"Vacillating" and "inconsistent" are the kindest terms that one can apply to the federal government's position on New Source Review (NSR) under the Clean Air Act (CAA) in recent years. Underlying these inconsistencies is an ongoing battle between some members
of the U.S. Environmental Protection Agency's (EPA) career staff and the agency's political leadership over the right to make policy - or in legal terms, over who exercises the agency's Chevron discretion.1
Today, we have the Kafkaesque spectacle of the federal government continuing a Clinton-era enforcement initiative to win billions of dollars from a few utility, refiners and manufacturing companies for allegedly "violating the law," while the same federal government (1) repeatedly reassured them that their repairs and replacements of equipment were legal at the time, (2) now again interprets the law to permit repairs of older plants, and (3) announces that no other enforcement cases will be brought against their competitors for doing the same things in the past. The law was created to prevent such arbitrary and inconsistent actions by rulers.2
The fundamental idea underlying NSR is a simple one: As new plants are built, they should incorporate the best pollution controls available. NSR benefits health only incidentally, but that is not the program's purpose or focus.
The program of "technology-based" NSR standards to supplement health-based standards elsewhere in the CAA has proved relatively uncontroversial when applied to truly new plants. The difficulty has come in applying the norm of "best available control technology" to existing plants. Technical amendments in 1977 that were never even debated on the floor of Congress extended the concept of applying the NSR-mandate for best available control technology to "modified" as well as new plants.3 For years, the concept of a "modified" plant was relatively clear: a plant was modified so as to trigger an obligation to install best available control technology if, but only if, changes were made that actually "increased emissions." EPA repeatedly defined the concept of increasing emissions to exclude things such as routine maintenance and replacement of broken equipment that did not cause the plant's emissions at its maximum design capacity to increase.
This stable situation continued until the late 1980s, when EPA staff began to advance new legal theories in litigation without appropriate review by the agency's policy-makers or the public.These innovative legal theories argued that either a "potential" increase in emissions would count or that an "incentive" to run a plant more often could count. After EPA lost "potential" in the early 1990s, it started to argue "restored capacity" and "increased utilization" as a backup. Those innovative legal theories were an attempt to make a fundamental change in the nature of the NSR program, but they were never approved or promulgated by EPA's leadership, much less reviewed by the Office of Management and Budget (OMB) or the White House, nor did they go through notice and comment rulemaking in the Federal Register. I should know; I had just been named, but was not yet confirmed, as General Counsel of EPA, and did not even know about the position that "my" agency was taking in court in the fateful WEPCO case until after the decision came down. The 1990 WEPCO decision4 by the well-respected Seventh Circuit Court of Appeals in Chicago partly rejected and partly accepted EPA's new legal theories. However, the WEPCO opinion was explicitly based on granting deference to EPA's litigation position and lower-level staffinterpretations,5 a view that seemed correct at the time, but has subsequently been squarely rejected by the Supreme Court.6
The case-by-case multifactor test that EPA argued for in its brief was incredibly bad policy that would never have survived a notice-and-comment rulemaking. Because this "test" involves weighing and balancing multiple factors on a case-by-case basis, it is virtually impossible for a company to determine whether any particular project triggers NSR. It is as if taking a tax deduction, subject to criminal penalties, depended on weighing and balancing six or seven factors on a case-by-case basis. The unpredictability created by WEPCO was condemned by both houses of Congress on both sides of the aisle during the 1990 CAA Amendments.7, 8 The first Bush administration then advised Congress that it intended to fix the WEPCO problem by creating safe harbors through an administrative rule by "exclud[ing] emission increases
due to demand growth and other factors unrelated to the modification."9 Congress accepted the proposed administrative fix so that there could be "a quick resolution of the WEPCO matter by EPA as appropriate."10, 11 That rule was duly promulgated in the Federal Register in July 1992 after review by OMB, and notice and comment by the public. The 1992 post-WEPCO rule12 restored a relatively stable and predictable situation, and was accompanied by a variety of public statements and assurances from the highest levels of EPA - including the EPA Administrator personally - that utilities could continue repairs and life extension projects at their plants without triggering NSR.13 If the official 1992 EPA interpretation had remained law, none or virtually none of the pending NSR enforcement cases could have been brought. In 1998, just prior to bringing the present NSR enforcement initiative, and without notice and comment rulemaking, the Clinton administration EPA quietly disavowed a key part of the 1992 rule, clearing the way to bring the current cases.14
The recent actions by the current Bush administration to reinterpret clear "safe harbors" back into the NSR program are clearly both good policy and politically courageous.15 They laudably attempt to restore the stability and predictability that is necessary to maintain
the reliability of our electricity supply and had been in the NSR program historically until some EPA staffers - illegitimately, in my view - began to write aggressive new interpretations into the
agency's briefs and interpretations without proper review and public participation. Unfortunately, however, three Democratic-appointees of the D.C. Circuit recently entered a stay of the safe harbor rule,16 creating further confusion as to which interpretation of NSRnow applies.The lower court decisions are also in disarray.As I said in my 2002 Senate testimony, "EPA's many changing interpretations of NSR over the years have created a legal mess of baffling complexity."Paradoxically, however, the administration has so far felt compelled - apparently by political considerations - to continue to prosecute a few companies under the very legal theories that it and most of its predecessors have disavowed. These companies merely did what they and their competitors will again be free to do in the future and what most past administrations had repeatedly told the world that it was legal to do at the time the acts occurred.
The key legal issue in these cases is who speaks for EPA (or in law talk, who controls the agency's Chevron deference17 to interpret the law). Is it the staff lawyers, who without any public participation or oversight write the briefs and argue the individual enforcement cases, or is it the supposedly higher levels of the agency, who are supposed to be in control because they were appointed by the President, confirmed by the Senate, and complied with the Administrative Procedure Act by going through notice and comment procedures before promulgating their interpretations in the Federal Register? (In legal-speak, this is the Mead issue, after a Supreme Court case18 that decided that litigation positions and statements by individual agency employees do not get much deference in court.) On this pivotal legal issue, Judge Bullock has clearly got the law right in United States vs. Duke Energy19 by deferring to EPA's official statements in the Federal Register, and Judge Sargus clearly got it wrong in United States vs. Ohio Edison20 by deferring to an interpretation first announced in enforcement litigation.Judge Bullock's sensible interpretation would return clarity and predictability to a program that badly needs it by returning NSR law to the positions that EPA has taken throughout most of its history, rather than the more recent, aggressive interpretations that staff is attempting to impose retroactively.
1Drew, C.; Oppel Jr., R.A. Lawyers at EPA Say It Will Drop Pollution Cases; The New York Times, Nov. 6, 2003, p. A-1. More evidence of the rearguard actions by some EPA attorneys to undermine the President's policies was provided by recent leaks to The New York Times, criticizing the decision not to bring more NSR cases.
2Ferullo, M., EPA's Enforcement Chief Defends Review of Existing New Source Review Investigations; BNA Daily Report for Executives Nov. 21, 2003. While EPAintends to drop cases that it has not yet filed "in light of the policy that is reflected in the new rule," quite inconsistently EPA maintains that it will nonetheless prosecute vigorously those cases it has already filed involving identical conduct.
3123 Cong. Rec. H36327-36334, S36250-36259 (Nov. 1, 1977).
4Wisconsin Electric Power Co. (WEPCO) vs. Reilly, 893 F.2d 901 (7th Cir. 1990).
5Ibid. at p. 915.
6United States vs. Mead Corp., 533 U.S. 218 (2001).
7H.Rep. 3030, LEXIS 1990 CAA Leg. Hist. 3021, 3685.
8S. Rep. 1630, LEXIS 1990 CAA Leg. Hist. 6946, 6951.
9Letter from Richard Schmalensee, Council of Economic Advisers, to Senator Wendall Ford, Chairman, Subcommittee on Energy Research and Development, Committee on Energy and Natural Resources (Sept. 11, 1990).
10Conference Report, LEXIS 1990 CAA Leg.Hist. 1451, 1794.
11Remarks, LEXIS 1990 CAA Leg.Hist. 10726, 10763.
12Fed. Regist. 1992, 57, 32314.
13Letter from William K. Reilly, EPA Administrator, to Congressman John D. Dingell, Chairman, Subcommittee on Oversight and Investigations, House Energy and Commerce Committee (April 19, 1989).
14Fed. Regist. 1998, 63, 39860.
15Fed. Regist. 2003, 68, 61248.
16NY v. EPA, No. 03-1380 (D.C. Cir., Dec. 24, 2003).
17Chevron vs. NRDC, 467 U.S. 837 (1984).
18United States vs. Mead Corp., 533 U.S. 218 (2001).
19United States vs. Duke Energy Corp., 2003 WL 22024780 (M.D.N.C. 2003).
20United States vs. Ohio Edison Co., 2003 WL 21910738 (S.D. Ohio 2003).
Published April 1, 2004.