The legal world abhors a vacuum. Thus, when the Supreme Court overruled Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc.,1 “a leading decision prescribing the standard of review”2 in statutory interpretation cases that had garnered over 18,000 citations by federal courts,3 scholars immediately began trying to chart what might follow.4 Last Term, in Seven County Infrastructure Coalition v. Eagle County,5 the Court laid down an early marker. Most immediately, the Court restrained the National Environmental Policy Act of 19696 (NEPA) by both requiring greater deference to agency decisionmaking in NEPA cases and limiting the required environmental assessment to the “effects of the project at issue.”7 Yet the way in which the Court reined in NEPA — categorizing the Court’s review as one of agency discretion rather than one of pure statutory interpretation — announces a worrying new phase in the post–Chevron deference world.8 Specifically, because the gap between review of law and review of discretion is larger under Loper Bright Enterprises v. Raimondo9 than it had been under Chevron, the Court can now, seemingly at will, use that threshold question of type of review to make unprincipled, statute-by-statute determinations of what level of deference an agency will receive.
In May 2020, the Seven County Infrastructure Coalition (“Coalition”)10 petitioned the Surface Transportation Board (“Board”) for approval to construct “an approximately 88-mile railroad line.”11 The railroad would transport primarily oil from the Uinta Basin, a roughly 12,000–square mile area in northern Utah and Colorado rich in crude oil and other minerals, to the “national rail network.”12 From there, the oil could be carried onward to refinement facilities, largely on the Gulf Coast.13 That rail connection would provide a major efficiency upgrade to the status quo, wherein trucks transport the Basin’s crude oil over high mountain passes.14 The Board conditionally approved the project in January 202115 but, as mandated by NEPA, withheld final approval until an environmental impact statement (EIS) was completed.16
Following a comment period that included six public meetings and over 1,900 comments, the Board published an EIS of over 3,600 pages in August 2021.17 The EIS addressed potential “significant and adverse impacts” of the railway, such as “disruptions to local wetlands, land use, and recreation,” as well as “minor impacts” like “air pollution and big-game movement around the construction site.”18 However, the EIS “did not fully analyze”19 the upstream or downstream effects that would predictably result from increasing oil transportation capacity: greater oil drilling and refining.20 The EIS mentioned that, on the high end, activities made possible by the project could contribute “up to approximately 0.8% of nationwide [greenhouse gas] emissions and 0.1% of global [greenhouse gas] emissions.”21 Nevertheless, according to the reviewing court, the Board did not fully consider these effects in its ultimate decision.22 Following the issuance of the EIS, the Board granted final approval for the railway in December 2021.23 With the project set to move ahead, environmental groups and Eagle County, Colorado — through which more oil would travel if the railway were to be built — each filed petitions for review in the D.C. Circuit.24
The D.C. Circuit vacated the Board’s final approval for the project.25 The panel first found both standing and jurisdictional requirements satisfied.26 Using the “arbitrary [and] capricious” standard from § 706(2)(A) of the Administrative Procedure Act27 (APA), the panel proceeded to address a number of objections raised by the petitioners under NEPA and other statutes.28 Most notably, in considering the NEPA claims, the panel agreed with the petitioners that “certain upstream and downstream” effects such as increased oil combustion were “reasonably foreseeable impacts” of the project that the final EIS “ignored.”29 Concluding that the “significant” issues with how the EIS toted up the project’s environmental effects constituted a violation of NEPA, the D.C. Circuit vacated the final order approving the project, vacated the EIS in part, and remanded the matter to the Board.30 The Coalition and the railway filed a petition for a writ of certiorari with the Supreme Court.31
The Court reversed.32 In a majority opinion written by Justice Kavanaugh,33 the Court first held that the D.C. Circuit failed to give the Board “the substantial judicial deference required in NEPA cases.”34 Given that lower courts had experienced “continuing confusion and disagreement” on how to review NEPA cases, the Court sought to “clarify the fundamental principles of judicial review applicable in those cases.”35 After Loper Bright, the Court noted, courts must review an agency’s interpretation of questions of law de novo.36 However, rather than finding a question of law, the Court characterized the determination of “whether a particular [EIS] is detailed enough”37 as “primarily [an] issue[] of fact.”38 The Court then assessed what qualifies as “significant environmental impacts and feasible alternatives” under NEPA as a “predictive and scientific judgment[].”39 Given the statute’s construction, the Court concluded that “[c]ourts should afford substantial deference and should not micromanage those agency choices” that are reasonable.40 The D.C. Circuit thus erred in its more searching review of the Board’s action.41 Turning to policy implications, the Court contended that NEPA’s growth from a “legislative acorn . . . into a judicial oak” had caught projects in endless delays such that “fewer projects ma[d]e it to the starting line [of construction].”42 Finally, because NEPA is a “purely procedural statute,” even a project with a deficient EIS can proceed if there is no “reason to believe that the agency might disapprove the project if it added more to the EIS.”43
Separately, as a matter of statutory interpretation, the majority found that the D.C. Circuit erred in requiring analysis of “upstream oil drilling or downstream oil refining” under NEPA.44 Rather, “the textually mandated focus” of NEPA is the “immediate project under consideration” (Project A) and not, critically, any “other future or geographically separate project[] that may be built . . . as a result” (Project B).45 The Court gave two reasons for this interpretation. First, Project A does not “proximate[ly] caus[e]” the environmental impacts of Project B, even if such impacts are “factually foreseeable” consequences of Project A,46 such that Project B’s impacts are not part of the “proposed action” for NEPA purposes.47 Rather, when deciding which environmental effects to consider, the agency gets to set “a ‘manageable line’ . . . that [includes] the effects of the project at hand, but not the effects of projects separate in time or place.”48 Second, in an EIS, agencies need only “analyze the effects of projects over which they . . . exercise regulatory authority” to fulfill their NEPA obligations.49 Because the Board could not approve subsequent drilling projects in the Uinta Basin or refining projects along the Gulf Coast, it did not have to analyze the effects of those activities.50
Justice Sotomayor concurred in the judgment,51 finding for the Coalition on narrower grounds.52 Following Department of Transportation v. Public Citizen,53 Justice Sotomayor interpreted NEPA to require consideration of environmental effects only when “such consideration would result in information on which the agency could act.”54 Here, the Board’s organic statute “contains a clear presumption in favor of approving new railways” and, “[a]s common carriers,” operational railways must serve any party who pays for their services.55 Thus, the Board “could not lawfully consider” upstream or downstream environmental effects in deciding whether to approve a railway project.56 Taken together, for Justice Sotomayor, the Board’s unique statutory obligations spared it from having to review the environmental effects at issue because any consideration could not influence the outcome of the Board’s approval process.57
While much of the initial commentary on Seven County has analyzed the economic impact of weakening NEPA’s procedural requirements,58 what the case reveals about judicial review in the post–Loper Bright era deserves just as much attention. Specifically, the Court’s underexplained decision to review the agency action as a question of discretion rather than as a question of law provides little guidance for lower courts and portends future statute-by-statute determinations of the standard of review.
Deference across judicial review in the pre–Loper Bright era made the question of what type of review a court was doing less important outside of very close cases.59 Between Chevron’s inception in 1984 and its overruling in 2024, courts reviewed agency action — whether for questions of law, fact, or discretion — with some level of deference to the agency.60 When reviewing a question of law, unless the statute was “unambiguous[],” a court would determine whether the agency action fit within “a permissible construction of the statute.”61 When reviewing a question of fact, a court would simply make sure that the agency’s factual findings were either “supported by substantial evidence in the record” for formal proceedings62 or not “arbitrary and capricious” for informal proceedings.63 Both standards were rather deferential.64 Finally, when a court was reviewing an agency’s exercise of discretion, the court would merely ensure that the agency’s choice was not “arbitrary or capricious,”65 while giving extra leeway for the agency’s “predictions, within its area of special expertise, at the frontiers of science.”66
In overruling Chevron, the Court raised the stakes for getting the standard of review correct by narrowing the circumstances under which agencies would receive deference. Loper Bright eliminated Chevron deference for agency interpretations of their organic statutes and instructed courts to “exercise their independent judgment in deciding whether an agency has acted within its statutory authority.”67 In the words of the Court, after Loper Bright, “[a]s a general matter, when an agency interprets a statute, judicial review of the agency’s interpretation is de novo.”68 The Loper Bright Court thus created a large gap between, on the one hand, the deference afforded to an agency in a review-of-discretion posture under § 706(2)(A) of the APA and, on the other, the de novo review of questions of law demanded by Loper Bright. Because of the now vastly different standards of review, the threshold question — whether the court is reviewing law or something else — takes on higher stakes in the post–Loper Bright world.
Seven County demonstrates how the choice between a review of law and a review of discretion can dramatically shape case outcomes. The D.C. Circuit conducted a searching appraisal of the Board’s EIS under arbitrary and capricious review,69 identifying three independent NEPA violations.70 Thus, the threshold question for the Supreme Court was whether the D.C. Circuit had adopted the correct level of deference to the agency.71 The Court wrote that “the meaning of ‘detailed’ [in NEPA] is a question of law to be decided by a court.”72 That logic would seem to suggest that a court would decide de novo what it means for an EIS to be “detailed.” However, as discussed above, the Court quickly pivoted away from a review of law into a more deferential review of discretion.73
The Court’s threshold bifurcation between what is required by the statutory language and what is left to “the exercise of agency discretion”74 seems like a choice unmoored from any guiding principles. It is easy to imagine the Court deciding this issue instead as a pure question of law. While the Court here assumed that the contents of the adjective “detailed” were left to the agency,75 the Court could have set those contours itself through de novo statutory interpretation. Put differently, a “detailed statement” in the NEPA context might be one that, in the eyes of a court, not just an agency, fulsomely discusses “the environmental impact of the proposed action” and “any [unavoidable] adverse environmental effects.”76 In fact, in other contexts, the Court has felt comfortable resolving ambiguities around at least equally vague language as pure questions of law.77 At the very least, there is some fuzziness over whether the meaning of “detailed” is an area of delegated discretion rather than a pure question of law. So it is notable that the Court’s discussion of that threshold bifurcation lasted only a few sentences.78 By defining the issue as a review of discretion under § 706(2)(A), the Court entitled the agency’s action to far greater deference than it would have received under a review of law.79 Unsurprisingly, once the question is whether the agency acted “within a broad zone of reasonableness,”80 the agency action — in this case, approving the railway — is likely to survive judicial review.
The Court’s approach in this case illustrates one of the problems that Chevron deference avoided: the challenge of sorting questions of law from questions of predictive judgment or delegated discretion.81 The cursory analysis that the Seven County Court offered shows only that the Court has failed to establish a consistent methodology for navigating this thicket. Moreover, given the broad, open-ended language of many statutes — language such as “detailed” or “proposed action” — demanding review of law in every instance is burdensome and pushes courts into zones of high uncertainty. Hence, the Court is already trying to selectively walk back its decision in Loper Bright.82
Yet, as Seven County demonstrates, an ad hoc approach to distinguishing questions of law from questions of discretion effectively empowers the Court to set the level of deference on a statute-by-statute, or even case-by-case, basis. Beyond approving the particular EIS before it, the Court used sweeping language — much more so than was required to dispose of the instant matter — implying that statutory interpretation in NEPA cases is never review of law. First, the Court repeatedly spoke of “NEPA cases” as a class wherein the agency receives deference.83 That categorical phrasing reaches much further than either exactly “what details need to be included in any given EIS”84 or which indirect effects count as part of the “proposed action”85 — the actual interpretations at issue — and suggests that the Seven County holding covers any statutory interpretation of NEPA. Second, the Court scolded courts that have “not applied NEPA with the level of deference demanded by the statutory text.”86 Yet the Court did not point to any statutory text that generates this “demand” across the universe of NEPA cases. Finally, the Court wrote that even if a court “might think that NEPA would support drawing a different line, a court should defer to an agency so long as the agency drew a reasonable and ‘manageable line.’”87 That edict sounds a lot like Chevron deference.88 Taken together, under this language, a future NEPA case where agency action would be reviewed de novo becomes much harder to imagine.89
Such statutory gerrymandering can create unpredictability and generate accusations of politicization.90 Lower courts will fixate on the Court’s language and try to achieve predictability in the level of deference afforded to agencies. Yet the Court can always disrupt that equilibrium with bespoke statutory interpretation, especially without a self-imposed doctrinal framework. For example, Seven County may just be a context-specific decision reflecting the reality that NEPA has never been a darling of the Court,91 such that the Court was especially willing to grant broad deference in order to pare it back. In contrast, when the Court does not trust the agency, it can demand review of law of the authorizing statute.92
A better approach would be for the Court to develop, in a common law fashion, a clear methodology for identifying which types of statutory language afford discretion to an agency and which types of language do not.93 For instance, going forward, “detailed” could trigger de novo review of law whereas “feasible” would be a question of discretion.94 The benefit of such an approach is that holdings would apply across statutes, regardless of the Court’s instincts about a particular statute or agency. Once words were classified in the “law” bucket or the “discretion” bucket, future courts would be bound to apply a uniform standard of review for the same type of statutory language. Yet this is far from the only possible approach to setting the level of deference post–Loper Bright.95 At base, any type of principled approach is preferable to an ad hoc toggling among types of judicial review. Such unpredictability is harmful to the rule of law and bodes poorly for stability in administrative law. For now, however, the Uinta Basin oil can flow faster than ever.96