Air pollution has no respect for state borders.1 Harmful pollutants generated by factories, power plants, and other sources in upwind states may travel hundreds of miles to reach downwind states, which receive no economic benefit in return for their degraded air quality.2 Last Term, in EPA v. EME Homer City Generation, L.P.,3 the Supreme Court upheld the Environmental Protection Agency’s (EPA) most recent interpretation of the Clean Air Act’s4 “Good Neighbor Provision,”5 which calls on the states to reduce their emissions of pollutants that prevent other states from attaining the minimum national air quality standards.6 In an apparent environmental victory, the Court allowed EPA to undertake a cost-effective allocation of emissions-reduction responsibilities and to immediately promulgate Federal Implementation Plans to bring those allocations into effect.7 Although the decision is not incompatible with the Court’s famous Whitman v. American Trucking Ass’ns8 opinion, which banned the consideration of costs in setting the national air quality standards,9 it nevertheless signals a troubling shift toward the permissibility of cost considerations in environmental regulation.
The Clean Air Act (CAA) directs EPA to set National Ambient Air Quality Standards (NAAQS) at levels “requisite to protect the public health,”10 establishing maximum allowable concentrations of several air pollutants that are emitted by numerous and diverse sources.11 Each state then has “primary responsibility”12 for attaining these NAAQS by promulgating federally enforceable State Implementation Plans (SIPs) for reducing in-state emissions.13 Under the Good Neighbor Provision, each state’s SIP must also “contain adequate provisions” to prevent in-state emissions “in amounts which will . . . contribute significantly to nonattainment in, or interfere with maintenance by, any other State with respect to [the NAAQS].”14 If EPA disapproves a SIP or finds that a state has failed to submit a SIP by the statutory deadline, EPA must promulgate a replacement Federal Implementation Plan (FIP) within two years of that finding.15
EPA’s attempts at implementing the Good Neighbor Provision have been rejected by the D.C. Circuit more often than not. In Michigan v. EPA,16 the D.C. Circuit upheld EPA’s decision in its “NOx SIP Call” regulation17 to cap reduction requirements at levels that could be achieved with “highly cost-effective controls,”18 despite a dissent from Judge Sentelle arguing that cost considerations were clearly prohibited.19 But in North Carolina v. EPA,20 the D.C. Circuit rejected EPA’s subsequent cap-and-trade-style Clean Air Interstate Rule21 (CAIR) on grounds of impermissible cost considerations and a failure to achieve “measurable progress”22 toward downwind state air quality goals.23 Most recently, in August 2011, EPA replaced CAIR with the Transport Rule,24 imposing mandatory emissions limits on each state that contributed at least one percent of any NAAQS of any downwind state.25 EPA chose state-specific limits based on what it determined were “significant cost thresholds” for achieving noticeable downwind air quality improvements.26 Having previously disapproved the relevant SIPs with respect to the states’ good-neighbor obligations,27 EPA simultaneously promulgated FIPs to immediately impose the appropriate emissions controls, particularly on fossil fuel–fired power plants.28
A divided panel of the D.C. Circuit vacated the Transport Rule in its entirety.29 Writing for the majority, Judge Kavanaugh30 held that EPA had exceeded its statutory authority in two independent ways. First, he held that EPA had crossed three “red lines”31 by effectively or potentially requiring too great an emissions reduction from particular upwind states.32 He reasoned that the Transport Rule might (1) require states to bring their emissions below the one percent contribution threshold that EPA had established in its first stage of analysis,33 (2) require states to make emissions reductions disproportionate to their respective contributions to downwind air pollution,34 and/or (3) require states to do more than necessary to achieve attainment of the NAAQS in downwind states.35 Second, he held that EPA had a duty to give the states extra time to promulgate new SIPs after it had quantified the minimum good-neighbor reductions that SIPs would have to achieve.36 Judge Rogers dissented, arguing that the court lacked jurisdiction to consider the three-red-lines argument on account of the petitioners’ failure to state those objections with reasonable specificity during the comment period.37 She also would have held that EPA’s SIP disapprovals afforded the agency complete discretion to promulgate FIPs “at any time” within the statutory two-year window.38
The Supreme Court reversed.39 In an opinion by Justice Ginsburg,40 the Court held that EPA had reasonably interpreted an ambiguous provision of the Clean Air Act.41 Justice Ginsburg began by acknowledging the complexity of the regulatory problem at issue, noting three major challenges: (1) the difficulty of identifying upwind sources of pollutants, (2) the nonuniformity of pollutant migration given changing winds, and (3) the chemical transformation whereby nitrogen oxide (NOx) and sulfur dioxide (SO2) gases from upwind sources create ozone and fine particular matter (PM2.5) pollution in downwind locations.42 She proceeded to chronicle Congress’s attempts to address interstate air pollution, which culminated in the 1990 Good Neighbor Provision.43 She then endeavored to describe the “two-step approach”44 employed by EPA in its recent Transport Rule.45 The step-one “screening” analysis, she explained, acted to screen out upwind states that contributed no more than one percent of the relevant NAAQS to any downwind state.46 The step-two “control” analysis attempted to achieve a “cost-effective allocation of emission reductions”47 by creating annual emissions “budget[s]”48 for each state based on EPA-determined “significant cost threshold[s].”49
Justice Ginsburg rejected the petitioners’ challenge to EPA’s FIP promulgation on the merits, concluding that the Clean Air Act’s SIP and FIP mandates are unambiguous.50 Rebuking the D.C. Circuit for reaching beyond the text of the statute, she held that nothing in the Clean Air Act calls for altering the SIP schedule based on whether EPA has quantified the good-neighbor obligations.51 She also rejected the argument that EPA had established a practice of giving the states more time after it quantified good-neighbor obligations in the NOx SIP Call and CAIR proceedings, highlighting that “the D.C. Circuit’s North Carolina decision admonished EPA to act with dispatch in amending or replacing CAIR.”52
Justice Ginsburg began her analysis of the Transport Rule with a relatively lengthy review of the Court’s seminal decision in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc.53 She emphasized that it is the agency’s responsibility to fill any “gap left open”54 by statutory ambiguity, and that courts should defer to the administering agency’s construction in such cases unless “arbitrary, capricious, or manifestly contrary to the statute.”55 She concluded that the Good Neighbor Provision indeed contains such a gap, in that it fails to specify how responsibility for a downwind state’s excess pollution should be allocated among contributing upwind states.56 She noted that the provision certainly does not mandate a proportionality approach, as advocated by the D.C. Circuit and by the dissent, and offered a series of simple hypotheticals to prove the inadequacy of such an approach in practice.57 She additionally argued that a proportionality approach would aggravate the potential for overregulation in contravention of the statute.58
Finally, Justice Ginsburg turned to the permissibility of EPA’s Transport Rule.59 She explained that “amounts” was the relevant ambiguous term,60 as it failed to dictate which amounts by which states.61 Given this ambiguity, she concluded that EPA’s Transport Rule calculus was a cost-effective and equitable way of making such an allocation, and she applauded the agency for developing an approach that would require a much lower overall cost and prevent states from free riding on other states’ earlier investments in pollution controls.62 She determined that nothing in the statute required EPA to disregard costs, and that the Transport Rule was a “‘reasonable’ way of filling the ‘gap left open by Congress.’”63 In a footnote, she distinguished American Trucking as dealing with an “absolute”64 provision that provided “express criteria” for setting the NAAQS.65
Justice Scalia penned a sharply worded dissent,66 arguing that the “plain logic” of the statute precludes the cost-based methodology devised by EPA in the Transport Rule, and that the statute is “pregnant with an obligation” for EPA to give the states another chance at their SIPs after quantifying the good-neighbor obligations.67 He asserted that, to the extent the proportionality approach is unworkable, the Good Neighbor Provision would simply be inoperative.68 He further argued that EPA had made no attempt to show that its chosen cost thresholds would not result in “gross over-control,”69 and that, moreover, cost considerations are clearly prohibited under the statute.70 Justice Scalia relied heavily on American Trucking, which he argued dealt with the interpretation of an even more ambiguous Clean Air Act provision than the Good Neighbor Provision.71 He noted that American Trucking had “refused to find implicit in ambiguous sections of the [Clean Air Act] an authorization to consider costs that has elsewhere, and so often, been expressly granted,”72 and that American Trucking thus demanded “a textual commitment of authority to the EPA to consider costs.”73 Finally, he argued that EPA’s early FIP promulgation was an abuse of discretion because of the significant harm it caused to the “cooperative federalism”74 structure of the statute.75
Even after the Court’s endorsement of cost considerations in Entergy Corp. v. Riverkeeper, Inc.,76 its landmark American Trucking decision was still susceptible of a broad reading that prohibited the consideration of costs in all ambiguous provisions of the Clean Air Act. EME Homer signaled the end of that broad reading by rejecting the notion that the structure of the Clean Air Act renders statutory silence with respect to costs unambiguous. Although EME Homer is not incompatible with Whitman, which explicitly left open the consideration of costs in implementation, it confined Whitman to a narrow interpretation with problematic consequences for environmental advocates.
Before EME Homer, the 2001 decision of American Trucking could plausibly be read as repudiating the consideration of costs under the Clean Air Act except where completely unambiguous.77 American Trucking involved a challenge by industry groups and several states to EPA’s promulgation of more stringent NAAQS for ozone and PM2.5 in the late 1990s.78 Faced with a D.C. Circuit opinion that upheld the challenge by resurrecting the long-ignored nondelegation doctrine,79 the Court unanimously endorsed the delegation of discretion to EPA in setting the NAAQS under Section 109 of the Clean Air Act80 even while holding that cost considerations were clearly precluded.81 The Court found it highly implausible that Congress might have buried the power to determine whether cost considerations should weaken national ambient air quality standards in the language “requisite to protect the public health”82 and “adequate margin of safety.”83 It explained that the cost factor is “both so indirectly related to public health and so full of potential for canceling the conclusions drawn from direct health effects”84 that, given that the NAAQS-setting provision is “the engine that drives nearly all of Title I of the [Clean Air Act],”85 cost would have been expressly mentioned had Congress meant it to be considered.86
Despite the apparent textual clarity of Section 109, the Court also announced the broader dictum that it had “refused to find implicit in ambiguous sections of the [Clean Air Act] an authorization to consider costs that has elsewhere, and so often, been expressly granted.”87 In this statement, the Court found coherence between its holding in American Trucking and its previous holding in Union Electric Co. v. EPA,88 which, like EME Homer, dealt with Section 11089 implementation plans,90 thus suggesting that the concept might apply broadly to the entire Clean Air Act. This reading is further bolstered by Justice Breyer’s American Trucking concurrence, which expressed disagreement with what he saw as “the Court’s presumption that any authority the Act grants the EPA to consider costs must flow from a ‘textual commitment’ that is ‘clear.’”91 Indeed, Justice Stevens would later employ American Trucking to argue that “we should not treat a provision’s silence as an implicit source of cost-benefit authority, particularly when such authority is elsewhere expressly granted and it has the potential to fundamentally alter an agency’s approach to regulation.”92
Although the broad reading of American Trucking was made more vulnerable by the Court’s 2009 decision in Entergy, which found cost-benefit analysis permissible under an ambiguous provision of the Clean Water Act,93 it was still not foreclosed prior to EME Homer. Entergy dealt with Section 316(b),94 a rather unique requirement in the Clean Water Act that cooling water intake structures adopt the “best technology available for minimizing adverse environmental impact.”95 In finding that Section 316(b)’s silence as to cost-benefit analysis or any other potentially relevant factors did not preclude EPA from concluding that cost-benefit analysis was permissible, Justice Scalia specifically distinguished American Trucking. He asserted that American Trucking “stands for the rather unremarkable proposition that sometimes statutory silence, when viewed in context, is best interpreted as limiting agency discretion,” and explained that Whitman’s “relevant ‘statutory context’ included other provisions in the Clean Air Act that expressly authorized consideration of costs, whereas § 109 did not.”96 Entergy thus left open the possibility of applying American Trucking broadly to the entire Clean Air Act, even if it foreclosed American Trucking’s applicability to the Clean Water Act.
But in finding cost considerations permissible under an admittedly ambiguous Clean Air Act provision with no express grant of authority to take costs into account, EME Homer signaled the end of that broad reading. The two cases were not so incompatible as Justice Scalia contended in his dissent,97 given that American Trucking expressly left open the permissibility of cost considerations in the development of NAAQS implementation plans.98 EPA’s multistate allocation of upwind emissions reductions under the Good Neighbor Provision is arguably analogous to the states’ allocation of emissions reductions amongst particular in-state sources, as both are constrained by the existing NAAQS. Just as the states can select any combination of emissions reductions so long as the result is compliance with the NAAQS,99 EPA can select any combination of upwind emissions reductions so long as the result is to eliminate “amounts [that] . . . contribute significantly” to other states’ inability to attain the NAAQS.100 In any event, American Trucking’s attention to the differences between Sections 109 and 110 suggests that it is susceptible of a narrow reading that limits its clear-statement requirement to the setting of the NAAQS under Section 109.101
EME Homer confined American Trucking to that narrow reading. Justice Ginsburg could have interpreted American Trucking as relying on the structure of the Clean Air Act to find that statutory silence with respect to cost considerations constitutes an unambiguous prohibition of such considerations. That is, she could have decided EME Homer at Chevron Step One, finding that provisions of the Clean Air Act that are individually ambiguous with respect to cost considerations are rendered unambiguous by the Act’s overall structure. Instead she read American Trucking as relying on the “requisite to protect public health” language to find that the NAAQS-setting provision specifically barred cost considerations by providing one exclusive criterion.102 EME Homer thus demonstrates the Court’s willingness to reach Chevron Step Two, rejecting the idea of a structurally derived unambiguity. Foreshadowing this result, Justice Kagan opined during oral argument that American Trucking had ultimately endorsed a “fundamentally silly”103 means of regulation, and that “most people, everybody” thinks regulating with attention to costs is “better.”104
To the contrary, however, most environmental advocates have deep concerns about the consideration of costs in regulation, especially when that consideration takes the form of cost-benefit analysis.105 Such concerns arise because of the difficulty of quantifying environmental harm, the often antiregulatory slant with which cost considerations are implemented, and the oversimplification of the political discourse when cost considerations become the focal point.106 Given the Clean Air Act’s significance to the protection of public health and the environment across the country, weakening its regulatory strength is likely to have widespread repercussions. Thus, although EPA’s position in EME Homer was supported by numerous environmental advocacy groups,107 its victory opened a doorway to cost considerations that environmental advocates may be fighting for years to come.