375 F.3d 537
SIERRA CLUB, Petitioner,
v.
ENVIRONMENTAL PROTECTION AGENCY, Respondent.
Intervening Respondents: State of Illinois; State of Missouri
No. 03-2839.
No. 03-3329.
United States Court of Appeals, Seventh Circuit.
Argued June 1, 2004.
Decided July 6, 2004.
Douglas R. Williams (argued), St. Louis University, Kathleen G. Henry, Bruce A. Morrison, Great Rivers Environment Law Center, St. Louis, MO, for Petitioner.
David A. Ullrich, E.P.A., Office of the Regional Counsel, Chicago, IL, Eileen T. McDonough (argued), Department of Justice, Environmental & Natural Resources Div., Washington, DC, for Respondent.
Mary E. Welsh, Office of the Attorney General, Chicago, IL, Timothy P. Duggan, Jefferson City, MO, for Intervenor.
Before EASTERBROOK, KANNE, and ROVNER, Circuit Judges.
EASTERBROOK, Circuit Judge.
Changes to the environmental laws in 1990 reduced the allowable levels of ozone pollution and set deadlines for attainment. Clean Air Act Title I, Part D, subpart 2, 42 U.S.C. §§ 7511 to 7511f. The St. Louis metropolitan area, initially classified as a "moderate" nonattainment zone, had until November 15, 1996, to comply. 42 U.S.C. § 7511(a)(1). A moderate jurisdiction that missed this deadline was to be reclassified automatically as a "serious" nonattainment area, 42 U.S.C. § 7511(b)(2)(A). That change requires additional costly anti-pollution steps. One of the principal differences between the 1990 legislation and its predecessor was this mandatory reclassification; the legislation leaves the EPA less discretion with respect to ozone than other pollutants. See Whitman v. American Trucking Associations, Inc., 531 U.S. 457, 481-86, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001). Nonetheless, when St. Louis failed to meet the deadline, the EPA decided that it had done well enough that its status should remain unchanged. In Sierra Club v. EPA, 311 F.3d 853 (7th Cir.2002), we held that dispensation unlawful and directed the EPA to apply the statute as written. See also Sierra Club v. EPA, 294 F.3d 155 (D.C.Cir.2002). Delay in compliance required a turn of the screw even though St. Louis was making progress.
While the proceedings that led to our 2002 decision were under way, St. Louis finally met the ozone standards. It asked the EPA for a formal decision that it satisfies the requirements for ozone. Before designating any area as in compliance, the EPA must make five determinations:
The Administrator may not promulgate a redesignation of a nonattainment area (or portion thereof) to attainment unless —
(i) the Administrator determines that the area has attained the national ambient air quality standard;
(ii) the Administrator has fully approved the applicable implementation plan for the area under section 7410(k) of this title;
(iii) the Administrator determines that the improvement in air quality is due to permanent and enforceable reductions in emissions resulting from implementation of the applicable implementation plan and applicable Federal air pollutant control regulations and other permanent and enforceable reductions;
(iv) the Administrator has fully approved a maintenance plan for the area as meeting the requirements of section 7505a of this title; and
(v) the State containing such area has met all requirements applicable to the area under section 7410 of this title and part D of this subchapter.
42 U.S.C. § 7407(d)(3)(E). The EPA made all of these findings in 2003 for the St. Louis metropolitan area (which spans the border between Missouri and Illinois, and hence requires consideration of multiple state plans). See 68 Fed.Reg. 25418 (May 12, 2003) (Missouri), 68 Fed.Reg. 25442 (May 12, 2003) (Illinois). The Sierra Club asks us to set aside these decisions. It does not contest the agency's finding that the St. Louis area now meets the national ambient air quality standard for ozone. Nor does it challenge the vital third finding: that "the improvement in air quality is due to permanent and enforceable reductions in emissions". But it insists that St. Louis lacks a proper "applicable implementation plan for the area under section 7410(k)" (requirement (ii)) and that the area's maintenance plan (requirement (iv)) does not meet all requirements of § 7505a. We start with the challenge to the maintenance plan.
A maintenance plan must take into account the sort of things, such as population growth and changes to the industrial base, that might cause existing pollution-control measures to become inadequate in the future even if they served well in the past. Ozone at or near ground level comes principally from chemical reactions involving its precursors, nitrous oxides (NOx) and volatile organic compounds. Both implementation plans and maintenance plans thus must provide for controls on the emission of these precursors. But what emissions are likely in the future, and what steps could reduce them by the required amount? Accurate projections depend on supplying good data to good models. All concrete requirements of § 7505a, to which requirement (iv) refers, have been satisfied. But a maintenance plan serves as an amendment to the local implementation plan, and 42 U.S.C. § 7511a(c)(2)(A) and (j)(1) thus may affect it. These subsections require both multi-state areas and serious nonattainment areas to "use photochemical grid modeling or any other analytical method determined [by the EPA], in [its] discretion, to be at least as effective." Missouri and Illinois have not promised to use photochemical grid modeling as part of their maintenance endeavors. The Sierra Club insists that all multi-state areas must use photochemical grid modeling as long as their maintenance plans are in effect. The EPA thinks that other tools can suffice — and the Sierra Club does not dispute this at the factual level. It contends, rather, that photochemical grid modeling is essential no matter how thorough and rigorous the maintenance plan may be. Unless the EPA makes a formal determination that some other modeling system is "at least as effective" — and the EPA did not make such a finding, even though it appears to believe that the St. Louis region's methods are at least as effective — then the method named in the statute is indispensable.
Thus we have a straightforward issue: must every multi-state area use photochemical grid modeling continually (at least until a formal equivalence finding has been made)? The EPA's view does not contradict the statute: § 7511a does not refer to maintenance plans at all, and it is only through the back door (because the maintenance plan amends the implementation plan) that this section enters the picture. What is more, § 7511a deals with pre-attainment requirements. This is the foundation of the agency's view that an area need not use photochemical grid modeling as part of a maintenance plan. That is not an inevitable reading of the statute, but the EPA receives the benefit of deference under Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), which American Trucking held applicable to the ozone subchapter. See 531 U.S at 481, 121 S.Ct. 903. Congress required nonattainment areas to shoulder more substantial burdens. St. Louis, like any other place that wants to keep a valuable attainment designation, has every incentive to choose adequate modeling tools. If it does not, and as a result slips out of compliance as its population or industry changes, then it must pay a steep price for backsliding. It is sensible for the federal agency to give localities that must pay the piper some opportunity to call the tune. Methods of projecting developments differ in accuracy and cost; allowing the affected region to make a cost-benefit comparison has much to recommend it. Recall that the Sierra Club does not contend that Missouri and Illinois have chosen irresponsibly, but only that the statute grants them (and the EPA) no option. The EPA's approach has received the approbation of the Sixth Circuit. See Wall v. EPA, 265 F.3d 426, 436 (6th Cir.2001). We see no reason to create a conflict.
Nor are we persuaded by the Sierra Club's argument that the maintenance plans fail to describe all contingency measures that may be applied if problems arise. The statute does not call for any particular degree of precision in the period after attainment (contrast § 7502(c)(9), which demands "specific measures" in the pre-attainment period), so again the EPA (and the affected states) had choices to make, choices that may be gainsaid only if obviously misguided. Intelligent decisions may depend on the nature of future developments. Missouri and Illinois have committed themselves to action; that they have reserved some discretion about the means does not spoil their plans. See Greenbaum v. EPA, 370 F.3d 527, 538-42, 2004 U.S.App. Lexis 10785 *24-38 (6th Cir. June 3, 2004). Cf. BCCA Appeal Group v. EPA, 355 F.3d 817 (5th Cir.2003).
Let us turn, then, to requirement (ii): that "the Administrator has fully approved the applicable implementation plan for the area under section 7410(k) of this title". This mentions § 7410(k), but the cross-reference is unilluminating. It does not answer one vital question: what kind of implementation plan is "applicable"? Although the parties have disputed many technical issues, most of their disagreement boils down to a single question: Is an "applicable" plan the same as the area's pre-attainment plan (as Sierra Club contends), or is it limited to those measures that have proved to be necessary to achieve compliance (the EPA's view)? The Sierra Club contends that every attainment plan for an area at the serious level must specify the implementation of all reasonably available control measures (though the D.C. Circuit disagrees, see Sierra Club, 294 F.3d at 162-63), and must ensure a 15% reduction in the emission of volatile organic compounds (though the Tenth Circuit disagrees, see Sierra Club v. EPA, 99 F.3d 1551, 1555-56 (10th Cir.1996)), but these are sidelights. The real dispute is whether St. Louis, having been promoted to the serious category by delay in meeting the national ozone standard, must use control measures appropriate to a serious nonattainment area as a condition of being designated as an attainment area.
In a nutshell, the EPA's view is that the "applicable" plan requires an area to continue doing whatever worked, and nothing more. In other words, the EPA wants the plan to contain all provisions that required some set of controls to be in place before the date the area met the national standard. Here's a concrete example, and a principal bone of contention between the parties. Moderate nonattainment areas must ensure that every source of more than 100 tons (annually) of precursor chemicals takes prescribed steps to curtail their emission. For serious nonattainment areas, the threshold falls to 50 tons. 42 U.S.C. § 7511a(c). Our 2002 decision concluded that St. Louis must be treated as a serious nonattainment area — although it was classified as a moderate area in 2000, when it filed the application to be reclassified as an attainment area. Even the plan applicable to a serious nonattainment area would allow newly covered sources some lead time to limit their emissions. 42 U.S.C. § 7511(a)(1), (a)(5). Before that time arrived, St. Louis met the national ozone standard. Sierra Club believes that businesses in the 50 to 100-ton range still must implement controls; the EPA believes that an "applicable" implementation plan need not require this. The Sierra Club's definition of "applicable," by contrast, is "whatever should have been in the plan at the time of attainment" rather than "whatever actually was in the plan and already implemented or due at the time of attainment."
Because the statute does not define "applicable," there is no ineluctable basis for a choice between these options. See Wall, supra, 265 F.3d at 438-40. Both are conceivable understandings of the law. Chevron therefore affords the EPA leeway. (The St. Louis designation is the result of notice-and-comment rulemaking under explicit statutory delegation; this is the core of Chevron's domain. See United States v. Mead Corp., 533 U.S. 218, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001).) The EPA's view is at least as sensible as the Sierra Club's, likely more so. Requirement (ii) reads: "the Administrator has fully approved the applicable implementation plan for the area under section 7410(k)". That's a curiously indirect way of requiring a plan to continue without change, or become more onerous. Why didn't the statute say: "the Administrator has determined that the area will continue to abide by the implementation plan that was, or should have been, in place"? A word such as "applicable" implies that there may be differences between the contents of the pre-attainment plan and those required for the post-attainment period. Against this the Sierra Club points to the way "applicable" is used in other parts of the Clean Air Act (e.g., § 7511a(i)), but that statute is too complex a compromise, and has been amended too many times, to indulge the assumption that all of its words must be used consistently in all of its subsections. "Applicable" is a protean word that takes color from context; it lacks a single, enduring meaning.
Under the Sierra Club's view, compliance does not have a payoff: the residents and businesses of St. Louis must take the same costly steps that would be required had the area been less successful. As the reason to take additional steps was to achieve an adequate reduction in ozone, it would be odd to require them even when they turned out to be unnecessary. Some parts of the Clean Air Act forbid cost-benefit analysis, see Union Electric Co. v. EPA, 427 U.S. 246, 96 S.Ct. 2518, 49 L.Ed.2d 474 (1976) — and subpart (2) on ozone is one of these, to the extent it mandates the progression to more severe controls until compliance has been achieved — but when the statute is ambiguous the EPA is free to take costs into account. That's the upshot of Chevron. See also Cass R. Sunstein, Risk and Reason: Safety, Law, and the Environment (2002); Stephen Breyer, Breaking the Vicious Circle: Toward Effective Risk Regulation (1993). The agency's approach strikes us as sensible.
Much of the Sierra Club's argument assumes that reclassification of St. Louis to serious nonattainment was some sort of punishment that its residents should not be allowed to escape. Not at all. "The St. Louis region" is an abstraction, a convenient collective phrase for millions of people whose own lives and fortunes are at issue. Reclassification was a combination of (a) goad (clean up or suffer expensive measures), and (b) palliative (sterner measures expedite compliance). Once an area has meet the national air quality standard, neither rationale calls for extra stringency; indeed, the statutory system would not be much of a goad if the tighter controls must continue even after attainment. It is not as if neighborhood bakeries and other smallish point sources were themselves blameworthy and in need of 20 lashes for transgressions.
One final subject requires brief comment. While the EPA was considering St. Louis's request for designation as an attainment area for ozone, the D.C. Circuit vacated some elements of the EPA's national regulations for the control of nitrous oxides (known as the NOx SIP Call). See Michigan v. EPA, 213 F.3d 663, 685, 695 (D.C.Cir.2000). After redesignating St. Louis, the EPA revised the NOx SIP Call. See 69 Fed.Reg. 21604 (Apr. 21, 2004). If the regulation adopted in 2004 implies a revision of the implementation and maintenance plans for the St. Louis region, the EPA will need to take action. But the Sierra Club's petitions for review do not raise any question about the implementation of the NOx SIP Call in St. Louis.
The petitions for review are denied.