Coalition for Responsible Reg v. EPA

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT Filed: December 20, 2012 No. 09-1322 COALITION FOR RESPONSIBLE REGULATION, INC., ET AL., PETITIONERS v. ENVIRONMENTAL PROTECTION AGENCY, RESPONDENT STATE OF MICHIGAN, ET AL., INTERVENORS Consolidated with 10-1024, 10-1025, 10-1026, 10-1030, 10-1035, 10-1036, 10-1037, 10-1038, 10-1039, 10-1040, 10-1041, 10-1042, 10-1044, 10-1045, 10-1046, 10-1234, 10-1235, 10-1239, 10-1245, 10-1281, 10-1310, 10-1318, 10-1319, 10-1320, 10-1321 No. 10-1073 COALITION FOR RESPONSIBLE REGULATION, INC., ET AL., PETITIONERS v. ENVIRONMENTAL PROTECTION AGENCY, RESPONDENT 2 AMERICAN FROZEN FOOD INSTITUTE, ET AL., INTERVENORS Consolidated with 10-1083, 10-1099, 10-1109, 10-1110, 10-1114, 10-1118, 10-1119, 10-1120, 10-1122, 10-1123, 10-1124, 10-1125, 10-1126, 10-1127, 10-1128, 10-1129, 10-1131, 10-1132, 10-1145, 10-1147, 10-1148, 10-1199, 10-1200, 10-1201, 10-1202, 10-1203, 10-1206, 10-1207, 10-1208, 10-1210, 10-1211, 10-1212, 10-1213, 10-1216, 10-1218, 10-1219, 10-1220, 10-1221, 10-1222 No. 10-1092 COALITION FOR RESPONSIBLE REGULATION, INC., ET AL., PETITIONERS v. ENVIRONMENTAL PROTECTION AGENCY, RESPONDENT LANGBOARD, INC. - MDF, ET AL., INTERVENORS Consolidated with 10-1094, 10-1134, 10-1143, 10-1144, 10-1152, 10-1156, 10-1158, 10-1159, 10-1160, 10-1161, 10-1162, 10-1163, 10-1164, 10-1166, 10-1182 3 No. 10-1167 AMERICAN CHEMISTRY COUNCIL, PETITIONER v. ENVIRONMENTAL PROTECTION AGENCY AND LISA PEREZ JACKSON, ADMINISTRATOR, U.S. ENVIRONMENTAL PROTECTION AGENCY, RESPONDENTS CHAMBER OF COMMERCE OF THE UNITED STATES OF AMERICA, ET AL., INTERVENORS Consolidated with 10-1168, 10-1169, 10-1170, 10-1173, 10-1174, 10-1175, 10-1176, 10-1177, 10-1178, 10-1179, 10-1180 On Petitions for Rehearing En Banc ______ Before: SENTELLE*, Chief Judge, and HENDERSON, ROGERS*, TATEL*, GARLAND, BROWN*, GRIFFITH, and KAVANAUGH*, Circuit Judges. 4 ORDER The petition of the Chamber of Commerce of the United States of America, joined by the State of Alaska, Peabody Energy Company, Southeastern Legal Foundation, et al., State Petitioners and Intervenors for Petitioners, for rehearing en banc; and the petition of the National Association of Manufacturers, et al. for rehearing en banc in No. 10-1073, et al. and No. 10-1167, et al., and the responses to the petitions were circulated to the full court, and a vote was requested. Thereafter, a majority of the judges eligible to participate did not vote in favor of the petitions. Upon consideration of the foregoing, it is ORDERED that the petitions be denied. FOR THE COURT: Mark J. Langer, Clerk BY: /s/ Jennifer M. Clark Deputy Clerk *Circuit Judges Brown and Kavanaugh would grant the petitions for rehearing en banc. * A statement by Chief Judge Sentelle and Circuit Judges Rogers and Tatel, concurring in the denials of rehearing en banc, is attached. * A statement by Circuit Judge Brown, dissenting from the denials of rehearing en banc, is attached. *A statement by Circuit Judge Kavanaugh, dissenting from the denials of rehearing en banc, is attached. SENTELLE, Chief Judge, ROGERS, Circuit Judge, and TATEL, Circuit Judge, concurring in the denials of rehearing en banc: In dissenting from the denials of rehearing en banc, Judge Brown primarily takes issue with EPA’s Endangerment Finding. But as she candidly acknowledges, see Dissenting Op. at 2 (Brown, J.), her quarrel is with the Supreme Court. In Massachusetts v. EPA, 549 U.S. 497 (2007), the Court expressly held that the Clean Air Act’s “sweeping definition of ‘air pollutant’ ” unambiguously includes greenhouse gases. See id. at 528–29. Moreover, in so holding, the Court expressly rejected many of the arguments her dissent now presses. In particular, it rebuffed EPA’s attempt to use “postenactment congressional actions and deliberations” to obscure “the meaning of an otherwise-unambiguous statute,” id. at 529, and found EPA’s reliance on FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120 (2000), “similarly misplaced,” Massachusetts v. EPA, 549 U.S. at 530. Seeking to revive the Brown & Williamson argument, Judge Brown suggests that the Court never considered the “far-reaching effects” of extending greenhouse gas regulation to stationary sources. See Dissenting Op. at 18 (Brown, J.). But this is inaccurate—the briefs before the Court explicitly raised the argument that interpreting “air pollutant” to include greenhouse gases could have tremendous consequences for stationary-source regulation. See, e.g., Brief of Respondent CO2 Litigation Group, Massachusetts v. EPA, 549 U.S. 497 (2007) (No. 05-1120), 2006 WL 3043971 at *19–*31. To the extent Judge Brown attempts to bypass Massachusetts v. EPA by focusing on the statutory condition that air pollution “reasonably be anticipated to endanger public health or welfare,” 42 U.S.C. § 7521(a)(1) (emphasis added), her quarrel is not just with the Supreme Court, but also with EPA’s assessment of the science. Of course, we agree that the statute requires EPA to find a particular causal nexus between the pollutant and the harm in order to regulate. See Dissenting Op. at 9 (Brown, J.). But that is exactly what 2 EPA did: it found that “greenhouse gases in the atmosphere may reasonably be anticipated both to endanger public health and to endanger public welfare.” Endangerment and Cause or Contribute Findings for Greenhouse Gases Under Section 202(a) of the Clean Air Act, 74 Fed. Reg. 66,496, 66,497 (Dec. 15, 2009). And, as the panel opinion explains, EPA’s scientific judgment about the causal relationship between greenhouse gases and climate change is a scientific determination entitled to “an extreme degree of deference.” Coalition for Responsible Regulation v. EPA, 684 F.3d 102, 120 (D.C. Cir. 2012) (quoting American Farm Bureau Federation v. EPA, 559 F.3d 512, 519 (D.C. Cir. 2009)). The dissent's suggestion that EPA was somehow statutorily precluded from finding the requisite nexus between greenhouse gases and harm to public health and welfare, see Dissenting Op. at 10–11 (Brown, J.), is belied by the Supreme Court's decision to remand precisely this question. See Massachusetts v. EPA, 549 U.S. at 532–35. Judge Kavanaugh’s dissent relates to the scope of the Prevention of Significant Deterioration (“PSD”) program, an aspect of the panel opinion Judge Brown also rejects. Specifically, Judge Kavanaugh disagrees with EPA’s longstanding interpretation of the term “any air pollutant,” 42 U.S.C. § 7479(1), arguing that, in the context of the PSD program, “any air pollutant” refers not to all pollutants regulated under the Clean Air Act, but only to the six NAAQS pollutants. Because taking the statute at its word and interpreting “any air pollutant” to include greenhouse gases would lead to what he considers absurd results, Judge Kavanaugh insists that EPA and this Court are obligated to read “any air pollutant” more narrowly. See Dissenting Op. at 3–10 (Kavanaugh, J.). This argument, however, hinges on the proposition that both readings are plausible interpretations of an ambiguous statutory provision. See Dissenting Op. at 2–3, 3 10 (Kavanaugh, J.). But as the panel opinion explains at length, the statute is clear. See Coalition for Responsible Regulation, 684 F.3d at 132–44. Congress did not say “certain ‘air pollutants.’ ” Dissenting Op. at 2 (Kavanaugh, J.). It said “any air pollutant,” and it meant it. See Coalition for Responsible Regulation, 684 F.3d at 136. Thus, unlike the unreasonable interpretation rejected in Kloeckner v. Solis, No. 11-184, slip op. at 7–13 (U.S. 2012), the panel’s interpretation of the statute is the only plausible one. Moreover—and again, as the panel opinion explains at length, see Coalition for Responsible Regulation, 684 F.3d at 135–36—considering “any air pollutant” in context buttresses rather than undermines the panel’s interpretation. The statute frames the purpose of the PSD program in broad—not NAAQS-specific—terms, emphasizing that the program’s goal is “to protect public health and welfare from any actual or potential adverse effect which . . . may reasonably be anticipate[d] to occur from air pollution.” 42 U.S.C. § 7470(1). And although certain aspects of the program are specifically directed at NAAQS pollutants, see, e.g., id. § 7473(b)(4), the program as a whole plainly has a more expansive scope. For instance, covered sources are required to (1) install the best available control technology for “each pollutant subject to regulation under [the Act],” id. § 7475(a)(4) (emphasis added), and (2) demonstrate that they will not cause or contribute to “any . . . applicable emission standard” under the Act, id. § 7475(a)(3) (emphasis added). In the end, we agree that “the question here is: Who Decides?” Dissenting Op. at 18 (Kavanaugh, J.). We also agree that “Congress (with the President) sets the policy through statutes, agencies implement that policy within statutory limits, and courts in justiciable cases ensure that agencies stay within the statutory limits set by Congress.” 4 Dissenting Op. at 18 (Kavanaugh, J.). Here, Congress spoke clearly, EPA fulfilled its statutory responsibilities, and the panel, playing its limited role, gave effect to the statute’s plain meaning. See Chevron, U.S.A., Inc. v. NRDC, 467 U.S. 837, 842–43 (1984) (“If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.”). To be sure, the stakes here are high. The underlying policy questions and the outcome of this case are undoubtedly matters of exceptional importance. The legal issues presented, however, are straightforward, requiring no more than the application of clear statutes and binding Supreme Court precedent. There is no cause for en banc review. BROWN, Circuit Judge, dissenting from the denial of rehearing en banc: In the summer of 1974, while waiting to start classes at UCLA, I was lucky enough to obtain a summer job house sitting in the pleasant, upscale neighborhood of Pasadena. Known mostly for its Rose Parade and Rose Bowl, Pasadena is one of the more scenic exurbs of Los Angeles. I inhabited a sparsely furnished, modest-but-pricey bungalow set among the lush landscape typical of southern California. This is a place where Birds of Paradise grow ten feet tall and the magenta blossoms of Bougainvillea fall like lavish draperies from redwood garden trellises. After staying in the house more than a month and spending a restless night listening to the agitated thrashings of the jacaranda trees in a fitful wind, I stumbled bleary-eyed into the kitchen, looked out the window, and stopped — utterly dumbfounded. There — looking like it was but a few feet beyond the back fence — stood a mountain. Not a foothill. Not an unobtrusive mesa. A mountain! Closer inspection revealed not a lone majestic peak, but a whole mountain range I later identified as the San Gabriels. In those days, the air in the Los Angeles basin was so thick with smog that a mountain, or even a nearby mountain range, could simply disappear. Although the Los Angeles basin was among the most notorious examples of the phenomenon, it was by no means unique and certainly not the worst. It was this crisis of ambient air quality that precipitated the enactment of the Clean Air Act (CAA). But as the CAA’s history, language, and structure make clear, Congress never intended the Act to serve as an environmental cure-all. It was targeted legislation designed to remedy a particular wrong: the harmful direct effects of poisoned air on human beings and their local environs. This is what Congress understood as “air pollution which may reasonably be anticipated to endanger public health” in the tailpipe emissions provision, 42 U.S.C. § 7521(a)(1). The Supreme Court in Massachusetts v. EPA, 549 U.S. 497 (2007), however, concluded otherwise. In dicta 2 too suggestive to ignore, the Court implicitly assumed that climate change could provide the basis for an endangerment finding in the tailpipe context. See id. at 532–33. Bound as I am by Massachusetts, I reluctantly concur with the Panel’s determination that EPA may regulate GHGs in tailpipe emissions. But I do not choose to go quietly. Because the most significant regulations of recent memory rest on the shakiest of foundations, Part I of this statement engages Massachusetts’s interpretive shortcomings in the hope that either Court or Congress will restore order to the CAA. Part II, by contrast, reflects my belief that Massachusetts does not compel the same result for Title V and the Prevention of Significant Deterioration of Air Quality (PSD) program. Although I agree with Judge Kavanaugh’s dissent, Coal. for Responsible Regulation v. EPA, Nos. 09- 1322, et al. (Kavanaugh, J., dissenting from denial of rehearing en banc), I approach the inflection point from a slightly different perspective. Part III concludes with a brief note on standing. Because I would vote for the full court to consider the propriety of extending Massachusetts to Title V and the PSD program, I respectfully dissent from this denial of rehearing en banc. I. A. The origins of the Clean Air Act are closely tied to fatal fogs and deadly air inversions that, for much of early post- industrial history, seemed to be the inevitable consequence of economic progress. See Arnold W. Reitze, Jr., A Century of Air Pollution Control Law: What’s Worked; What’s Failed; 3 What Might Work, 21 ENVTL. L. 1549, 1575 (1991).1 Initially regulated at the local and state level, air pollution became the focus of the federal government only after World War II. See id. at 1585–86. In October 1948, a severe temperature inversion in the industrial city of Donora, Pennsylvania increased air pollution to such an extent that traffic “ ‘was virtually stopped because of lack of visibility.’ ” The inversion killed 20 people, id., and prompted the federal government to begin researching air pollution. Id. at 1586. By 1961, President Kennedy included a plea for “an effective air pollution program” in his Special Message on the Natural Resources. Id. Public pressures for legislation only increased when a “Killer Smog” engulfed London in December 1962, killing at least 340, and a similar inversion in New York City allegedly claimed the lives of 200. Id. Eventually, legislation recommended by President Kennedy in February 1963 led to the enactment of the CAA, which President Johnson signed into law on December 17, 1963. Id. at 1586–87. Seven years later, President Nixon signed The Clean Air Amendments of 1970. The 1970 Amendments authorized the EPA to prescribe national ambient air quality standards (NAAQS) and created the statutory framework that still exists today. B. It was no happy accident that congressional draftsmen titled the legislation the “Clean Air Act.” Ambient air quality was the point, purpose, and focus of the CAA. Congress had set its sights on the “dirty, visible ‘smokestack’ emissions,” 136 CONG. REC. H2771-03 (1990) (statement of Rep. Roe), 1 Inversions, sometimes known as “Londoners,” occur “when a layer of hot air warmed by . . . water exists above cooler ground- level air and traps smoke and particulate matter under the warmer air.” Id. 4 and smog caused by vehicle emissions. The CAA was the means by which Congress would grapple with urban air pollution and its attendant health effects, including impaired breathing, heart disease, lung damage and lung disease, and even death. If pollution was the problem, these ills were the specific harms Congress sought to combat. Even a cursory glance at the legislative history, with its numerous charts, graphics, and statistics detailing cancer and death rates, will bear this point out. See, e.g., Hearings on Air Pollution — 1968 Before the Subcomm. on Air and Water Pollution of the Sen. Comm. on Pub. Works, 90th Cong. 2nd Sess., pt. 2, 608– 20 (1968) (statement of Dr. Samuel S. Epstein, Children’s Cancer Research Foundation.) (“Air Pollution — 1968”). With the enactment of the 1990 Amendments, Congress expanded the Act beyond its singular emphasis on urban air quality to address hazardous — i.e., toxic — air pollutants, acid rain, and stratospheric ozone. In regulating hazardous pollutants, Congress reemphasized the need for a close and tangible nexus between pollutant and harm. The legislative record, for example, continued to conceive of dangers in terms of their direct effects on human health and well-being. See, e.g., S. Rep. No. 101-228, at 3388 (1989), reprinted in 1990 U.S.C.C.A.N. 3385 (“Air pollution can silently damage our lungs and heart or act swiftly in the case of exposure to toxic air pollutants. Rigorous regulation of toxic air pollutants is needed to avoid risk of serious, irreversible damage to human health.”). To the extent the regulation of stratospheric ozone and acid rain suggest a broader nexus between pollutant and harm to human health, the very particular way in which Congress handled these exceptions goes a long way toward proving the rule: Congress only expands the CAA through considered legislative acts. 5 In addressing these transnational phenomena, the legislature did not spin regulations out of whole cloth. With ozone concerns, for example, Congress developed solutions through international negotiations, the implementation of which led to the creation of a separate title of the CAA. See NRDC v. EPA, 464 F.3d 1, 3 (D.C. Cir. 2006). Likewise, years of contentious discussions with Canada helped bring about the acid rain provisions in the 1990 Amendments. See generally Dennis A. Leaf, Intergovernmental Cooperation: Air Pollution from an U.S. Perspective, 18 CAN.-U.S. L.J. 245 (1992). Simply put, when Congress became aware of new dangers, it acted judiciously in crafting workable remedies that, when they obtained the necessary political support, were worked into their own discrete provisions under the Act. Neither Congress nor the EPA attempted to force these distinct problems into existing, ill-suited regulatory schemes. Congressman Waxman, one of the strongest proponents of stringent air pollution controls and a key force behind the 1990 Amendments, has stated that “in recent experience, no legislation has received more scrutiny during its consideration.” The Honorable Henry A. Waxman, An Overview of the Clean Air Act Amendments of 1990, 21 ENVTL. L. 1721, 1724 (1991). Hyperbole or not, the admission is telling. The history of the CAA is one of hard- fought incremental gains through which Congress remedied particular environmental wrongs with tailored remedies. Said the Congressman: Discrete and extensive new programs are included to grapple with high ambient pollution levels (urban and regional smog), hazardous air pollution, acid rain, and depletion of the stratospheric ozone layer. Each of these programs [was] tailored to the problem it [sought] to address, and each [was] quite different in its approach.” 6 Id. at 1811. Political necessity has forced Congress to calibrate its amendments to the CAA with great specificity and care. Where our Representatives have acted with such caution, any suggestion that Congress has — through a single word — conferred upon EPA the authority to steamroll through Congressional gridlock, upend the Senate’s rejection of the Kyoto Protocol, and regulate GHGs for the whole of American industry must necessarily fail. The legislature, recall, does not “hide elephants in mouseholes.” Whitman v. Am. Trucking Assocs., 531 U.S. 457, 468 (2001). But we needn’t rely on interpretative canons alone to make this point. In drafting the 1990 Amendments, Congress considered — and expressly rejected — proposals authorizing EPA to regulate GHGs under the CAA. See S. Rep. No. 101- 228, at 377 (1989), as reprinted in 1990 U.S.C.C.A.N. 3385, 3760. Even the Executive objected that an attempt to control Carbon Dioxide (CO2) emissions — emissions not harmful to health — in order to prevent global warming was premature. See Administration’s Amendments — Hearings Before the Subcomm. On Health and the Env’t of the Comm. on Energy and Commerce, 101st Cong., 1st Sess. (1989) (includes Bush Administration Report on S. 1630). The Executive’s critique noted that “unilateral action aimed at addressing a global problem” through a standard limiting tailpipe emissions would not be an effective means of safeguarding the global environment and would “necessarily punish national interests.” Id. at 792, 813. That Congress has never deviated from its decision to not regulate GHGs under the CAA was not for lack of opportunity. Congress has considered and rejected countless other bills in the years since the 1990 Amendments that would have authorized GHG regulation. By one estimate, 7 Congressmen have proposed over 400 bills concerning GHGs between 1990 and 2009. See Abigail R. Moncrieff, Reincarnating the “Major Questions” Exception to Chevron Deference As A Doctrine of Noninterference (or Why Massachusetts v. EPA Got It Wrong), 60 ADMIN. L. REV. 593, 636–37 (2008) (tracking proposals). Congress’s inability to break this nearly quarter-century long deadlock is incredibly suggestive: this is not an area of policymaking where the legislature has acted rashly or unthinkingly in delegating authority to agencies. At bottom, Congress understood the dangers of “any air pollutant” in § 7521(a)(1) in terms of the ill-effects caused those who inhale the pollutants, not the broad, attenuated consequences of climate change. The CAA was drafted not to combat the threat of flooding or the menace of heat waves, see Endangerment and Cause of Contribute Findings for Greenhouse Gases, 74 Fed. Reg. 66,496, 66,526 (Dec. 15, 2009) (“EPA Endangerment Finding”), but the choking, stifling, and degenerative effect of airborne pollutants on human beings and their affected localities. Congress has long quantified this harm in terms of mortality rates, see, e.g., Air Pollution — 1968, 564 (statement of Dr. Roger S. Mitchell, Director, Webb-Waring Institute for Medical Research), not acreage of “costal land” lost. Massachusetts, 549 U.S. at 522. To put matters pointedly: the injury sufficient to establish standing need not suffice to establish endangerment as well. Congress was of course free to circumvent this close cause-health effect nexus by devising a separate provision for GHG regulation, much as it did for stratospheric ozone, but it did no such thing. And nothing in the legislative history suggests that Congress has deviated from this status quo. 8 The plain language of the CAA only underscores the Act’s non-applicability to GHGs insofar as it requires the harm be of the sort “reasonably [] anticipated to endanger.” 42 U.S.C. §7251(a)(1) — a term we know to have a discrete meaning. C. In the present case, this Court had “little trouble” disposing of the argument that the “PSD program is specifically focused solely on localized air pollution” because it is “quite clear . . . the PSD program was intended to protect against precisely the types of harms caused by greenhouse gases.” CRR Slp. Op. 62–63 (emphasis added). Massachusetts notwithstanding, this statement is a curious thing in light of the uncontradicted legislative history just discussed.2 So too is the court’s reliance on the statutory text, particularly its finding that “the CAA expressly provides that effects on ‘welfare’ means ‘effects on . . . weather . . . and climate.’ ” Slp. Op. 62-63 (citing 42 U.S.C. § 7602(h)). As a textual matter, there is nothing “quite clear” about it. The Supreme Court has declared that GHGs like CO2 are pollutants within the meaning of the Act. Under the CAA, however, EPA can regulate a pollutant only if the administrator finds that the GHG causes or contributes to “air pollution which may reasonably be anticipated to endanger public health or welfare.” 42 U.S.C. §7251(a)(1) (emphasis added). But in locating the CAA’s conception of “harm” in § 7602(h), the definition of “welfare,” and not §7251(a)(1) 2 As noted, the weather and climate issues targeted by the CAA involve direct, deleterious, localized effects caused by polluted air people breathe or suspended pollutants that may be deposited on land and crops by precipitation. 9 generally, this court effectively skirted the operative statutory language — “may reasonably be anticipated” — and rendered it nugatory. This was in error. Section 7602(h) defines only the potential objects of harm; the “reasonably be anticipated” language of §7251(a)(1) supplies the requisite nexus between the pollutant and the objects of its harm. The two provisions must be read together if the statute is to be interpreted faithfully. To put matters another way, the “may reasonably be anticipated” language must do some analytical work in the endangerment determination lest it be deemed surplusage. See, e.g., Conference of State Bank Supervisors v. Conover, 715 F.2d 604, 627 (D.C. Cir. 1983) (“[I]n construing a statute, we ‘are obliged to give effect, if possible, to every word Congress used.’ ” (quoting Reiter v. Sonotone Corp., 442 U.S. 330, 339 (1979))). And in view of the CAA’s legislative history, the nature of that work is clear. In order to reasonably anticipate that a pollutant will contribute to air pollution that endangers public health or welfare, the Agency would have to conclude that pollution created by CO2 or another GHG is a reasonably direct cause of the damage to public health and welfare. To find that CO2 may ultimately endanger public health and welfare because sea levels will rise tells us nothing about whether CO2 concentrations in the ambient air directly harm public health and welfare. The ingredients of a Killer Smog are few and specific; the process through which an air inversion traps particulate matter close to the ground is well understood. With both there is a direct correlation between reducing the concentration of the pollutant and reducing the negative health effects. Questions of public health impacts from air pollution have consistently been based on the direct — that is, inhalational — effects of exposure to the pollutant. See, e.g., Joint Opening Brief of Non-State Petitioners and Supporting Intervenors at 58, Coal. for Responsible Regulation v. EPA, 10 No. 09-1322 (May 20, 2011); NRDC, Inc. v. EPA, 902 F.2d 962, 973 (D.C. Cir. 1990) (concluding that EPA may not consider the health effects of increased unemployment when setting new health-based NAAQS) In contrast, any harm to human health and welfare flowing from climate change comes at the end of a long speculative chain. The dissent in Massachusetts pointed out that EPA had described in great detail the scientific uncertainty that precluded even forming a judgment as to whether greenhouse gases endanger public welfare. See 549 U.S. at 553–55 (Scalia, J., dissenting). In that earlier defense of its refusal to form a judgment, EPA explained how predicting climate change involved a “complex web of economic and physical factors,” including: [o]ur ability to predict future global anthropogenic emissions of GHGs and aerosols; the fate of these emissions once they enter the atmosphere (e.g., what percentage are absorbed by vegetation or are taken up by the oceans); the impact of those emissions that remain in the atmosphere on the radiative properties of the atmosphere; changes in critically important climate feedbacks (e.g., changes in cloud cover and ocean circulation); change in temperature characteristics (e.g., average temperatures, shifts in daytime and evening temperatures); changes in other climatic parameters (e.g., shifts in precipitation, storms); and ultimately the impact of such changes on human health and welfare (e.g., increases or decreases in agricultural productivity, human health impacts). Id. If there can be this much logical daylight between the pollutant and the anticipated harm, there is nothing EPA is not authorized to do. If this finding is valid, in a world where six 11 degrees of separation is the compass of all humankind, the right endangerment finding would allow EPA to rule the world. But as this Court has noted before, EPA’s authority to regulate is constrained, not enlarged, by the relationship of the term “will endanger” to other sections of the CAA. See Ethyl v. EPA, 541 F.2d 1, 29 (D.C. Cir. 1976) (en banc). Of course, nothing here should be taken to imply that a particular GHG does not contribute to climate change. I mean only to suggest that a pollutant might contribute to the nebulous mélange of potential drivers of climate change without having any direct, deleterious impact within the meaning of the CAA. I emphasize too that this is not a problem with science. This is a problem of statutory interpretation. Climate change, with its geologic timeframe and its many uncertainties and imponderables, is and will probably remain a subject of some controversy. EPA finds the science sufficiently convincing for its purposes and it is entitled to a certain amount of deference on questions related to its technical expertise. But it is not necessary to quibble with the science of climate change to conclude that the endangerment finding fails on textual and logical terms. There is simply a point at which a difference in degree becomes a difference in kind and we have passed this point many times over in the course of this tortured litigation. The Supreme Court, however, has refused to recognize as much for tailpipe emissions. II. A. But we need not follow Massachusetts off the proverbial cliff and apply its reasoning to the unique Title V and PSD provisions not considered in that case. The cascading layers 12 of absurdity that flow from that interpretive exercise make clear that the plain language of the CAA compels no such result. As EPA’s own rulemaking documents have so unabashedly explained: To apply the statutory PSD and title V applicability thresholds literally to sources of GHG emissions would bring tens of thousands of small sources and modifications into the PSD program each year, and millions of small sources into the title V program. These extraordinary increases in scope of the permitting programs would mean that the programs would become several hundred-fold larger than what Congress appeared to contemplate. PSD and Title V Greenhouse Gas Tailoring Rule; Final Rule, 75 Fed. Reg. 31,514, 31,533 (Jun. 3, 2010) (“Final Tailoring Rule”). Completely oblivious to the irony, EPA added: For our authority to take this action, we rely in part on the “absurd results” doctrine, because applying the PSD and title V requirements literally (as previously interpreted narrowly by EPA) would not only be inconsistent with congressional intent concerning the applicability of the PSD and title V programs, but in fact would severely undermine congressional purpose for those programs. Id. at 31,541–42. And again: [I]n this case because a literal reading of the PSD and title V applicability provisions results in insurmountable administrative burdens. Those insurmountable administrative burdens — along with the undue costs to sources — must be considered “absurd results” that 13 would undermine congressional purpose for the PSD and title V programs. Id. at 31,547. In precincts outside Washington, D.C., this litany might cause a regulator to pause and consider whether results so at odds with Congressional presuppositions could ever be justified as falling within the literal meaning of an enactment. EPA, however, proposes that the absurd result can be easily eliminated by ramping up and gradually phasing in the requirements. Faced with the choice of reconsidering the legitimacy of an endangerment finding that sets in motion such a cluster of chaos or rewriting the statute, the agency has blithely done the latter. This is an abuse of the absurdity and administrative necessity doctrines as neither can be invoked to preempt legislative prerogatives. Permitting a statute “to be read to avoid absurd results allows an agency to establish that seemingly clear statutory language does not express the ‘unambiguously expressed intent of Congress,’ ” but it does not grant the agency “a license to rewrite the statute.” Mova Pharmaceuticals v. Shalala, 140 F.3d 1060, 1068 (D.C. Cir. 1998). But that is not the worst of it. The real absurdity — apparently as invisible to the EPA as the San Gabriels once were to me — cannot be cured by phase in, no matter how subtly Byzantine. The real absurdity is that this unprecedented expansion of regulatory control, this epic overreach, may very well do more damage to the wellbeing of Americans than GHGs could ever do.3 3 See, e.g., Joint Reply Br. of Non-State Petitioners and Supporting Intervenors at *1, No. 09-1322 (Nov. 14 2011) (“Nor does [EPA] dispute that the new rules will impose massive burdens on a struggling economy, or that its program of vehicle standards 14 B. A second, more elementary consideration counsels against the mechanical application of Massachusetts’s tailpipe emissions determination to these distinct CAA provisions: deference to Congress. As articulated in Food & Drug Administration v. Brown & Williamson Tobacco Corp., 529 U.S. 120 (2000), the Supreme Court’s “major questions” canon gives form to the judicial intuition so strongly implicated here: Congress should not be presumed to have deferred to agencies on questions of great significance more properly resolved by the legislature. If there was ever a regulation in recent memory more befitting such a presumption than the present, I confess I do not know of it. On familiar facts, the Supreme Court in Brown & Williamson rebuffed the FDA’s expansionist effort to bring tobacco products within its regulatory ambit. The agency’s regulation rested on a strained interpretation of the Food, Drug, and Cosmetic Act, 21 U.S.C. § 301 et seq., in which it defined nicotine as a “drug” and cigarettes and smokeless tobacco as “combination products” used to deliver nicotine to the body. See Brown & Williamson, 529 U.S. at 125–27. Applying Chevron U.S.A. Inc. v. Natural Resources Defense Council, 467 U.S. 837 (1984), the Court first considered the statutory structure. “[I]f tobacco products were within the FDA’s jurisdiction,” the majority concluded, the normal operation of the “Act would require the FDA to remove them from the market entirely,” and this would “contradict will affect global mean temperatures by no more than 0.01 degree Celsius by 2100”). 15 Congress’ clear intent as expressed in its more recent, tobacco-specific legislation.” Brown & Williamson, 359 U.S. at 143. As the present case confirms, such absurdity is all but inevitable where an agency attempts to regulate that which “simply do[es] not fit” within its regulatory scheme. Id. The Court next considered Congress’s 35 year history of tobacco- specific legislation, finding it “clear” that this “legislation has effectively ratified the FDA’s previous position that it lacks jurisdiction to regulate tobacco.” Id. at 156. The Court then closed its lengthy Chevron discussion with an appeal to first principles. The “inquiry into whether Congress has directly spoken to the precise question at issue,” the Court explained, “is shaped, at least in some measure, by the nature of the question presented.” Id. at 159. Chevron deference operates on the assumption “that a statute’s ambiguity constitutes an implicit delegation,” but this tenuous fiction need not hold true in every situation. Id. “In extraordinary cases,” the Court went on, “there may be reason to hesitate before concluding that Congress has intended such an implicit delegation.” Id. (referencing Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 ADMIN. L. REV. 363, 370 (1986) (“A court may also ask whether the legal question is an important one. Congress is more likely to have focused upon, and answered, major questions, while leaving interstitial matters to answer themselves in the course of the statute’s daily administration”)).4 4 MCI Telecommunications Corporation v. AT&T Co., 512 U.S. 218 (1994), a case the Brown & Williamson Court found “instructive,” Brown & Williamson, 529 U.S. at 160, had advanced a similar logic. In concluding Congress had spoken to the meaning of the term “modify” as it appears in § 203(b) of the Communications Act of 1934, the Court rejected FCC’s far more expansive interpretation. The Court assumed in dicta that it was 16 Declaring Brown & Williamson “hardly [the] ordinary case,” the Court reasoned: Contrary to its representations to Congress since 1914, the FDA has now asserted jurisdiction to regulate an industry constituting a significant portion of the American economy. In fact, the FDA contends that, were it to determine that tobacco products provide no “reasonable assurance of safety,” it would have the authority to ban cigarettes and smokeless tobacco entirely. Owing to its unique place in American history and society, tobacco has its own unique political history. Congress, for better or for worse, has created a distinct regulatory scheme for tobacco products, squarely rejected proposals to give the FDA jurisdiction over tobacco, and repeatedly acted to preclude any agency from exercising significant policymaking authority in the area. Given this history and the breadth of the authority that the FDA has asserted, we are obliged to defer not to the agency's expansive construction of the statute, but to Congress’ consistent judgment to deny the FDA this power. Brown & Williamson, 529 U.S. at 159-60. In view of the language, structure, and history of the CAA, I am simply unable to distinguish this logic from the present case in any meaningful way. To the contrary, with “highly unlikely that Congress would leave the determination of whether an industry will be entirely, or even substantially, rate- regulated to agency discretion — and even more unlikely that it would achieve that through such a subtle device as permission to ‘modify’ rate-filing requirements.” MCI, 512 U.S. at 231. Certainly the same might be said here as well. 17 only the slightest of modifications one could rework the above text to apply to GHG emissions.5 Although the Massachusetts Court distinguished Brown & Williamson, it did so only in the context of tailpipe emissions. Its reasoning does not extend to Title V and the PSD program. In the Court’s view, Brown & Williamson had “found critical at least two considerations that have no counterpart in [Massachusetts].” 549 U.S. at 531. First, whereas the regulation of tobacco under the FDCA would have necessarily led to a ban on tobacco products — an outcome that clashed with the “common sense” intuition that Congress never meant to remove those products from circulation — the expansion of EPA’s “jurisdiction would lead to no such extreme measures 5 Perhaps: Contrary to its representations in Massachusetts v. EPA, the EPA has now asserted jurisdiction to regulate industries constituting a significant portion of the American economy. In fact, the EPA contends that, because greenhouse gases can be regulated as tailpipe emissions, it is obligated to regulate all stationary sources at admittedly “absurd” levels. Owing to its ubiquitous place in the planet’s life cycle, greenhouse gases have their own unique political history. Congress, for better or for worse, has declined to create a distinct regulatory scheme for greenhouse gases, squarely rejected proposals to give the EPA jurisdiction over greenhouse gases, and repeatedly acted to preclude any agency from exercising significant policymaking authority in the area. Given this history and the breadth of the authority that the EPA has asserted, we are obliged to defer not to the agency's expansive construction of the statute, but to Congress’ consistent judgment to deny the EPA this power. 18 [because] EPA would only regulate emissions” and “there is nothing counterintuitive to the notion that EPA can curtail the emission of substances that are putting the global climate out of kilter.” Id. But the Court spoke too soon. In the present litigation, EPA argued — and a Panel of this Court readily agreed — that in regulating tailpipe emissions under 42 U.S.C. § 7521, it is obligated to regulate stationary sources under Title V and the PSD program as well. As a threshold matter, the Massachusetts Court never considered these far- reaching effects. It limited its brief discussion on the merits to the tailpipe emissions question squarely before it. In this way, the Court never considered the differing ways in which the CAA regulates tailpipes and stationary sources. With tailpipe emissions, the inclusion of greenhouse gasses within the term “air pollutant” does not directly expand or contract the universe of vehicles and engines subject to the new standards. Consequently, the regulation’s impact will fall primarily on those manufacturers already complying with existing emission requirements. And even then, the Court explained, EPA “would have to delay any action ‘to permit the development and application of the requisite technology, giving appropriate consideration to the cost of compliance.’ ” Massachusetts, 549 U.S. at 531 (quoting § 7521(a)(2)). Not so with the regulation of stationary sources. Insofar as 42 U.S.C. § 7479(1) defines “major emitting facility” to include those facilities with the “potential to emit” either 100 or 250 “tons per year or more of any air pollutant,” the statutory term is necessarily tied to CAA’s jurisdictional scope. Inescapably, then, the regulation of greenhouse gasses as “air pollutants” will radically expand the universe of covered entities far beyond Congress’s intentions. EPA’s decidedly extra-textual Tailoring Rule only confirms the ludicrousness of this result. Nor can it be said that the statutory safeguards operate in the same way as § 19 7521(a)(2). Permitting authorities may well be able to determine on a case-by-case basis what constitutes the “best available control technology” for a particular emitting facility, 42 U.S.C. § 7479(3), but this is of little consolation for the small business owner who previously fell outside the CAA. At bottom, this outcome clashes with the “common sense” understanding that Congress would not have intended such a broad, unchecked expansion of the CAA to potentially millions of businesses from all walks of industry. The Supreme Court in Massachusetts simply did not have occasion to consider this absurd and “counterintuitive” outcome, but we do — and we must. Second, the Court determined that the “unbroken series of congressional enactments” referenced in Brown & Williamson “made sense only if adopted ‘against the backdrop of the FDA’s consistent and repeated statements that it lacked authority under the FDCA to regulate tobacco.’ ” Massachusetts, 549 U.S. at 531.6 By contrast, EPA had “not identified any congressional action that conflicts in any way with the regulation of greenhouse gases from new motor vehicles.” Id. And even if it had, “Congress could not have acted against a regulatory ‘backdrop’ of disclaimers of regulatory authority” because “EPA had never disavowed the authority to regulate greenhouse gases, and in 1998 it in fact affirmed that it had such authority.” Id. When read in context, however, it is clear that the Court’s reasoning was building toward a wholly unspectacular point: because EPA’s legislative history failed to establish congressional intent with 6 The suggestion here seems to be that Congress’s decision to regulate tobacco products would not, by itself, evince its intent to proscribe agencies from doing the same. Doing so in light of FDA’s statements, however, had the effect of implicitly codifying the agency’s long-held view. 20 the same weight and precision as Brown & Williamson, it did not justify “read[ing] ambiguity into a clear statute.” Id. That logic is inapplicable here. In the absence of lexical clarity — which the Court had found in in CAA’s “sweeping definition of ‘air pollutant,’ ” id. at 528 — we need legislative history and other indicia of congressional intent to inform our understanding of how GHGs are to be regulated under other CAA provisions.7 The Massachusetts Court’s effort to distinguish Brown & Williamson is thus unavailing where we deal not with the definitional scope of “any pollutant” and tailpipe emissions, but the particular dangers Congress sought to combat in enacting Title V and the PSD program. When read in conjunction with the CAA’s history, structure, and language, the intuitive logic of the “major questions” doctrine makes clear that the Panel erred in extending Massachusetts. 7 Consider the role of NAAQS in this regulatory system. EPA in Massachusetts had observed that NAAQS were established to “address air pollution problems that occur primarily at ground level” as well as “concentrations of substances in the ambient air and the related public health and welfare problems.” Massachusetts, 549 U.S. at 558–59 (Scalia, J., dissenting). EPA thus reasoned that the regulation of the buildup of CO2 in the upper reaches of the atmosphere — the process alleged to cause global climate change — was not akin to regulating the concentration of a substance that is polluting the air and was “beyond the scope of CAA’s authorization to regulate.” Id. In other words, EPA maintained that had Congress intended the CAA to regulate greenhouse cases and global climate change, it would have provided some better tool than NAAQS. That defense — offered in response to a demand to regulate tailpipe emissions — applies with even greater potency to Title V and the PSD program. In fact, although EPA now claims it is authorized to regulate greenhouse gases and global climate change, the agency acknowledges that the regulatory framework is as ill-suited to the task as ever. 21 Congress simply did not intend for EPA to convert the “Clean Air Act” to the “Warm Air Act” writ large. But that is exactly what the federal courts have done. As the Chief Justice observed in his Massachusetts dissent, impatience is not a juridical principle that can be sustained under our constitutional framework. See Massachusetts, 549 U.S. at 535–36 (Roberts, C.J., dissenting). It certainly fares no better as a default measure of institutional choice under Chevron. As Massachusetts recognized, an agency can only exercise the authority Congress has delegated to it. See 549 U.S. at 534–35 (noting that EPA must “ground its reasons for action or inaction in the statute” and “exercise its discretion within defined statutory limits.”). Absurdity can never figure as an adequate substitute for authority in this threshold assessment. Nor can absurdity cure the agency’s failure to establish that the statute unambiguously compels its interpretation or that its interpretation, though discretionary, is actually consistent with statutory text, structure, and purposes. The agency seeks to avoid these pesky constraints here by invoking Massachusetts, but Article III judges cannot be a legitimate source of legislative authority. By deferring to the distorted claim of delegation advanced here, this Court has transformed Chevron from a useful, albeit accidental, touchstone into an idol to which we surrender our constitutional faith. III. In rejecting State Petitioners’ challenge to the Tailoring Rule for want of standing, the Panel invoked that famed preceptor of American civics, Schoolhouse Rock, to great effect. Slp. Op. at 79. (“As a generation of schoolchildren knows, ‘by that time, it’s very unlikely that [a bill will] become a law. It’s not easy to become a law.’ ”). I certainly 22 do not quarrel with such dispositive authority. Lawmaking is neither easy nor certain. In an ordinary case, the mere possibility of “corrective legislation” will not establish that redress is “likely, as opposed to merely speculative.” Lujan, 504 U.S. at 561. But it bears repeating that this is not an ordinary case. Where the choice is between non-action or a confessedly “absurd” regulation poised to impress countless billions of dollars in costs on American industry, we have transcended the realm of the speculative. For once, the comparison with Massachusetts is apt. The Supreme Court found standing on the basis of an estimated rise in sea level of 20 to 70 centimeters by the year 2100, see Massachusetts, 549 U.S. at 542 (Roberts, C.J, dissenting) — a prediction based almost entirely on conjecture. Is it any more speculative to say that specific projections of billions of dollars in actual regulatory costs would not suffice to compel Congress to act? The Panel’s alternative contention fares better: because Congress could remedy the issue in countless ways, not all of which inure to State Petitioners’ benefit, the inquiry is “inherently speculative.” See Op. at 79. This argument benefits from the genuine uncertainty in Congress over what, if any, role EPA should play in GHG regulation. But therein lies a frighteningly obtuse logic. If EPA actions are ultra vires precisely because disagreement on the Hill prevented Congress from altering the status quo and authorizing such regulation, how then can the very same deadlock be used to defeat Petitioners’ standing to challenge the Rule through which EPA effectuates its absurdist scheme? The Court cannot have it both ways. At bottom, bad decisions make bad law. In denying rehearing en banc, this Court has read Massachusetts to its illogical ends and it is American industry that will have to pay. That this Court did so is unsurprising, but certainly not 23 fated. Massachusetts does not compel this outcome for the PSD and Title V provisions. Had this Court interrogated its own assumptions and yielded not to Massachusetts’s telos but sound constitutional principles, it would have found that the matter properly belongs before Congress, not courts or agencies. As Schoolhouse Rock long ago explained: Ring one, Executive, Two is Legislative, that’s Congress. Ring three, Judiciary. See it’s kind of like my circus, circus.8 And what a circus it is. For these reasons, I respectfully dissent from the denial of rehearing en banc. 8 “Three Ring Government,” Schoolhouse Rocks, available at http://www.schoolhouserock.tv/ThreeRing.html. KAVANAUGH, Circuit Judge, dissenting from the denial of rehearing en banc: This case is plainly one of exceptional importance. A decision in either direction will have massive real-world consequences. The U.S. Chamber of Commerce describes the EPA regulations at issue here as “the most burdensome, costly, far-reaching program ever adopted by a United States regulatory agency.” Petition for Rehearing En Banc at 1. On the other hand, EPA issued these regulations to help address global warming, a policy issue of major long-term significance to the United States. Put simply, the economic and environmental policy stakes are very high. Of course, our role is not to make the policy choices or to strike the balance between economic and environmental interests. That job is for Congress and the President when considering and enacting legislation, and then as appropriate for the Executive Branch – here, EPA, under the ultimate supervision of the President – when exercising its authority within statutory constraints. Our job as a court is more limited: to ensure that EPA has acted within the authority granted to it by Congress. In this case, I conclude that EPA has exceeded its statutory authority. I respectfully disagree with the panel opinion’s contrary conclusion, and given the overall importance of the case, I respectfully dissent from the denial of rehearing en banc. I A This case concerns EPA’s implementation of the Prevention of Significant Deterioration provisions of the Clean Air Act. The Prevention of Significant Deterioration program – which is codified in Sections 7470 to 7479 of Title 42 – is designed to maintain state and local compliance with 2 the National Ambient Air Quality Standards, known as the NAAQS. The NAAQS are currently established for six air pollutants: carbon monoxide, lead, nitrogen dioxide, ozone, particle pollution, and sulfur dioxide. As relevant here, the Prevention of Significant Deterioration statute requires stationary facilities that emit certain “air pollutants” to obtain permits before beginning new construction. See 42 U.S.C. §§ 7475(a)(1), 7479(1). To obtain a permit, the facility must undergo a lengthy, costly process to analyze the new construction’s impact on air quality and to try to demonstrate its compliance with the relevant emissions limits. A central question in this case is how to construe the term “air pollutant” for purposes of this statutory permitting requirement. In particular, the question is whether the term “air pollutant” here covers not just the NAAQS pollutants, which can cause breathing problems or other health issues, but also greenhouse gases such as carbon dioxide, which contribute to global warming. Under the broader interpretation of “air pollutant” that encompasses greenhouse gases, a far greater number of facilities would fall within the Prevention of Significant Deterioration program and have to obtain pre-construction permits. That in turn would impose significantly higher costs on businesses and individuals that are building new commercial or residential property. In considering a different Clean Air Act program targeted at motor vehicle emissions, the Supreme Court said that the term “air pollutant” meant “all airborne compounds of whatever stripe,” which included greenhouse gases such as carbon dioxide. Massachusetts v. EPA, 549 U.S. 497, 529 (2007). But all parties here, including EPA, agree that the Massachusetts v. EPA interpretation of the term “air pollutant” cannot control in this case, for purposes of this very 3 different Clean Air Act program for stationary facilities. Rather, as the parties agree, we must look to the text and context of the Prevention of Significant Deterioration statute to determine what “air pollutant” covers here. Looking at the relevant statutory text and context, there would initially appear to be two plausible interpretations of the term “air pollutant” for purposes of the Prevention of Significant Deterioration statute: (i) more broadly, an airborne compound that is deemed harmful and is regulated by EPA in any Clean Air Act program, which would include greenhouse gases such as carbon dioxide; or (ii) more narrowly, the six air pollutants that are regulated by EPA in setting and enforcing the NAAQS, which would cover carbon monoxide, lead, nitrogen dioxide, ozone, particle pollution, and sulfur dioxide, but would not include greenhouse gases such as carbon dioxide. EPA chose the broader interpretation of “air pollutant,” thereby greatly expanding the reach of the Prevention of Significant Deterioration statute. But that broader interpretation has a glaring problem, as EPA itself recognized. In the context of the Prevention of Significant Deterioration statute, EPA’s broader interpretation would not mesh with other provisions of the statute and would lead to absurd results. That’s because the Prevention of Significant Deterioration statute requires pre-construction permits for facilities with the potential to emit more than 250 tons per year (or, for some facilities, 100 tons per year) of any covered pollutant. See 42 U.S.C. §§ 7475(a)(1), 7479(1). That would be a very low trigger for emissions of greenhouse gases because greenhouse gases are emitted in far greater quantities than the NAAQS pollutants. As a result, the low trigger would mean a dramatically higher number of facilities would 4 fall within the program and have to obtain pre-construction permits. In an unusual twist, EPA openly acknowledged the unreasonableness – indeed, the absurdity – caused by its interpretation of the statute. If the Prevention of Significant Deterioration program were interpreted to require pre- construction permits based on emissions of greenhouse gases, EPA candidly stated that the result would be “so contrary to what Congress had in mind – and that in fact so undermines what Congress attempted to accomplish with the PSD requirements – that it should be avoided under the ‘absurd results’ doctrine.” 74 Fed. Reg. 55,292, 55,310 (Oct. 27, 2009). But faced with those absurd consequences from the broader interpretation of the statute, EPA surprisingly did not choose the seemingly obvious option of adopting the narrower and more sensible interpretation of the term “air pollutant” for the Prevention of Significant Deterioration statute – the interpretation limited to NAAQS air pollutants. Instead, EPA plowed ahead with the broader interpretation. And then, to try to deal with the absurd repercussions of that interpretation for the Prevention of Significant Deterioration statute, EPA re-wrote the very specific 250-ton trigger in the permitting requirement of the statute, unilaterally raising that trigger for greenhouse gas emissions from 250 tons to 100,000 tons – a 400-fold increase. See 75 Fed. Reg. 31,514 (June 3, 2010). EPA believed that re-writing the statute’s permitting-triggers provision in this way would reduce the number of facilities that would require pre-construction permits and thereby “tailor” the absurdity – that is, alleviate some of the absurdity 5 caused by interpreting “air pollutant” to cover greenhouse gases. 1 This is a very strange way to interpret a statute. When an agency is faced with two initially plausible readings of a statutory term, but it turns out that one reading would cause absurd results, I am aware of no precedent that suggests the agency can still choose the absurd reading and then start re- writing other perfectly clear portions of the statute to try to make it all work out. And just recently, the Supreme Court reminded the Executive Branch and the lower courts that this is not the proper way to interpret a statute: Instead of “reading new words into the statute” to avoid absurd results, as the Government had urged in that case, the Court said that the statute should be interpreted so that “no absurdity arises in the first place.” Kloeckner v. Solis, No. 11-184, slip op. at 13 (U.S. 2012). Even limited to this case alone, the practical implications of accepting EPA’s approach are obviously major. And if this case stands as a precedent that influences other agency decisionmaking, the future consequences likewise could be significant: Agencies presumably could adopt absurd or otherwise unreasonable interpretations of statutory provisions and then edit other statutory provisions to mitigate the 1 At the same time, EPA reserved the right to ratchet the trigger all the way back down to 250 tons, thereby bringing more and more facilities under the program at EPA’s unilateral discretion. EPA’s assertion of such extraordinary discretionary power both exacerbates the separation of powers concerns in this case and underscores the implausibility of EPA’s statutory interpretation. Put simply, the statute cannot be read to grant discretion to EPA to raise or lower the permitting triggers as EPA sees fit. 6 unreasonableness. Allowing agencies to exercise that kind of statutory re-writing authority could significantly enhance the Executive Branch’s power at the expense of Congress’s and thereby alter the relative balance of powers in the administrative process. I would not go down that road. B In my view, the statutory issue here is reasonably straightforward. The Prevention of Significant Deterioration statute’s definition of “major emitting facility” subjects a facility to the permitting requirement based on the facility’s emissions of “air pollutants.” See 42 U.S.C. §§ 7475(a)(1), 7479(1). In the context of the Prevention of Significant Deterioration program as a whole, it seems evident that the term “air pollutant” refers to the NAAQS air pollutants. To begin with, as explained above, interpreting “air pollutant” in this context to refer to the NAAQS air pollutants would avoid the absurd consequences that EPA’s broader interpretation creates – namely, the exponential increase in the number of facilities that would be required to obtain pre- construction permits. That single point alone provides dispositive support for the narrower, NAAQS-specific interpretation. See, e.g., Taniguchi v. Kan Pacific Saipan, Ltd., 132 S. Ct. 1997, 2004-05 (2012) (statutory context supports narrower rather than broader reading of statutory term). Moreover, other provisions in the Prevention of Significant Deterioration statute likewise plainly use the term “air pollutant” to refer to the NAAQS air pollutants. The Prevention of Significant Deterioration program is codified in Sections 7470 to 7479 of Title 42. Of relevance here, Section 7473 sets guidelines for areas designated as in attainment of 7 the NAAQS and requires that the “concentration of any air pollutant” in those areas not exceed certain concentrations permitted by the NAAQS. 42 U.S.C. § 7473(b)(4). The term “air pollutant” in Section 7473(b)(4) necessarily refers to the NAAQS air pollutants. In addition, several other provisions in the Prevention of Significant Deterioration statute similarly refer to Section 7473(b)(4)’s maximum concentrations for NAAQS pollutants. Each of those references thus also necessarily employs a NAAQS-specific use of the term “air pollutant.” See, e.g., 42 U.S.C. § 7473(c)(1) (listing exclusions from “the maximum allowable increases in ambient concentrations of an air pollutant”); § 7474(a)(B) (redesignations cannot cause “concentrations of any air pollutant” to exceed the maximum); see also § 7475(a)(3)(A) (facility may not cause air pollution in excess of “maximum allowable concentration for any pollutant”). So it’s clear that a variety of provisions in the Prevention of Significant Deterioration statute use “air pollutant” to refer to a NAAQS air pollutant. And we presume that, unless otherwise indicated, the term “air pollutant” is used the same way throughout the Prevention of Significant Deterioration statute – and here, we have no reason to conclude otherwise. See IBP, Inc. v. Alvarez, 546 U.S. 21, 34 (2005) (“identical words used in different parts of the same statute are generally presumed to have the same meaning”). By contrast, when Congress wanted, in the Prevention of Significant Deterioration statute, to refer to a broader set of pollutants than the NAAQS pollutants, it did so expressly. Thus, a facility that requires a pre-construction permit because of its emissions of NAAQS pollutants must employ the best available control technology for emissions not just of “air pollutants” but of “each pollutant subject to regulation 8 under this chapter,” which – now that EPA has regulated greenhouse gases in other parts of the Clean Air Act – does include greenhouse gases. 42 U.S.C. § 7475(a)(4). By its terms, Section 7475(a)(4) thus applies to greenhouse gases, not just the NAAQS. Importantly, however, Congress did not employ the language “each pollutant subject to regulation under this chapter” in the statutory provision setting forth which facilities must obtain a pre-construction permit, the provision at issue in this case. And the policy distinction drawn in Section 7475(a)(4) is rather intuitive: Congress designed the statute’s permitting requirement based on facilities’ NAAQS emissions, but, once those facilities are subject to the permitting requirement, they must also meet a range of other minimum environmental standards. 2 The overall objectives of the Prevention of Significant Deterioration statute also suggest that “air pollutant” refers to the NAAQS air pollutants for purposes of the permitting requirement. Importantly, the Prevention of Significant Deterioration statute applies only in areas that have met the NAAQS – that is, areas that do not have excessive emissions of the NAAQS air pollutants. If the purpose of this statute were in part to address global warming by requiring pre- construction permits for facilities that emit greenhouse gases, as EPA’s reading suggests, why would the statute target the construction of facilities only in areas that are in compliance with the NAAQS – and not elsewhere in the United States? 2 Section 7479(1) – the definition of “major emitting facility” – speaks of “any” air pollutant. But the word “any” just begs the question of what the term “air pollutant” covers in the Prevention of Significant Deterioration program. It’s either any air pollutant regulated under the Clean Air Act or any of the NAAQS air pollutants. 9 That would make little sense, which in turn further suggests that EPA has misread the statute. Moreover, as its name indicates, the Prevention of Significant Deterioration statute is designed primarily to prevent “deterioration” of an attainment area’s air quality. The relevant air quality standards that define whether an area is in attainment are the NAAQS. In a statute expressly linked to the NAAQS and designed to ensure that air quality does not “deteriorate” with respect to the NAAQS, it is somewhat illogical to read the statute as requiring pre-construction permits simply because a facility may emit substances that will not affect attainment of the NAAQS. Under EPA’s approach, a facility could be covered by the permitting requirement even if it emits no NAAQS air pollutants at all. That, too, makes little sense and suggests that EPA has misread the statute. A separate canon of interpretation further demonstrates that EPA’s broad reading of the term “air pollutant” is at odds with Congress’s design. By requiring a vastly increased number of facilities to obtain pre-construction permits, EPA’s interpretation will impose enormous costs on tens of thousands of American businesses, with corresponding effects on American jobs and workers; on many American homeowners who move into new homes or plan other home construction projects; and on the U.S. economy more generally. Yet there is literally no indication in the text or legislative record that Members of Congress ever contemplated – much less intended – such a dramatic expansion of the permitting requirement of the Prevention of Significant Deterioration statute. Courts do not lightly conclude that Congress intended such major consequences absent some indication that Congress meant to do so. See 10 FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 159-61 (2000). Here, as elsewhere, we should not presume that Congress hid an elephant in a mousehole. For all of those reasons – the statutory text, the absurdity principle, the statutory context as demonstrated by related statutory provisions, the overarching objectives of the statute, the major unintended consequences of a broader interpretation – the Prevention of Significant Deterioration statute as a whole overwhelmingly indicates that the permitting requirement is based on emissions of the NAAQS air pollutants. And just to reiterate, the simple and absolutely dispositive point in this case is the following: The broader interpretation of “air pollutant” adopted by EPA produces what even EPA itself admits are absurd consequences. When an agency is faced with two plausible readings of a statutory term, but one reading would cause absurd results, the agency cannot choose the absurd reading. Here, therefore, EPA was required to adopt the narrower and more sensible interpretation of “air pollutant,” the interpretation limited to the NAAQS pollutants. As the Supreme Court has said, “interpretations of a statute which would produce absurd results are to be avoided if alternative interpretations consistent with the legislative purpose are available.” Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 575 (1982). Such an “alternative interpretation[] consistent with the legislative purpose” is readily available here. II If that were the end of the analysis, I would not hesitate to conclude that EPA had adopted an impermissibly broad reading of the term “air pollutant” for purposes of the 11 permitting provision of the Prevention of Significant Deterioration statute. But before reaching that conclusion definitively, we need to consider whether EPA’s approach was mandated by the Supreme Court’s decision in Massachusetts v. EPA, 549 U.S. 497 (2007). In that case, the Supreme Court considered the general statutory term “air pollutant” as applied to a different aspect of the Clean Air Act – the motor vehicle emissions program. The Court there interpreted “air pollutant” very broadly to mean “all airborne compounds of whatever stripe,” including greenhouse gases. Id. at 529. Does Massachusetts v. EPA dictate EPA’s broader interpretation of “air pollutant” in the different context of the Prevention of Significant Deterioration statute? The panel opinion seemed to think so; its conclusion appears to have been heavily if not dispositively influenced by Massachusetts v. EPA. See, e.g., Coalition for Responsible Regulation, Inc. v. EPA, 684 F.3d 102, 134, 136 (D.C. Cir. 2012). In my view, however, the holding in Massachusetts v. EPA does not control the result in this case. Indeed, as explained more fully below, even EPA has concluded that Massachusetts v. EPA does not control here. The decision in Massachusetts v. EPA concerned the motor vehicle emissions program, a point the Supreme Court expressly noted many times in its opinion. The case did not purport to say that every other use of the term “air pollutant” throughout the sprawling and multi- faceted Clean Air Act necessarily includes greenhouse gases. Each individual Clean Air Act program must be considered in context. 3 3 As an analogy, take the familiar example of “no vehicles in the park.” Assume that a court has decided that the term “vehicles” generally includes bicycles, and that no bicycles are allowed in the 12 Importantly, in Massachusetts v. EPA, the Supreme Court explicitly relied on the fact that the Clean Air Act’s “capacious definition of ‘air pollutant,’” did not appear “counterintuitive” or produce “extreme” consequences in the context of motor vehicle emissions. 549 U.S. at 531-32. But, as explained above, EPA’s capacious definition of “air pollutant” is counterintuitive and does produce extreme consequences in the context of the Prevention of Significant Deterioration statute, as EPA itself acknowledges. Moreover, in this case, an alternative and sensible interpretation of the term “air pollutant” is readily discernible from the text, context, and structure of the Prevention of Significant Deterioration statute as a whole – namely, the NAAQS- specific interpretation. To be sure, as noted earlier, the same words used in different parts of an Act are often construed to have the same meaning. See IBP, Inc. v. Alvarez, 546 U.S. 21, 34 (2005). If that were an inflexible command, the Massachusetts v. EPA interpretation of “air pollutant” would certainly control here park. Next assume that another park regulation states that “all park service vehicles must have reinforced gas tanks.” In that latter regulation, context tells us that the term “vehicles” obviously does not include bicycles. Bicycles are still vehicles in the abstract, but the gas-tank regulation logically applies only to a specific subset of vehicles (namely, motor vehicles). So it is with “air pollutant” as used in different parts of the Clean Air Act. Massachusetts v. EPA held that the term “air pollutant” generally includes greenhouse gases. But that does not mean that the term “air pollutant” can never be used in a narrower sense. Greenhouse gases may qualify as “air pollutants” in the abstract, but context tells us that the Prevention of Significant Deterioration program uses the term “air pollutant” to refer only to a subset of all air pollutants (namely, the NAAQS pollutants). 13 and throughout the entire Clean Air Act. But as the Supreme Court recently reminded us – in the context of interpreting the Clean Air Act – “the natural presumption that identical words used in different parts of the same act are intended to have the same meaning is not rigid and readily yields whenever there is such variation in the connection in which the words are used as reasonably to warrant the conclusion that they were employed in different parts of the act with different intent.” Environmental Defense v. Duke Energy Corp., 549 U.S. 561, 574 (2007) (internal quotation marks and ellipsis omitted). As instructed by the Supreme Court, we must interpret statutory terms based on their context and in light of the statute as a whole, even if that approach on some occasions means that the same term applies differently in different parts of a statute. See, e.g., General Dynamics Land Systems, Inc. v. Cline, 540 U.S. 581, 596-97 (2004) (term “age” has different meanings within Age Discrimination in Employment Act); United States v. Cleveland Indians Baseball Co., 532 U.S. 200, 212-13 (2001) (term “wages paid” has different meanings within Social Security Act Amendments of 1939); Robinson v. Shell Oil Co., 519 U.S. 337, 343-44 (1997) (term “employee” has different meanings within Title VII). The Supreme Court’s application of that interpretive principle in Environmental Defense v. Duke Energy – a decision issued on the same day as Massachusetts v. EPA – is illuminating. There, the Supreme Court confronted the Clean Air Act’s definition of a stationary source “modification.” See 549 U.S. at 567-68. That term was relevant to both the New Source Performance Standards program and the Prevention of Significant Deterioration program. The Court ruled that EPA could interpret the term “modification” differently for each of those two Clean Air Act programs, even though “the terms share a common statutory definition.” 14 Id. at 574. In so holding, the Court analyzed the two programs’ different regulatory goals, noting that a “given term in the same statute may take on distinct characters from association with distinct statutory objects calling for different implementation strategies.” Id. The Supreme Court’s interpretive approach in Environmental Defense v. Duke Energy – which recognizes that the meaning of a statutory term in the Clean Air Act may vary based on the particular program at issue – shows that the Massachusetts v. EPA interpretation of “air pollutant” in the context of the motor vehicle emissions program does not necessarily require the same interpretation of “air pollutant” in the context of the Prevention of Significant Deterioration program. In Massachusetts v. EPA, the Supreme Court emphasized that the regulation of greenhouse gases in the motor vehicle emissions program would not be “counterintuitive” and would not lead to any “extreme measures.” 549 U.S. at 531. Greenhouse gas standards would simply be added to the other regulations already applicable to manufacturers of new motor vehicles, and any such standards would take into account both cost and technological feasibility. See 42 U.S.C. § 7521(a). By contrast, the regulation of greenhouse gases in the Prevention of Significant Deterioration program would be both counterintuitive and extreme. Tens of thousands of businesses and homeowners would be swept into the Clean Air Act’s purview for the first time and hit with permitting costs averaging $60,000, not to mention the additional costs of trying to construct and maintain the facility in compliance with the relevant emissions limits and technological standards. See 75 Fed. Reg. 31,514, 31,556 (June 3, 2010). In addition, the costs associated with a vastly expanded permitting requirement would deter numerous projects from 15 even starting in the first place. The major differences between the motor vehicle emissions program and the Prevention of Significant Deterioration program thus neatly fit the Environmental Defense v. Duke Energy paradigm of “distinct statutory objects calling for different implementation strategies.” In reaching that conclusion, it bears mention that the Clean Air Act is a very complicated statute encompassing several distinct environmental programs. It is no surprise, then, that the motor vehicle emissions program and the Prevention of Significant Deterioration program are not the only parts of the Act to employ a term like “air pollutant” in a context-dependent way. For example, the visibility program applies to facilities based on their emissions of “any pollutant.” 42 U.S.C. § 7491(g)(7). In the context of that program, EPA has interpreted the term “any pollutant” to mean “any visibility-impairing pollutant,” which obviously does not include greenhouse gases. 40 C.F.R. pt. 51, App. Y, § II.A. Similarly, the nonattainment program applies to areas that have been designated as nonattainment “for any air pollutant.” 42 U.S.C. § 7501(2). In the context of that program, the term “air pollutant” is logically limited to the NAAQS air pollutants, which are the only pollutants for which an area can be designated as nonattainment. Id. § 7407(d)(1)(A). All of that simply underscores that a court should exercise caution before reflexively importing the interpretations applicable to one Clean Air Act program into a distinct Clean Air Act program. Any lingering doubt that Massachusetts v. EPA does not control here is dispelled when we recall that EPA itself has rejected Massachusetts v. EPA’s interpretation of “air pollutant” for the Prevention of Significant Deterioration 16 statute. The Court in Massachusetts v. EPA said that “air pollutant” meant “all airborne compounds of whatever stripe.” 549 U.S. at 529. EPA has acknowledged, however, that such a broad definition cannot possibly extend to the use of the term “air pollutant” in the Prevention of Significant Deterioration statute. EPA understood that it would be absurd to require pre-construction permits because of emissions of any airborne compound, including emissions of airborne compounds that have not been deemed harmful and regulated under the Clean Air Act. To avoid rendering the Prevention of Significant Deterioration statute an absurdity, EPA construed “air pollutant” to mean certain air pollutants – in particular, “any regulated air pollutant.” The critical point for present purposes – and it really is a critical point in thinking about the significance of Massachusetts v. EPA to the present case – is that EPA itself recognized that the Massachusetts v. EPA definition of “air pollutant” cannot and does not control how to interpret “air pollutant” in the Prevention of Significant Deterioration context. As it tries to justify its broad interpretation of the Prevention of Significant Deterioration statute, EPA cannot simultaneously latch on to Massachusetts v. EPA and reject Massachusetts v. EPA. If Massachusetts v. EPA does not control here – and even EPA admits that it does not – then we are back where we started. EPA was faced with two initially plausible interpretations of “air pollutant” for purposes of the permitting requirement of the Prevention of Significant Deterioration statute. One interpretation created patent absurdities and made little sense given the other statutory provisions. The other interpretation fit comfortably and sensibly within the statutory text and context. EPA 17 nonetheless chose the first option. In my view, EPA’s reading of the statute was impermissible. An agency cannot adopt an admittedly absurd interpretation and discard an eminently sensible one. Given all of this, the case seems reasonably straightforward. So how did the panel opinion reach the opposite conclusion? I respectfully have three main points of disagreement. First, as I read it, the panel opinion was decisively influenced by Massachusetts v. EPA’s interpretation of “air pollutant” in the context of the motor vehicle emissions program. But in light of the material differences between the motor vehicle emissions program and the Prevention of Significant Deterioration program, the Massachusetts v. EPA interpretation cannot control here, as even EPA acknowledges. Second, the panel opinion attempted to buttress its choice of a broad interpretation of the term “air pollutant” by pointing to Section 7475(a)(4), the provision in the Prevention of Significant Deterioration program requiring covered facilities to use the best available control technology. But as explained above, Section 7475(a)(4) actually cuts the other way because it specifically refers to “each pollutant subject to regulation under this chapter,” which now does include greenhouse gases – whereas, by contrast, other statutory provisions in the Prevention of Significant Deterioration program clearly employ a NAAQS-specific interpretation of the unadorned term “air pollutant.” Third, the panel gave insufficient weight to the most critical point in this case, the absurd consequences of EPA’s broad interpretation. This was a mistake because the ultimate clincher in this case is one simple point: EPA chose an admittedly absurd reading over a perfectly natural reading of the relevant statutory text. An agency cannot do that. 18 III In finding EPA’s statutory interpretation legally impermissible, I do not in any way want to diminish EPA’s vital policy objectives. EPA’s regulations for the Prevention of Significant Deterioration statute may well be a good idea as a matter of policy. The task of dealing with global warming is urgent and important. But as in so many cases, the question here is: Who Decides? The short answer is that Congress (with the President) sets the policy through statutes, agencies implement that policy within statutory limits, and courts in justiciable cases ensure that agencies stay within the statutory limits set by Congress. A court’s assessment of an agency’s compliance with statutory limits does not depend on whether the agency’s policy is good or whether the agency’s intentions are laudatory. Even when that is true, we must enforce the statutory limits. See Hamdan v. United States, 696 F.3d 1238 (D.C. Cir. 2012) (ruling that Executive Branch exceeded statutory authority in wartime prosecution of al Qaeda member). In cases like this one, the bedrock underpinnings of our system of separation of powers are at stake. To be sure, courts must be wary of undue interference with an agency’s action implementing its statutory responsibilities. See American Radio Relay League, Inc. v. FCC, 524 F.3d 227 (D.C. Cir. 2008) (separate opinion of Kavanaugh, J.); see also Desert Citizens Against Pollution v. EPA, 699 F.3d 524 (D.C. Cir. 2012); National Environmental Development Association’s Clean Air Project v. EPA, 686 F.3d 803 (D.C. Cir. 2012); American Petroleum Institute v. EPA, 684 F.3d 1342 (D.C. Cir. 2012); ATK Launch Systems, Inc. v. EPA, 669 F.3d 330 (D.C. Cir. 2012); Natural Resources Defense Council v. EPA, 661 F.3d 662 (D.C. Cir. 2011); Medical 19 Waste Institute & Energy Recovery Council v. EPA, 645 F.3d 420 (D.C. Cir. 2011). To take one salient and important example, the statutory scheme gives EPA significant discretion in setting the NAAQS for the NAAQS air pollutants – a discretion the courts must respect. But at the same time, undue deference or abdication to an agency carries its own systemic costs. If a court mistakenly allows an agency’s transgression of statutory limits, then we green-light a significant shift of power from the Legislative Branch to the Executive Branch. The Framers of the Constitution did not grant the Executive Branch the authority to set economic and social policy as it sees fit. Rather, the Framers gave Congress, along with the President, that legislative role (subject to constitutional limits), and they assigned the Executive Branch the executive power to issue rules and enforce the law within the limits set by Congress. 4 It is true that the legislative process can be cumbersome and frustrating, and the Executive Branch often is well- intentioned in wanting to address pressing policy concerns quickly, before the sometimes glacial congressional machinery can be stirred to action. 5 The legislative process 4 In protecting national security, the Executive has some Article II authority to act in certain circumstances in the Nation’s defense even without specific congressional authorization. This is known as Youngstown category two. See Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 637 (1952) (Jackson, J., concurring). There is no general Youngstown category two authority in the domestic social and economic realms, where the Executive must have statutory authority in order to act. 5 In 2009, the House of Representatives passed a global warming bill that was supported by the President. But the Senate did not pass it. In the early 2000s, Senators McCain and Lieberman 20 can be slow because the Constitution makes it far harder to enact legislation than to block it: Under the Constitution, three different entities must agree in order to enact legislation – the House, the Senate, and the President (or two-thirds of both the House and the Senate to override a President’s veto). But the Framers knew the legislative process would be laborious. They designed it that way. The time and difficulty of enacting new legislation has never justified an agency’s contravention of statutory limits. The Framers specifically contemplated, moreover, that there would be situations where the Executive Branch confronts a pressing need that it does not have current authority to address. In those circumstances, the Constitution’s Recommendations Clause provides that the President may “recommend” to Congress “such Measures as he shall judge necessary and expedient.” U.S. CONST. art. II, § 3. Importantly, the separation of powers and checks and balances of our system are designed not just to ensure that the Branches operate within the proper spheres of their authority, but also to protect individual liberty. As the Supreme Court has explained many times, “while a government of opposite and rival interests may sometimes inhibit the smooth functioning of administration, the Framers recognized that, in the long term, structural protections against abuse of power were critical to preserving liberty. . . . The failures of . . . regulation may be a pressing national problem, but a judiciary that licensed extraconstitutional government with each issue of comparable gravity would, in the long run, be far worse.” Free Enterprise Fund v. Public Company Accounting sought to pass global warming legislation, but no law was ultimately enacted. Numerous other bills have been introduced over the years, and various legislative efforts are ongoing. 21 Oversight Board, 130 S. Ct. 3138, 3157 (2010) (internal quotation marks, alterations, and citations omitted). As a court, it is not our job to make the policy choices and set the statutory boundaries, but it is emphatically our job to carefully but firmly enforce the statutory boundaries. That bedrock separation of powers principle accounts for my concern about this case. Here, as I see it, EPA went well beyond what Congress authorized for the Prevention of Significant Deterioration statute. I respectfully disagree with the panel’s resolution of this issue, and given the overall importance of the case, I respectfully dissent from the denial of rehearing en banc.