilities to maximize “co-benefits” of emission controls. Controls such as
scrubbers and fabric filters, both of which are widely used today to control SO and
2
particulates, have the side effect of reducing mercury emissions to some extent.
Under EPA’s cap-and-trade regulations, both the 2010 and 2018 mercury emission
standards are set to maximize use of these co-benefits. Thus, few controls would be
required to specifically address mercury emissions before the 2020s, the costs
specific to controlling mercury would be minimal, and emissions would decline to
about 50% of the 1999 level in 2020.
Besides citing the cost advantage of relying on co-benefits, the EPA has claimed
that technology specifically designed to control mercury emissions (such as activated
carbon injection, ACI) would not be generally available until after 2010. This
assertion is widely disputed. ACI and fabric filters have been in use on municipal
waste and medical waste incinerators for a decade, and have been successfully
demonstrated in at least 16 full-scale tests at coal-fired power plants, for periods as
long as a year. Manufacturers of pollution controls and many others maintain that
if the agency required the use of ACI and fabric filters at power plants, reductions in
mercury emissions as great as 90% could be achieved at reasonable cost in the near
future. Relying on these assertions, about 20 states have promulgated requirements
stricter than the federal program, with several requiring 80% to 90% mercury
reductions before 2010. (For additional information, see CRS Report RL33535,
26 U.S. EPA, Office of Air Quality Planning and Standards, Regulatory Impact Analysis of
the Clean Air Mercury Rule
, March 2005, Table 7-3, p. 7-5, at [http://www.epa.
gov/ttn/atw/utility/ria_final.pdf]. For further discussion, see CRS Report RL32868,
Mercury Emissions from Electric Power Plants: An Analysis of EPA’s Cap-and-Trade
Regulations
, by James E. McCarthy.

CRS-15
Mercury Emissions from Electric Power Plants: States Are Setting Stricter Limits,
by James E. McCarthy.)
It is unclear whether the EPA has legislative authority to establish a cap-and-
trade program for mercury: many argue that the agency is required by the statute to
impose MACT standards on each individual plant once it has decided to control
mercury emissions. Questions also have arisen regarding the role of industry
lobbyists in crafting portions of the EPA proposal. For many of these reasons, 45
Senators wrote EPA Administrator Leavitt at the beginning of April 2004 to request
that he withdraw the mercury proposal and start over. In June, 2004, 178 House
members wrote Leavitt that they hoped further review “will lead to a stronger final
rule.” On February 3, 2005, the EPA Inspector General echoed these comments,
concluding that EPA senior management instructed the staff to develop a standard
that would result in emissions of 34 tons annually, instead of basing the standard on
unbiased analysis. Nevertheless, the agency acted to make the final rule less stringent
rather than strengthening it.27 The agency’s cost-benefit analysis also did not include
several peer-reviewed studies that indicated stricter utility mercury rules would have
yielded large benefits. Thus, opponents, including at least 15 states, have filed suit
to overturn the rule.28
In September 2005, the Senate considered, but narrowly defeated, a challenge
to the rule under the Congressional Review Act.29
New Source Review
A related issue that has driven some of the debate over the regulation of power
plant emissions is whether the EPA has adequately enforced existing regulations,
using a process called New Source Review (NSR). The New Source Review debate
has occurred largely in the courts. The EPA took a more aggressive stance on NSR
under the Clinton Administration, filing lawsuits against 13 utilities for violations at
51 plants in 13 states. The Bush Administration has taken action against an
additional half a dozen utilities but has made little headway in settling the original
suits or in bringing them to trial. In the meantime, it has proposed major changes in
the NSR regulations that critics argue will weaken or eliminate New Source Review
as it pertains to modifications of existing plants.
The controversy over the NSR process stems from the EPA’s application of
New Source Performance Standards to existing stationary sources of air pollution that
have been modified. The Clean Air Act requires that plants undergoing
27 Office of the Inspector General, U.S. EPA, Additional Analyses of Mercury Emissions
Needed Before EPA Finalizes Rules for Coal-Fired Electric Utilities
, Feb. 3, 2005, p. 10,
available at [http://www.epa.gov/oig/reports/2005/20050203-2005-P-00003.pdf].
28 New Jersey v. EPA, No. 05-1097 (D.C. Cir.) Filed Mar. 29, 2005.
29 For discussion of the Congressional Review Act and how it applied to the mercury rule,
see CRS Report RS22207, Congressional Review of EPA’s Mercury Rule, by James
McCarthy and Richard Beth.

CRS-16
modifications meet NSR requirements by installing best available pollution controls,
but industry has often avoided the NSR process by claiming that changes to existing
sources were “routine maintenance” rather than modifications. In the 1990s, the EPA
began reviewing records of electric utilities, petroleum refineries, and other industries
to determine whether the changes were, in fact, routine. As a result of these reviews,
since late 1999, the EPA and the Department of Justice have filed suit or
administrative actions against numerous large sources of pollution, alleging that they
made major modifications to their plants, extending plant life and increasing output,
without undergoing required New Source Reviews and without installing best
available pollution controls.
Of the utilities charged with NSR violations, 11 have settled with the EPA,
generally without going to trial. Under the settlements, they have agreed to spend
about $5 billion over the next decade on pollution controls or fuel switching to
reduce emissions at their affected units. Combined, these companies will reduce
pollution by about 775,000 tons annually. Since July 25, 2000, the agency has also
reached 17 agreements with petroleum refiners representing three-fourths of industry
capacity. The refiners agreed to settle potential charges of NSR violations by paying
fines and installing equipment to eliminate 315,000 tons of pollution.
About half the utilities charged with NSR violations have not settled with the
EPA. They and other critics of the agency’s enforcement actions claim that the EPA
reinvented the rules. They also contend that a strict interpretation of what constitutes
routine maintenance will prevent them from making changes that would have
previously been allowed without a commitment of time and money for permit
reviews and the installation of expensive pollution control equipment. This provides
disincentives for power producers, refiners, and others to expand output at existing
facilities, they maintain.
The first case involving one of the nonsettling utilities went to trial in February
2003. In an August 7, 2003, decision, the U.S. District Court for the Southern
District of Ohio found that Ohio Edison had violated the Clean Air Act 11 times in
modifying its W. H. Sammis power plant. The company subsequently settled the
case, agreeing to spend $1.1 billion to install controls that are expected to reduce
pollution by 212,000 tons annually.30 In a second case, decided in April 2004 and
currently on appeal to the U.S. Supreme Court, Duke Energy was found not to have
violated the act despite undertaking modifications that increased total emissions
without undergoing New Source Review. The U.S. District Court for the Middle
District of North Carolina, in a decision upheld by the Fourth Circuit Court of
Appeals, held that since the maximum hourly emissions rate did not increase as a
result of the modifications, even if annual emissions did increase, the company was
not required to undergo NSR and install more stringent pollution controls.31
While pursuing these enforcement actions, the Bush Administration has
promulgated a number of changes to the NSR regulations that would make future
30 United States v. Ohio Edison Co., No. C-2-99-1181, [S.D. Ohio].
31 United States v. Duke Energy Corp., 278 F.Supp. 2d 619 [M.D.N.C. 2003] affirmed, 411
F. 3d 539 [4th Cir., 2005], petition for cert. Filed [No. 05-848].

CRS-17
enforcement of NSR less likely. In December 2002 and October 2003, the agency
promulgated five sets of changes to the NSR rules. The most controversial were new
regulations defining what constitutes routine maintenance.32 The new regulations
would have exempted industrial facilities from undergoing NSR (and thus from
installing new emission controls) if they were replacing safety, reliability, and
efficiency-rated components with new, functionally equivalent equipment, and if the
cost of the replacement components was less than 20% of the replacement value of
the process unit. Using this benchmark, few, if any, plant modifications would
trigger new pollution controls.
These changes were highly controversial. The Administration and its supporters
characterized them as streamlining or improving the program; others saw them as
permanently “grandfathering” older, more polluting facilities from ever having to
meet the clean air standards required of newer plants. Fifteen states, three
municipalities, and several environmental groups filed suit to block the “equipment
replacement / routine maintenance” rule. The rule was stayed by the U.S. Court of
Appeals for the D.C. Circuit on December 24, 2003. On March 17, 2006, a three-
judge panel of the court unanimously struck the rule down. In its decision, the court
held that the EPA’s attempt to change the NSR regulations was “contrary to the plain
language” of the Clean Air Act.33
The EPA proposed further changes to the NSR regulations on October 20, 2005,
and September 8, 200634; these regulations have yet to be promulgated. Under the
October 2005 proposal, power plants could modify existing facilities without
triggering NSR, provided that the facility’s “maximum hourly emissions achievable”
after the changes were no greater than the same measure at any point during the past
five years. By focusing on the hourly rate, rather than the previous measure (annual
emissions), the new rule would effectively allow increases in annual emissions any
time a modification led to an increase in the hours of operation of a facility. The
agency’s proposal stated that this change would establish a uniform national
emissions test, in conformance with the Fourth Circuit’s decision in the Duke Energy
case, and it downplayed the significance of the change in light of “substantial
emissions reductions from other CAA [Clean Air Act] requirements that are more
efficient.” But internal EPA documents released by an environmental group indicate
that the proposed rule was strongly opposed by the Air Enforcement Division, whose
Director concluded that it would adversely affect the agency’s NSR enforcement
cases and is largely unenforceable as written.35
32 These changes appeared in the Federal Register on Oct. 27, 2003 (68 FR 61247).
33 State of New York v. EPA, No. 03-1380, 2006 Westlaw 662746 [D.C. Cir., Mar. 17,
2006].
34 70 FR 61081, Oct. 20, 2005. The September 2006 proposal had not yet appeared in the
Federal Register as of this writing, but it is available on the EPA’s website at
[http://www.epa.gov/nsr/documents/dapn_analysis_9-8-06.pdf]. It would limit application
of NSR by allowing plants to consider emissions only from the unit undergoing
modification, rather than the entire plant, in determining whether NSR applies.
35 Memorandum of Adam M. Kushner, Director, Air Enforcement Division, U.S. EPA, to
(continued...)

CRS-18
Thus, there appears to be a conflict between the EPA’s regulatory actions and
its enforcement stance. While the agency stated in promulgating the equipment
replacement rule that “we do not intend our actions today to create retroactive
applicability for today’s rule,” continued pursuit of the enforcement actions filed
during the Clinton Administration would create a double standard for utilities, with
one set of rules applicable to those utilities unlucky enough to have been cited for
violations prior to promulgation of the new rule, and a different standard applicable
afterward. Despite earlier agency denials that the rule would affect ongoing
investigations, in early November 2003, the EPA’s enforcement chief, J. P. Suarez,
and another EPA official were reported to have indicated that the agency would drop
enforcement actions against 47 facilities that had already received notices of
violation, and would drop investigations of possible violations at an additional 70
power companies. Agency staff who were involved in the enforcement actions note
that the prospect of an NSR rollback caused utilities already charged with violations
to withdraw from settlement negotiations over the pending lawsuits, delaying
emission reductions that could have been achieved in the near future.36 (For
additional information, see CRS Report RS21608, Clean Air and New Source
Review: Defining Routine Maintenance
, and CRS Report RL31757, Clean Air: New
Source Review Policies and Proposals
, both by Larry Parker.)
At Congress’s direction, the National Academy of Sciences began a review of
the NSR program in May 2004. An interim report, released in January 2005, said the
committee had not reached final conclusions, but it also said, “In general, NSR
provides more stringent emission limits for new and modified major sources than
EPA provides in other existing programs” and “It is ... unlikely that Clear Skies
would result in emission limits at individual sources that are tighter than those
achieved when NSR is triggered at the same sources.”37 The final report, issued July
21, 2006, found that
[m]ore than 60% of all coal-fired electricity-generation capacity in the United
States currently lacks the kinds of controls for SO and NO emissions that have
2
x
been required under NSR. Also, the older facilities are more likely than newer
facilities to undergo maintenance, repair, and replacement of key components,
so a substantial portion of emissions from the electricity-generating sector is
potentially affected by the NSR rule changes.38
35 (...continued)
William Harnett, Director, Information Transfer and Program Integration Division, Office
of Air Quality Planning and Standards, Aug. 25, 2005, p. 1.
36 See, for example, “Departing EPA Official Issues Broadside at Administration Air,
Enforcement Programs,” Daily Environment Report, Mar. 1, 2002, p. AA-1. Also, “Second
Former EPA Enforcement Official Raps Bush’s New Source Review Reforms,” Daily
Environment Report
, Oct. 22, 2002, p. A-9.
37 National Research Council of the National Academies, Interim Report of the Committee
on Changes in New Source Review Programs for Stationary Sources of Air Pollutants
(Washington, DC: The National Academies Press, 2005), p. 27.
38 National Research Council of the National Academies, New Source Review for Stationary
Sources of Air Pollutants
(Washington, DC: The National Academies Press, 2006),
(continued...)

CRS-19
Nevertheless, the report reached ambivalent conclusions. On the one hand, the report
stated, “It is reasonable to conclude that the implementation of the ERP [the proposed
Equipment Replacement Provision] could lead to SO and NO emission increases
2
x
in some locations and decreases in others.”39 On the other hand,

the committee concluded overall that, because of a lack of data and the
limitations of current models, it is not possible at this time to quantify with a
reasonable degree of certainty the potential effects of the NSR rule changes on
emissions, human health, energy efficiency, or on other relevant activities at
facilities subject to the revised NSR program.40
Besides the NAS study, on April 21, 2003, the National Academy of Public
Administration released a report commissioned by Congress that made sweeping
recommendations to modify NSR. The study panel recommended that Congress end
the “grandfathering” of major air emission sources by requiring all major sources that
have not obtained an NSR permit since 1977 to install Best Available Control
Technology or Lowest Achievable Emissions Rate control equipment. In the interim,
the NAPA panel concluded, the EPA and the Department of Justice should continue
to enforce NSR vigorously, especially for changes at existing facilities.41
38 (...continued)
Prepublication Copy, p. 3.
39 Ibid., p. 5.
40 Ibid., p. 2.
41 National Academy of Public Administration, A Breath of Fresh Air: Reviving the New
Source Review Program
, Summary Report, April 2003, p. 3.