In a 5-4 decision, a divided Supreme Court stayed the enforcement of the Environmental Protection Agency’s (“EPA”) Federal Implementation Plan (“FIP”) holding that the emission control measures set to be used in upwind states to improve ozone levels in downwind states was arbitrary and capricious. Ohio et al. v. Env’t Prot. Agency, No. 23A349, No. 23A350, No. 23A351, No. 23A384, 2024 WL 3187768 at *11 (S. Ct. June 27, 2024). The Court reasoned that EPA’s FIP rested on the assumption that all upwind states would adopt emission-reduction measures to ensure effective downwind air quality improvements. EPA, however, failed to reasonably explain if the FIP would be operable if some upwind states fell out of the plan. In the short term, the stay provides temporary relief to industry groups and states that challenged various aspects of the FIP in the D.C. Circuit; in the long term, the Court’s ruling spells trouble for the FIP’s efficacy if and when federal courts rule on the merits of the rule’s legality.     

To control air pollution, the Clean Air Act (“CAA”) directs EPA to set National Ambient Air Quality Standards (“NAAQS”) for certain air pollutants, such as ozone. Under the CAA, a state is required to prepare a State Implementation Plan (“SIP”) to demonstrate how the state will comply with NAAQS. A state’s SIP must also account for emissions that travel outside a state to neighboring downwind states—a  provision often referred to as the “Good Neighbor Provision.” EPA must approve a state’s SIP so long as it meets all applicable requirements, which include compliance with the Good Neighbor Provision. If, however, a SIP fails to meet all applicable requirements, EPA must issue a FIP for a noncompliant state unless the state corrects the deficiencies in its SIP within the timeline required by the CAA.

After EPA set new NAAQS for ozone in 2015, several states proposed SIPs. Several SIPs asserted that they did not need to adopt new emission standards to comply with the Good Neighbor Provision because they were not linked to downwind air-quality problems, or they could identify no additional cost-effective method of controlling emissions beyond those already implemented. In 2022, EPA announced its intention to disapprove twenty-three SIPs on the grounds that the states failed to adequately comply with the Good Neighbor Provision. In accordance with the CAA, EPA issued its proposed SIP disapprovals for public comment before finalizing them. During the public comment period, however, EPA proposed a FIP to bind all twenty-three states. The uniform FIP used emission measures for nitrogen oxide, a precursor to ozone, in upwind states to achieve downwind air quality improvements.

To calculate the appropriate emissions measures EPA used the following methodology: (1) it identified costs associated with emission control measures using nationwide data; (2) then it predicted the extent to which upwind states’ nitrogen-oxide emissions would fall if relevant facilities in each upwind state adopted the measure; and (3) it estimated the extent to which ozone levels would fall on a state-by-state basis. EPA focused on the “knee of the curve,” or the point in which upwind states were likely to produce very little emissions reductions, to determine a uniform level of costs for the emission measures. Once the agency published the proposed FIP for public comment, commenters warned of a potential pitfall: would the uniform level of costs for the emission control measures change if the number of upwind states included in the FIP changed? Commenters questioned whether the curve’s knee was dependent upon the inclusion of all of the upwind states in the FIP to determine cost-effective emission measures.

Despite the comments and pending litigation challenging the SIP disapprovals, EPA issued its final FIP with the addition of a severability provision. The provision allowed the agency to continue to implement the FIP even if states dropped out. Courts, however, stayed the FIP in twelve states, which accounted for over seventy percent of the emissions EPA planned to address with its FIP. After a D.C. circuit court refused to stay the enforcement of the FIP in a state that was challenging its SIP disapproval, the Supreme Court granted certiorari.

Justice Gorsuch delivered the opinion of the Court by first reviewing the standard for a stay. A stay requires the court to determine (1) whether the applicant is likely to succeed on the merits; (2) whether it will suffer irreparable injury without a stay; (3) whether the stay will substantially injure other parties interested in the proceedings; and (4) where the public interest lies. Since both sides had strong arguments about the harms they faced and equities involved, the Court stated that the issuance of a stay hinged on who was likely to prevail at the end of the litigation.

The Court sided with the applicants challenging the Rule, holding that they are likely to prevail on their claim that the FIP was not reasonably explained.  The Court stated that an agency action is arbitrary and capricious, and therefore must be set aside, when it is neither reasonable nor reasonably explained.  The issue here was the latter.  In the Court’s view, public comments had put the agency on notice of an overarching concern regarding the FIP:  if a state were to drop out of the FIP, due to a court’s reversal of any one of the SIP disapprovals for example, it would potentially change the uniform level of costs because the state’s particular technologies and industries would drop out too.  But rather than “address the applicants’ concern” on this issue in its final rule, EPA “sidestep[ped] it” and failed to respond in substance to the concern.  Indeed, the Court put particular emphasis on the fact that the government could not represent with certainty whether the cost-effective analysis would yield the same results if conducted for just one state. As a result, Justice Gorsuch concluded that the applicants are likely to prevail on the claim that the FIP was not reasonably explained.

Justice Gorsuch rejected three alternative arguments raised by the government. First, the government asserted that it had offered a reasonable response to applicants’ concern, which was the severability provision.  But in the Court’s view, the severability provision was beside the point. EPA needed to explain why its FIP would continue to offer cost-effective improvements in downwind air quality if only a subset of the states originally covered were included. Next, the government asserted that if the FIP lacked a reasoned response to the applicants’ concern it was because the concern was not raised with reasonable specificity during the public comment period. Justice Gorsuch clarified that the reasonable specificity requirement was more a question of notice rather than if an identical argument was raised during the public comment period. In this case, EPA had notice because commenters alerted the agency that should some states no longer participate in the FIP, the agency would need to conduct a new assessment and modeling of contribution to determine what emission control measures maximized cost effectiveness for downwind ozone air quality improvements. Finally, Justice Gorsuch rejected the government’s procedural argument that the applicants needed to file a motion for reconsideration of the final rule before presenting the objection in court. Justice Gorsuch reasoned that EPA was aware of the applicants’ objections when it added the severability provision to the final FIP. Thus, the applicants were not required to return to EPA to again raise a concern that EPA already had a chance to address. Based on these findings, Justice Gorsuch stayed the enforcement of the FIP pending the review of applicants’ SIP disapprovals in lower courts.

Justice Barrett delivered the dissenting opinion, arguing, in pertinent part, that the challengers did not meet the standard for emergency relief. Justice Barrett opined that the challengers were unlikely to prevail on the merits because EPA expressly explained its methodology. Specifically, EPA derived technology performance averages based on nationwide data and applied it to the relevant industry standard on a uniform basis for each emitter in the covered states. EPA only assessed the FIP’s impact based on the particular states included to ensure the emissions limits did not result in more reductions than necessary to achieve the NAAQS. Thus, EPA’s site-specific data in the overcontrol analysis did not mean that the entire FIP’s emission limits depended on the number of states it covered. Finally, Justice Barrett defended the severability provision by reminding the majority that if the FIP were not severable, EPA would have to go back to the drawing board whenever a single state were removed from the FIP. This would be contrary to EPA’s obligation to ensure compliance with all NAAQS by the next applicable attainment date.

Notwithstanding the dissent, the Court’s decision halts EPA’s enforcement of the FIP for the foreseeable future, and the Court’s holding that the challengers of the FIP are likely to succeed on the merits does not bode well for the long-term future of the FIP.  Indeed, the Court’s willingness to chime in on the merits of the FIP based upon a petition for emergency relief stands apart from how the Court often operates as a decision-maker of last resort.  The decision is one of a series issued by the Court this term skeptical of agency overreach. 

Share | Email| Print| Comments|  Topics: Administrative Procedures Act, Clean Air Act, United States Supreme Court