Harvard Law Review Harvard Law Review Harvard Law Review

Administrative Law

Major Question Objections

The Supreme Court constantly mediates the relationship between law and administration. Of the tools available for that task, none sees more use than the ubiquitous Chevron1×1. Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984). doctrine, which establishes the conditions for judicial deference to administrative interpretations of law. In part because of its centrality to administrative law, in part because the Court regularly tinkers with the doctrine, Chevron has inspired “a virtual cottage industry within the [legal] academy”2×2. Peter M. Shane & Christopher J. Walker, Foreword, Chevron at 30: Looking Back and Looking Forward, 83 Fordham L. Rev. 475, 495 (2014). that churns out a seemingly endless stream of Chevron-related praise, criticism, and analysis.3×3. See Gary Lawson & Stephen Kam, Making Law out of Nothing at All: The Origins of the Chevron Doctrine, 65 Admin. L. Rev. 1, 2 (2013) (“[T]he scholarship on Chevron could fill a small library.”); id. n.4 (counting over 8600 Chevron-related articles as of 2012); cf. Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 1019 (2005) (Scalia, J., dissenting) (predicting the Court’s latest Chevron innovation would create “a wonderful new world . . . full of promise for administrative-law professors in need of tenure articles”). One pocket of Chevron confusion concerns the mercurial “major question exception,” under which the Court occasionally refuses to defer to an agency’s interpretation of an economically or politically significant statutory provision.4×4. See, e.g., FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 159–61 (2000); id. at 159 (citing Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 370 (1986) (“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.”)); see also Mayburg v. Sec’y of Health & Human Servs., 740 F.2d 100, 106 (1st Cir. 1984) (Breyer, J.) (expressing the same). The academic commentary has alternately denounced and defended various rationales for the exception.5×5. See, e.g., John F. Manning, The Nondelegation Doctrine as a Canon of Avoidance, 2000 Sup. Ct. Rev. 223 (criticizing the exception for upsetting legislative compromises); 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 (2008) (generally supportive of the Brown & Williamson approach); Cass R. Sunstein, Chevron Step Zero, 92 Va. L. Rev. 187, 236–42 (2006) (criticizing the major question exception cases). The Supreme Court’s latest resort to the major question exception, in King v. Burwell,6×6. 135 S. Ct. 2480 (2015). has sparked a fresh round of critical examination.7×7. See, e.g., Christine Kexel Chabot, Selling Chevron, 67 Admin. L. Rev. 481, 497–504 (2015); Abbe R. Gluck, The Supreme Court, 2014 Term — Comment: Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62, 65, 93–96 (2015); John F. Manning, How Chevron Shifts Power to the Courts (Jan. 7, 2016) (unpublished manuscript) (on file with the Harvard Law School library).

This Note enters the fray, though with a sense of hesitation: like a pickled monster, the rare and freakish major question exception may make a dubious case study.8×8. Apologies to Oliver Wendell Holmes, Sr., The Autocrat of the Breakfast Table, No. II, 2 New Eng. Mag. 134, 137 (1832) (“A natural law is not disproved by a pickled monster.”). Still, the Court continues to invoke the exception, and arguably to make it more robust, while the lower courts continue to follow the Supreme Court’s lead.9×9. For a recent and high-profile example from the lower courts, see Texas v. United States, 809 F.3d 134, 181 (5th Cir. 2015), cert. granted, 136 S. Ct. 906 (2016); id. at 217–18 (King, J., dissenting); see also Esquivel-Quintana v. Lynch, 810 F.3d 1019, 1032 (6th Cir. 2016) (Sutton, J., concurring in part and dissenting in part) (listing the “extraordinary case[]” exception alongside other routine Chevron exceptions (quoting King, 135 S. Ct. at 2488)). Perhaps now is a good time to reappraise the major question exception, while it is still a minor excrescence on administrative law. In the hope of producing some order, this Note argues as follows. First, it proposes to abandon the fruitless quest to rationalize the disorderly major question cases in terms of conventional doctrine, and suggests it might be better to regard them as episodes of vaguely equitable intervention, where the Court’s “olfactory sense detects the odor of administrative waywardness.”10×10. Colin S. Diver, Statutory Interpretation in the Administrative State, 133 U. Pa. L. Rev. 549, 598 (1985) (suggesting the Court’s olfactory sense may not be a reliable guide). It then argues that the Court’s functionally equitable mode in major question cases conflicts with its own understanding of Chevron doctrine. The Court has invoked the major question exception both while conducting a pre-Chevron threshold inquiry and while interpreting statutes within the Chevron framework. The former confuses proxies for congressional delegation with proxies for agency opportunism; the latter short-circuits arbitrary and capricious review under the Administrative Procedure Act11×11. Pub. L. No. 79-404, 60 Stat. 237 (codified as amended in scattered sections of 5 U.S.C.). (APA).

This Note proceeds in three Parts. Part I briefly reviews the elements of Chevron. Part II uses five cases to introduce the major question exception and to illustrate its haphazard application and elusive rationale(s). Part III begins by arguing that no coherent principle animates the major question cases, which are better understood as equitable-mode cases that aim to prevent agencies from taking undue advantage of Chevron’s simple deference rule. Section III.A then argues that deploying the major question exception as part of a pre-Chevron threshold inquiry confuses proxies for delegation with proxies for agency opportunism, while section III.B argues that deploying it while interpreting statutes under the Chevron framework effectively curtails arbitrary and capricious review.

I. Background: Chevron

Suppose the Food and Drug Administration (FDA) issues a rule intended to implement some portion of its organic statute, the Food, Drug, and Cosmetic Act12×12. 21 U.S.C. §§ 301–399d (2012). (FDCA). Suppose further that a regulated party sues in court to invalidate the FDA’s rule, claiming that the rule is based on an impermissible interpretation of the FDCA. Before 1984, a baroque body of case law determined what weight to accord the FDA’s interpretation.13×13. See Stephen G. Breyer et al., Administrative Law and Regulatory Policy 273–82 (7th ed. 2011). The Supreme Court’s 1984 decision in Chevron streamlined the inquiry. Formally, Chevron instructs courts to engage in a two-step process, asking first whether Congress has “directly spoken” to the interpretive question — that is, whether the statutory language forecloses the agency’s interpretation — and second, in the face of ambiguity, whether the agency’s interpretation is reasonable.14×14. Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984). Functionally, Chevron simply tells the court to defer to the FDA’s interpretation so long as it is reasonable.15×15. Put another way, “[t]he court need not conclude that the agency construction was the only one it permissibly could have adopted to uphold the construction, or even the reading the court would have reached if the question initially had arisen in a judicial proceeding.” Id. at 843 n.11. Instead, the court’s task is to “identify[] the statute’s range of reasonable interpretations” and determine whether the agency’s interpretation falls within that range. Matthew C. Stephenson & Adrian Vermeule, Chevron Has Only One Step, 95 Va. L. Rev. 597, 602 (2009); see also Holder v. Martinez Gutierrez, 132 S. Ct. 2011, 2017 (2012); Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 218 & n.4 (2009).

Chevron rests on the premise that when Congress delegates a statutory provision to an agency to implement, it intends for the agency to exercise primary interpretive authority over the provision.16×16. Chevron, 467 U.S. at 843–44; see also City of Arlington v. FCC, 133 S. Ct. 1863, 1868 (2013); Breyer, supra note 4, at 369–70; Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 516–17. For a summary of alternative justifications for Chevron, see Evan J. Criddle, Chevron’s Consensus, 88 B.U. L. Rev. 1271, 1275–98 (2008); see also United States v. Mead Corp., 533 U.S. 218, 241–43 (2001) (Scalia, J., dissenting) (likening Chevron deference to the rules governing mandamus). That premise in turn rests on the idea that interpretation inevitably entails policy choices, which Congress thinks agencies, with their comparative advantages of subject-matter expertise and political accountability, are better situated than courts to make.17×17. See Chevron, 467 U.S. at 843–45. Deference, that is, presupposes delegation. The Court has long agreed on that much; but the Justices do not all agree on how to determine whether Congress has delegated a statutory provision to an agency.18×18. For a concise but thorough discussion, see John F. Manning, Inside Congress’s Mind, 115 Colum. L. Rev. 1911, 1929–34 (2015). An agency’s organic statute might include, for example, (i) broad standards that invite policy choices, accompanied by rulemaking or adjudicative authority; (ii) terms or phrases that are not self-defining or whose statutory definitions leave room for interpretive discretion; or (iii) provisions that do not fit well together or that pull in different directions, perhaps because of sloppy drafting. From first to last, it is most to least likely that Congress made an intentional choice about which branch, executive or judiciary, should resolve the statutory indeterminacy.19×19. Bracket the question whether Congress can have a (discoverable) collective intent in the first place. See, e.g., Max Radin, Statutory Interpretation, 43 Harv. L. Rev. 863, 870–72 (1930). After the first or possibly second on the list, it is doubtful that Congress even knew indeterminacy existed, let alone that it had an intention about who should resolve it.20×20. Cf. Scalia, supra note 16, at 517. The Court generally seems to agree that it cannot determine whether Congress actually intended to delegate interpretive authority to an agency in any given provision.21×21. See Manning, supra note 18, at 1932–34. Admittedly, though, the Justices may not hold this view with perfect consistency. See, e.g., Christensen v. Harris County, 529 U.S. 576, 597 (2000) (Breyer, J., dissenting) (suggesting that Chevron deference depends on whether “Congress actually intended to delegate interpretive authority to the agency”). So Chevron requires courts to impute to Congress a fictional intent to delegate.22×22. See Breyer, supra note 4, at 370; Manning, supra note 18, at 1932–34; Scalia, supra note 16, at 517. The disagreement arises over how to construct that fictional intent. On one view, associated most closely with Justice Scalia, courts should pick “a stable background rule against which Congress can legislate,” and so should presume that whenever Congress leaves statutory provisions ambiguous, it means for administering agencies to resolve any indeterminacies within the bounds of reasonable interpretation.23×23. City of Arlington v. FCC, 133 S. Ct. 1863, 1868 (2013) (Scalia, J.); see also Scalia, supra note 16, at 514–17; Sunstein, supra note 5, at 203 (describing Justice Scalia’s views). In other words, ambiguity alone should signal delegation. On another view, associated most closely with Justice Breyer, courts should engage in case-by-case inquiry into Congress’s hypothetical intentions, in order to “allocate the law-interpreting function between court and agency in a way likely to work best within any particular statutory scheme.”24×24. Breyer, supra note 4, at 371; see also Sunstein, supra note 5, at 199–200 (describing Justice Breyer’s views). If a provision is ambiguous, a court’s job is to decide whether, all things considered, Congress would more likely want an agency or a court to resolve the ambiguity.

Reasonable minds can disagree about how best to construct fictional congressional intent to delegate in Chevron cases; in part, the decision reduces to “the pervasive choice between standards and rules.”25×25. Sunstein, supra note 5, at 205; see also id. (“If the choice between rules and standards turns in part on the costs of error and the costs of decisions, then [Justices Breyer and Scalia] might be seen as disagreeing about exactly how to assess those costs.” (footnote omitted)). But Justice Scalia was surely right about an important and related point: the question whether Congress has delegated interpretive authority to an agency cannot be separated neatly from the question whether a court should defer to an agency’s interpretation in a given case.26×26. See City of Arlington, 133 S. Ct. at 1871 (Scalia, J.) (“In practice, it does not appear to matter whether delegated authority is viewed as a threshold inquiry.” (quoting Harry T. Edwards & Linda A. Elliott, Federal Standards of Review 146 (2007))); see also id. at 1868–71. The more robust the antecedent inquiry into whether Congress (fictionally) intended to delegate a question to an agency — Justice Breyer would consider a host of factors27×27. See id. at 1875 (Breyer, J., concurring in part and concurring in the judgment) (“[Courts should assess] the interstitial nature of the legal question, the related expertise of the Agency, the importance of the question to administration of the statute, the complexity of that administration, and the careful consideration the Agency has given the question over a long period of time.” (quoting Barnhart v. Walton, 535 U.S. 212, 222 (2002))). — the less room for meaningful inquiry into the reasonableness of the agency’s interpretation. And there is something odd about a robust, Chevron-antecedent delegation inquiry; Chevron, after all, ultimately aims to empower expert and politically accountable agencies, which seems to entail judicial restraint.28×28. See Gluck, supra note 7, at 65 (“Chevron aim[s] to minimize the role of courts in a landscape dominated by statutory law.”).

But the Court has not been able to resist the impulse to split the question whether Congress intended to delegate interpretive authority from the question whether courts should defer to an agency’s interpretation of a statutory provision in a given case. United States v. Mead Corp.29×29. 533 U.S. 218 (2001). introduced a Chevron-antecedent threshold inquiry to determine what kinds of agency interpretations should trigger the deferential Chevron framework — that is, to decide what falls under Chevron’s “domain.”30×30. Thomas W. Merrill & Kristin E. Hickman, Chevron’s Domain, 89 Geo. L.J. 833 (2001). The account here is simplified. For more detail on the case law leading up to Mead, see, for example, Thomas W. Merrill, The Mead Doctrine: Rules and Standards, Meta-Rules and Meta-Standards, 54 Admin. L. Rev. 807, 809–11 (2002). Under Mead, Chevron applies “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law” — as through a grant of rulemaking or adjudicative authority — “and that the agency interpretation claiming deference was promulgated in the exercise of that authority.”31×31. Mead, 533 U.S. at 226–27. Mead is no masterpiece of clarity, and it left open the possibility that “some other indication of comparable congressional intent” besides a general grant of rulemaking or adjudicative authority might signal a delegation. Id. at 227. But the case has generally been understood to mean that grants of rulemaking or adjudicative authority satisfy Mead’s threshold inquiry. See City of Arlington, 133 S. Ct. at 1874. Importantly, “Mead’s analysis of delegation is entirely procedural, not substantive.”32×32. Adrian Vermeule, Introduction: Mead in the Trenches, 71 Geo. Wash. L. Rev. 347, 355 (2003). It rests on the idea that Congress should be understood to delegate interpretive authority to agencies “when it provides for a relatively formal administrative procedure tending to foster . . . fairness and deliberation.”33×33. Mead, 533 U.S. at 230. Mead amounts to an intermediate approach to constructing fictional congressional intent: ambiguity alone is not sufficient to signal intent to delegate, but the focus on procedural indicators keeps the test from being all-things-considered. It also allows Congress a predictable way to signal its intent (by specifying procedures), and agencies a way to take advantage of delegations if they want judicial deference (by using the specified procedures).

In sum: Chevron presumes that when Congress delegates a statutory provision to an agency to implement, it intends for the expert and politically accountable agency to exercise primary interpretive authority over the provision and for courts to defer to the agency’s reasonable resolution of any statutory indeterminacies. Deference presupposes delegation; but some disagreement exists over how best to construct (fictional) congressional intent to delegate, and a robust inquiry into what constitutes a delegation can shade into the question whether to defer to a given interpretation. Mead mediates between the rule-like (Scalian) and all-things-considered (Breyeresque) approaches to identifying delegations.

II. The Major Question Exception: Five Cases

The foregoing account describes Chevron’s operation in the vast run of administrative law cases in the federal courts. But an odd series of cases has occasionally punctuated the Supreme Court’s otherwise unremarkable Chevron docket. These cases seem to establish a “major question” exception to Chevron deference: the Court will not defer to an agency’s interpretation of certain “economically and politically significant” statutory provisions, even if all the other preconditions for Chevron deference obtain — even if, that is, Congress gave the agency a general grant of rulemaking or adjudicative authority, and the agency’s interpretation of the contested “major” provision emerged from the exercise of that authority and at least facially accords with the statute’s text. Perplexingly, the major question cases coexist with certain other cases that undercut most conceivable rationales for the major question exception. The following discussion presents five illustrative cases: three major question cases and two “anti-cases.” The aim of the discussion is twofold: to recount the devel-opment of the major question exception over time and to show that the exception has never been justified by any coherent rationale.

The major question exception made a prominent early appearance in FDA v. Brown & Williamson Tobacco Corp.,34×34. 529 U.S. 120 (2000). which invalidated a set of FDA rules that regulated tobacco products, for the first time, as “restricted devices” under the FDCA.35×35. Id. at 129, 146; see also Regulations Restricting the Sale and Distribution of Cigarettes and Smokeless Tobacco to Protect Children and Adolescents, 61 Fed. Reg. 44,396, 44,400, 44,403–07 (Aug. 28, 1996) (discussing the legal authority that the FDA cited for the rule invalidated by Brown & Williamson); Regulations Restricting the Sale and Distribution of Cigarettes and Smokeless Tobacco Products to Protect Children and Adolescents, 60 Fed. Reg. 41,314, 41,349–50 (proposed Aug. 11, 1995) (same). Although the statute’s literal language seemed to encompass tobacco products,36×36. See Brown & Williamson, 529 U.S. at 162 (Breyer, J., dissenting); Jody Freeman & David B. Spence, Old Statutes, New Problems, 163 U. Pa. L. Rev. 1, 74 (2014); Manning, supra note 5, at 233. the Court found the FDA’s interpretation impermissible, in part because it would create unacceptable tension with other parts of the statute,37×37. See Brown & Williamson, 529 U.S. at 133–43. But see id. at 174–81 (Breyer, J., dissenting) (disputing tension). in part because other tobacco-related legislation implicitly or expressly “preempted” the agency’s regulations.38×38. See id. at 143–59 (majority opinion). But see Manning, supra note 5, at 226 (“[The Court’s] heavy reliance on postenactment legislative history . . . seems out of character.”). Then the Court added a third ground: Congress simply could not have intended to leave such an important question, concerning the basic reach of the FDCA and its application to the vast tobacco industry, to an administrative agency.39×39. Brown & Williamson, 529 U.S. at 159–61. Taking account of “the nature of the question presented,” the Court seemingly weighted the scale against Chevron deference in “extraordinary cases”40×40. Id. at 159 (emphasis added). where agencies interpret statutory provisions of “economic and political magnitude.”41×41. Id. at 133. The authority to which the Court appealed for this concept was (ironically, since he dissented) Justice Breyer’s own earlier article, cited supra note 4. See Brown & Williamson, 529 U.S. at 159. Altogether, the Court concluded that Congress had “directly spoken to the question at issue and precluded the FDA from regulating tobacco products,”42×42. Id. at 160–61. unambiguously foreclosing the agency’s interpretation.

Brown & Williamson vexed commentators, who explored and criticized possible rationales for the fledgling major question exception. The simplest: reasonable legislators would not want agencies to make especially significant policy decisions, either because a different institution is better placed to make such decisions or because of the danger of agency self-aggrandizement.43×43. See Sunstein, supra note 5, at 232–33. A slightly more elaborate version of the “just too big” rationale is that Congress would not intend for agencies to decide on the scope of their “jurisdiction,” or the bounds of their own authority — which they seem to be doing when they decide questions about a statute’s basic reach. See id. at 234–35. For the Court’s later disposal of this particular worry, see the discussion of City of Arlington, infra text accompanying notes 58–60. But the “just too big” rationale is incoherent. Agencies decide “big” questions all the time (Chevron itself “involved a significant rethinking” of a previous interpretation, with large consequences44×44. Sunstein, supra note 5, at 232. ), and more importantly, “majorness” does not correlate with ordinary reasons for denying an agency deference. With major questions, agency expertise and political accountability “might well be more relevant, not less”; “there is no reason to think that Congress . . . seek[s] a judicial rather than administrative judgment.”45×45. Id. at 233. Justice Breyer argued similarly in dissent in Brown & Williamson. See Brown & Williamson, 529 U.S. at 190–91 (Breyer, J., dissenting). Alternatively, and more seriously, perhaps major questions present nondelegation concerns by arguably transferring legislative power to the executive branch. On this view, the Brown & Williamson Court was warning Congress that it would “aggressively narrow[]” statutes that might be read to give agencies sweeping powers over an industry or the economy.46×46. Manning, supra note 5, at 237; see also Sunstein, supra note 5, at 244–45. Note that this kind of nondelegation concern is slightly different from the more familiar one, which emphasizes the need for discretion-constraining “intelligible principles”; this kind simply aims to avoid conferring too much power overall, it seems. Manning, supra note 5, at 227 n.26. Leaving aside the questionable legitimacy of the nondelegation doctrine itself,47×47. For the argument that there is no constitutional or theoretical warrant for the nondelegation doctrine, see generally Eric A. Posner & Adrian Vermeule, Interring the Nondelegation Doctrine, 69 U. Chi. L. Rev. 1721 (2002). this rationale seems self-defeating: “If the point of the nondelegation doctrine is to ensure that Congress makes important statutory policy, a strategy that requires the judiciary, in effect, to rewrite the terms of a duly enacted statute cannot be said to serve the interests of that doctrine.”48×48. Manning, supra note 5, at 228; see also id. at 256–57.

The major question exception made a puzzling reappearance in an “anti-case,” Massachusetts v. EPA,49×49. 549 U.S. 497 (2007). which held that the Clean Air Act50×50. 42 U.S.C. §§ 7401–7671q (2012). authorized the Environmental Protection Agency (EPA) to regulate greenhouse gas emissions from new motor vehicles.51×51. Massachusetts, 549 U.S. at 532. EPA had denied a rulemaking petition after concluding that greenhouse gases were not “air pollutants” within the Act’s meaning.52×52. Id. at 510–13. The agency was “urged on in this view” by the decision in Brown & Williamson: using a broadly worded statute to regulate greenhouse gas emissions would have even “greater economic and political implications than FDA’s attempt to regulate tobacco.”53×53. Control of Emissions from New Highway Vehicles and Engines, 68 Fed. Reg. 52,922, 52,928 (Sept. 8, 2003). The Court replied that “EPA’s reliance on Brown & Williamson . . . [was] misplaced,”54×54. Massachusetts, 549 U.S. at 530. and strained to distinguish the structure and history of the FDCA from that of the Clean Air Act.55×55. Id. at 530–31. The Court essentially ignored the “majorness” parallel between the two cases, however, prompting observers to claim that it had “unceremoniously killed” the major question exception.56×56. See, e.g., Moncrieff, supra note 5, at 598. In a way, it did something subtler. By willingly finding a broad grant of administrative authority over a large swath of the economy, the Court seemingly eliminated the nondelegation and anti-aggrandizement rationales for the major question exception; but by rejecting the agency’s (self-restraining!) answer to a “big” question, the Court arguably invigorated a form of the “just too big” rationale — the idea that reasonable legislators would not want agencies to make especially significant policy decisions one way or the other.57×57. Cf. Jody Freeman & Adrian Vermeule, Massachusetts v. EPA: From Politics to Expertise, 2007 Sup. Ct. Rev. 51, 76–77 (arguing similarly); David Baake, Obituary: Chevron’s “Major Questions Exception,” Harv. Envtl. L. Rev. Blog (Aug. 27, 2013, 5:43 PM), http://harvardelr.com/2013/08/27/obituary-chevrons-major-questions-exception [http://perma.cc/P65H-TDDM] (arguing similarly). For the view that the Court in Massachusetts v. EPA was simply “engaged in a different enterprise” from that of the Brown & Williamson Court — namely, that it was engaged in a kind of “expertise-forcing” — see Freeman & Vermeule, supra, at 77–92. Still, the Court’s rejection of Brown & Williamson’s concerns about agency overreach at best converted the major question exception into a wholly new principle.

Although it did not address the major question exception head-on, another “anti-case” somewhat eroded the very concept of a special category of “big questions.” City of Arlington v. FCC58×58. 133 S. Ct. 1863 (2013). confirmed that the Chevron framework applies to statutory provisions that define the scope of an agency’s authority (its “jurisdiction”). Writing for the majority, Justice Scalia rejected the claim that Chevron should not apply, which he insisted rested on a false premise: “that there exist two distinct classes of agency interpretations,” “big, important ones” that “define the agency’s ‘jurisdiction,’” and “humdrum, run-of-the-mill stuff” that “simply [apply] jurisdiction the agency plainly has.”59×59. Id. at 1868. That purported distinction is a “mirage,” he explained; “[n]o matter how it is framed, the question a court faces when confronted with an agency’s interpretation of a statute it administers is always, simply, whether the agency has stayed within the bounds of its statutory authority.”60×60. Id. To be clear, City of Arlington renounced only the claimed distinction between jurisdictional and nonjurisdictional provisions in organic statutes, but similar logic might dispute the difference between “major” and “minor” questions.

Yet the major question exception made a comeback in Utility Air Regulatory Group v. EPA61×61. 134 S. Ct. 2427 (2014). (UARG), the progeny of Massachusetts v. EPA. EPA determined that the Clean Air Act, as interpreted in Massachusetts v. EPA, required the agency to regulate greenhouse gas emissions from stationary sources as well as moving sources. To simplify somewhat, regulating emissions from stationary sources would also abruptly increase the number of sources subject to the statute’s licensing requirements, in such a way that the licensing program would become unworkable. To fix that problem, EPA “tailor[ed]”62×62. Id. at 2437 (quoting Prevention of Significant Deterioration and Title V Greenhouse Gas Tailoring Rule, 75 Fed. Reg. 31,514 (June 3, 2010) (codified as amended in scattered parts of 40 C.F.R.)). the rather explicit requirements of the licensing program so that it covered only certain kinds of stationary sources.63×63. See id. at 2436–38. The Court would have none of that. Massachusetts v. EPA, the Court explained, merely clarified that “air pollutants” could include greenhouse gas emissions, but EPA was free to give the phrase a “narrower, context-appropriate meaning” where it occurred in specific operative provisions.64×64. Id. at 2439; see also id. at 2439–41. Here, moreover, EPA must give the phrase a narrower meaning to avoid rendering the licensing program unworkable.65×65. Id. at 2442. But that is not all, said the Court. “EPA’s interpretation is also unreasonable because it would bring about an enormous and transformative expansion in EPA’s regulatory authority without clear congressional authorization.”66×66. Id. at 2444. Citing Brown & Williamson, the Court insisted that “[t]he power to require permits for the construction and modification of tens of thousands, and the operation of millions, of small sources nationwide falls comfortably within the class of authorizations that we have been reluctant to read into ambiguous statutory text.”67×67. Id. This sounds a good deal like a return to the nondelegation or anti-aggrandizement versions of the major question exception, the versions that Massachusetts v. EPA supposedly zapped.

The most recent major question case is King v. Burwell.68×68. 135 S. Ct. 2480 (2015). King sustained an Internal Revenue Service (IRS) regulation interpreting a key provision of the Patient Protection and Affordable Care Act69×69. Pub. L. No. 111-148, 24 Stat. 119 (2010) (codified as amended in scattered sections of the U.S. Code). (ACA). Apocalyptic commentary notwithstanding, the case presented the Court with a quotidian Chevron question: was the IRS’s interpretation reasonable? The ACA, as codified at § 36B of the Internal Revenue Code, authorizes tax credits for certain health insurance plans “enrolled in through an Exchange established by the State under [section] 1311 of the [ACA].”70×70. 26 U.S.C. § 36B(b)(2)(A) (2012); id. § 36B(c)(2)(A)(i). Based on statutory cross-references and other contextual signals, the IRS issued a regulation interpreting § 36B to allow tax credits for plans purchased on either state- or federally- created exchanges.71×71. King, 135 S. Ct. at 2487; Health Insurance Premium Tax Credit, 77 Fed. Reg. 30,377, 30,378 (May 23, 2012) (codified as amended at 26 C.F.R. pt. 1 (2015)). The lower courts split on the interpretive question, but all analyzed it under Chevron.72×72. See Halbig v. Burwell, 758 F.3d 390, 398 (D.C. Cir. 2014); King v. Burwell, 759 F.3d 358, 367 (4th Cir. 2014); Oklahoma ex rel. Pruitt v. Burwell, 51 F. Supp. 3d 1080, 1088 (E.D. Okla. 2014); King v. Sebelius, 997 F. Supp. 2d 415, 426 (E.D. Va. 2014); Halbig v. Sebelius, 27 F. Supp. 3d 1, 16 (D.D.C. 2014). The Supreme Court, however, summoned Brown & Williamson and skirted Chevron altogether: this was an “extraordinary case[],”73×73. King, 135 S. Ct. at 2488 (quoting FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 159 (2000)). implicating “a question of deep ‘economic and political significance,’” a case that resisted the usual Chevron framework — so the Court took the question for itself.74×74. Id. at 2489 (quoting UARG, 134 S. Ct. 2427, 2444 (2014)). The majority found the contested provision ambiguous and in the end settled on the same interpretation as the IRS; but it did so independently, without deferring to the agency or (ostensibly) giving the agency’s views any particular weight.75×75. See id. at 2492–96.

King’s casual invocation of the major question exception glossed over several curiosities. First, the Court unequivocally treated the exception as a threshold matter: the “economic and political significance” of the question meant that the Chevron framework did not apply at all,76×76. See id. at 2488–89 (quoting UARG, 134 S. Ct. at 2444). even though the IRS’s regulation met Mead’s procedural requirements.77×77. Congress gave the IRS both general and specific grants of rulemaking authority, see 26 U.S.C. §§ 36B(g), 7805(a) (2012), and the IRS promulgated its regulation through notice-and-comment rulemaking, see Health Insurance Premium Tax Credit, 77 Fed. Reg. 30,377 (May 23, 2012) (codified as amended at 26 C.F.R. pt. 1 (2015)). Second, the Court had never before agreed with an agency’s interpretation of a “big” question. The major question exception had always reinforced various other reasons for rebuffing an agency, as a sort of tiebreaker in close cases. Third, the Court did not face a new and unexpected expansion of agency power, which seemed to be a latent concern in Brown & Williamson and UARG. Fourth and finally, besides the “economic and political significance” of the question and the centrality of the tax credits to the statutory scheme, the Court added another reason why the Chevron framework should not apply: the ACA is a health insurance statute, and the IRS, a tax agency, has “no expertise in crafting health insurance policy of this sort.”78×78. King, 135 S. Ct. at 2489. The Court ignored the fact that the IRS deliberately took its interpretation from the Department of Health and Human Services, which does have health-policy expertise. See Health Insurance Premium Tax Credit, 77 Fed. Reg. at 30,378; 26 C.F.R. § 1.36B-1(k) (2012) (incorporating HHS regulation). The Court also ignored the fact that the IRS regularly helps administer parts of statutes, like the ACA, that rely on the tax system to implement policy goals beyond raising revenue. See Kristin E. Hickman, The (Perhaps) Unintended Consequences of King v. Burwell, 2015 Pepp. L. Rev. 56, 66–70. The Court did not specify whether lack of expertise alone might be decisive, or whether it simply complemented “majorness” in this case.

The “major question” category is somewhat fluid, and other cases could be included here.79×79. Cass Sunstein, for example, counts Babbitt v. Sweet Home Chapter of Communities for a Great Oregon, 515 U.S. 687 (1995), and MCI Telecommunications Corp. v. AT&T Co., 512 U.S. 218 (1994), as major question cases. See Sunstein, supra note 5, at 193 & n.30, 236–40. Gonzales v. Oregon, 546 U.S. 243 (2006), might be another candidate. See id. at 267. But these cases sufficiently illustrate the muddled state of things. Neither the purpose of nor the doctrinal rationale for the major question exception has ever been made explicit, and both seem to shift from case to case. Functionally, the exception may operate to prevent an agency from interpreting the text of its organic statute in a newly expansive way, but King did not involve a changed interpretation, and Massachusetts v. EPA forced a newly expansive interpretation. The exception usually operates against the agency — the agency usually loses — but not in King, which sustained the IRS’s regulation. Formally, the exception has had nondelegation overtones, but only sometimes; and anyhow, enforcing nondelegation by “aggressively narrowing” broad statutory grants is a questionable practice.80×80. See supra text accompanying note 48. Alternatively, the exception has sought to route especially “big” or consequential questions away from agencies and to courts, apparently on the theory that Congress would prefer judicial to administrative resolution of such questions. But the Court has never explained why agencies’ comparative advantages — subject-matter expertise and political accountability — shouldn’t be brought to bear all the more when the interpretive question is a “major” one. Nor has it offered a positive account for why courts should answer big questions. As it has developed so far, the protean major question exception has an air of judicial improvisation.

III. The Major Question Exception: An Equitable Intervention

The major question exception defies doctrinal justification, and it is tempting to dismiss it as an inexplicable, but fortunately rare, judicial wild card. Perhaps, however, it can be explained motivationally. Rather than attempt to press the disorderly cases into the mold of something like the nondelegation doctrine, perhaps it is better to characterize the major question cases as occasions where the Court thought it necessary to engage in vaguely equitable intervention. To put it in a favorable light, the major question exception might function like the anti-abuse principle in tax: as a standard that complements a simple rule by preventing its opportunistic use to produce unexpected and disruptive results.81×81. See David A. Weisbach, Formalism in the Tax Law, 66 U. Chi. L. Rev. 860, 860 (1999). Just as anti-abuse standards license courts “to override the literal words of a statute or regulation” in order to thwart transactions that “literally compl[y] with” but also manipulate the rules to achieve unexpected tax liability outcomes,82×82. Id. so the major question exception seems to give courts the equitable power to prevent “violations of [an] implicit pact”83×83. Id. at 880. — that agencies will not take advantage of Chevron’s simple pro-agency deference rule in order to advance interpretations that are somehow misaligned with general expectations about how a given statutory scheme should work. The strongest argument for a limited Chevron anti-abuse exception is that it avoids the need for a more complex set of deference rules to capture uncommon cases, and so it allows Chevron to operate as a simple rule most of the time.84×84. Cf. id. at 876 (“The analysis leaves us between the Scylla of certain but unduly complex rules and the Charybdis of uncertain but less complex standards. The goal of anti-abuse rules is to find a way out.”).

This understanding of the major question exception jibes with the Court’s apparent concerns in the cases examined. In Brown & Williamson, the Court worried that the FDA might upset a kind of equilibrium among Congress, administration, and the industry over tobacco regulation.85×85. See FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 137–39, 143–59 (2000); Moncrieff, supra note 5, at 623–24. In the Court’s view, the FDA took advantage of Chevron’s simple rule — agencies resolve statutory ambiguities — to violate a tacit understanding that tobacco, with its “unique political history,” was outside the statute’s ambit, even if within its literal bounds.86×86. Brown & Williamson, 529 U.S. at 159–60. In Massachusetts v. EPA, with circumstances strikingly similar to those in Brown & Williamson except for the agency’s refusal to regulate, EPA arguably took advantage of the Court’s own anti-abuse principle in order to mask a partisan political decision. UARG might have been a Brown & Williamson redux, or the Court’s correction of its own overreach in Massachusetts v. EPA, or both. And King might have been a preemptive strike to stop the IRS from later taking advantage of what the Court viewed as a sloppy drafting error to change its interpretation in a way that would threaten the ACA’s operation.87×87. For the suggestion that the King Court recognized but avoided directly confronting the possibility of a drafting error, see Gluck, supra note 7, at 76–79, 96, 100–05.

If it is right to understand the major question exception as a kind of equitable intervention — as a sort of Chevron safety valve in cases where Chevron would normally require judicial deference, but the Court thinks an agency has taken (or may take) undue advantage of the simple deference rule — then several questions follow. For the sake of simplicity, the argument here sets two particular questions aside: whether it is possible for courts to discern how a statutory scheme “should” work, and whether it is realistic to think courts can accurately identify instances of agency opportunism. Even assuming affirmative answers to those questions, there is a further question about internal consistency: does the major question exception as an anti-abuse principle fit with the rest of administrative law? This Part argues that it does not. Administrative law does have roots in the federal courts’ equity jurisdiction,88×88. John F. Duffy, Administrative Common Law in Judicial Review, 77 Tex. L. Rev. 113, 118 (1998) (“Judicial review in the early administrative era grew up in the federal equity jurisdiction, which was filled with judge-made law fashioned within the tradition of equity.”); id. at 121–30; Henry E. Smith, Equity and Administrative Behavior, in Equity and Administration 326, 328, 343–356 (P.G. Turner ed., 2016). and the extent to which the APA displaced the courts’ more general equitable powers is open to dispute.89×89. Compare Duffy, supra note 88, at 119, 181–211 (arguing that after the APA, “the courts’ method of analysis should have changed: Statutory law should have assumed the dominant position in cases covered by the APA (which means just about all cases reviewing federal administrative action),” id. at 119), with Smith, supra note 88, at 353 (arguing that the APA did not displace all equity jurisdiction, and that the APA itself might incorporate equitable principles). But the Court’s functionally equitable mode in major question cases clashes with its own understanding of Chevron and administrative law more generally. The Court has now called on the major question exception both as part of a pre-Chevron threshold inquiry and while interpreting statutes within the Chevron framework. Section III.A argues that incorporating the major question exception into the pre-Chevron threshold inquiry confuses proxies for congressional delegation with proxies for agency opportunism. Section III.B then argues that invoking the major question exception while interpreting statutes within the Chevron framework effectively short-circuits arbitrary and capricious review under § 706(2)(A) of the APA.

A. Major Questions at the Threshold

King v. Burwell is the latest word on the major question exception. Read for all it is worth, the case marks a notable shift: the Court may now consider “majorness” as a threshold matter when deciding whether the Chevron framework applies in a given case. The significance of this shift is debatable, and evaluating it requires returning to the earlier discussion about the putative distinction between two questions: whether Congress has delegated a statutory provision to an agency, and whether a court should defer to an agency’s interpretation of a given provision.90×90. See supra text accompanying notes 26–28. The two questions readily collapse into one another. But the claim here is that to the extent they can be distinguished, as the Court tries to distinguish them, it makes no sense to incorporate the major question exception into the pre-Chevron threshold inquiry.

To illustrate the collapsibility of the two inquiries before King, consider Brown & Williamson again. Justice Breyer, who dissented in that case, would have deferred to the FDA’s interpretation, not as a matter of course, but because he thought, all things considered, that Congress would prefer the FDA to resolve that particular big question (in part because it was well suited to resolution by a politically accountable actor).91×91. Brown & Williamson, 529 U.S. at 190–91 (Breyer, J., dissenting). The majority opinion, which Justice Scalia joined, refused to defer to the agency because it found the statute unambiguous, in part because the Court thought it unlikely that Congress would assign the FDA such a big question.92×92. See supra text accompanying notes 36–42. If the line between the two approaches seems thin, it is. Both amount to judicial conjectures about congressional intent to delegate; one bases the decision whether to defer directly on that conjecture, while the other feeds the conjecture into the determination about ambiguity.93×93. Cf. Christensen v. Harris County, 529 U.S. 576, 590 n.* (2000) (Scalia, J., concurring in part and concurring in the judgment) (“The implausibility of Congress’s leaving a highly significant issue unaddressed (and thus ‘delegating’ its resolution to the administering agency) is assuredly one of the factors to be considered in determining whether there is ambiguity . . . .”). All the same, at least formally, Brown & Williamson was a Chevron case; it made no attempt to inquire into “majorness” as a pre-Chevron threshold matter, and the economic and political significance of the question was just one consideration among several in the decision not to defer to the FDA.94×94. See Brown & Williamson, 529 U.S. at 132–33; see also City of Arlington v. FCC, 133 S. Ct. 1863, 1872 (2013) (describing Brown & Williamson as a case decided within the Chevron framework).

Now return to King. There are two ways of understanding the Court’s move in that case. One possibility is that King marks a complete (and sudden) shift to an all-things-considered version of Mead’s pre-Chevron delegation inquiry.95×95. See City of Arlington, 133 S. Ct. at 1875 (Breyer, J., concurring in part and concurring in the judgment) (quoting Barnhart v. Walton, 535 U.S. 212, 222 (2002)). King’s grab bag of reasons for hesitating to conclude that Congress “intended . . . an implicit delegation”96×96. King v. Burwell, 135 S. Ct. 2480, 2489 (2015). to the IRS — “economic and political significance,” centrality of the tax credits to the ACA, the amount of money at stake, purported lack of agency expertise97×97. Id. at 2488–89. — might support this view. If so, there is little more to say. Under the all-things-considered version of Mead, the delegation inquiry essentially swallows the deference inquiry. Courts have complete discretion to decide whether Congress has delegated a provision to an agency, and can therefore reach the deference question only when they want to — a bit like peeking at the merits in standing cases.98×98. William A. Fletcher, The Structure of Standing, 98 Yale L.J. 221 (1988) (showing how jurisdictional and merits questions collapse); see also Transcript of Oral Argument at 7, Am. Elec. Power Co. v. Connecticut, 131 S. Ct. 2527 (2011) (No. 10-124) (Kennedy, J.) (“[W]e all know that you sometimes have to peek at the merits to see if there’s standing. There’s a little cheating that goes on.”). Put another way, this view amounts to saying, “Congress ‘intends’ courts to decide whether Congress ‘intends’ agencies to decide” any given question. The second part is essentially meaningless; it effectively amounts to a judicial promise to defer whenever courts feel like deferring. Cf. Manning, supra note 7.

Alternatively, and somewhat more modestly, the King Court was engaged in the same kind of equitable intervention as in earlier major question cases, one that it saves for “extraordinary” circumstances where it thinks the agency has taken (or might in the future take) undue advantage of Chevron’s simple deference rule. If so, King’s innovation was to decide that the economic and political significance of a question can remove it from Chevron’s framework altogether, rather than being one factor within the Chevron framework that informs the Court’s evaluation of an agency’s interpretation. Assume arguendo that King’s innovation is meaningful — assume, that is, that the pre-Chevron threshold inquiry is sufficiently distinct from the question whether the agency’s interpretation is reasonable within the Chevron framework, so that the stage at which the Court considers “majorness” can affect the outcome of the case. King’s incorporation of the major question exception into the pre-Chevron threshold inquiry not only clouds the conventional view of the threshold inquiry as a device for identifying congressional delegations, but also fails to pick out a domain where equitable intervention might make sense.

To see why, it helps to understand the Court’s use of “majorness” as a kind of proxy. Mead’s threshold inquiry already relies on procedural proxies — statutory authorization to promulgate rules or to engage in adjudication is taken as a proxy for delegation. Courts also rely on proxies when determining whether to operate in a functionally equitable mode — equity’s domain usually consists in preventing fraud, accident, and mistake, and “proxies like inadequacy of the legal remedy and disproportionate hardship help pick out this domain.”99×99. Henry E. Smith, Equity as Second-Order Law: The Problem of Opportunism 65 (Harvard Pub. Law Working Paper No. 15-13, 2015), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2617413 [http://perma.cc/9ZAW-KYL5]. Bare “majorness,” of course, is not a good proxy for delegation; as discussed earlier, the Court recognizes all kinds of economically and politically significant delegations (the delegation in Chevron itself is Exhibit A).100×100. See supra text accompanying notes 44–45. Of course, it is also fair to question whether Mead’s procedural proxies accurately capture congressional intent to delegate, but the point here is one about consistency. But neither is “majorness” a good proxy for opportunistic behavior by agencies. The size and importance of the tax credit question in King, for example, says nothing about whether the IRS, in a moment of caprice, might later switch its interpretation and undermine the stability of the ACA. Assuming a connection between major questions and agency opportunism, in fact, comes close to reviving a theory the Court repudiated in City of Arlington: that allowing agencies to interpret “big, important”101×101. City of Arlington v. FCC, 133 S. Ct. 1863, 1868 (2013). provisions that define the scope of their authority necessarily “leaves the fox in charge of the henhouse.”102×102. Id. at 1874. Others have noted the parallels between Chief Justice Roberts’s dissent in City of Arlington and his majority opinion in King. See, e.g., Hickman, supra note 78, at 58; Jody Freeman, The Chevron Sidestep: Professor Freeman on King v. Burwell, Harv. L. Sch. Envtl. L. Program Blog, http://environment.law.harvard.edu/2015/06/the-chevron-sidestep [http://perma.cc/J3JG-Q93J]. In contrast to the Court’s own instructions in City of Arlington, King “establish[es] an arbitrary and undefinable category”103×103. City of Arlington, 133 S. Ct. at 1874. of questions — “major questions” — that fails to correspond with the vaguely defined kind of behavior the Court apparently wishes to counter.

At this point, it is worth stepping back from the intricacies of the steps of Chevron to consider the larger picture. Chevron ultimately aims to shift resolution of statutory indeterminacies to expert and politically accountable agencies. The more robust the pre-Chevron threshold requirements are — the more complicated it is to enter Chevron’s domain — the less meaningful is the inquiry into the reasonableness of the agency’s interpretation and the more power the courts actually exercise over the interpretive question. King’s gloss on the major question exception seems to confuse various concepts — “majorness,” congressional intent to delegate, and proxies for agency opportunism — in a way that leaves courts with broad, ex post discretion to decide whether courts or agencies should resolve statutory indeterminacies with far-reaching consequences. Perhaps it is exaggerating to compare King’s rendition of the major question exception to the notorious Chancellor’s Foot standard.104×104. Cf. The Table Talk of John Selden 62–63 (David Irving ed., 1854) (“[E]quity is according to the conscience of him that is chancellor, and as that is larger or narrower, so is equity. ’Tis all one as if they should make the standard for the measure we call a foot, a chancellor’s foot; what an uncertain measure would this be? One chancellor has a long foot, another a short foot, a third an indifferent foot: ’tis the same thing in the chancellor’s conscience.”). But if the point of Chevron is to restrain the courts in a certain domain, the major question exception effectively arrogates significant power right back.105×105. Cf. Manning, supra note 7.

B. Major Questions, Chevron, and the APA

If King’s incorporation of the major question exception into the pre-Chevron threshold inquiry is misguided, however, what about the cases where the Court has applied the exception while interpreting statutes under the Chevron framework — cases like Brown & Williamson? If the major question exception in those cases amounted to nothing more than an interpretive rule of thumb that broadly worded statutes should not lightly be taken to grant especially sweeping powers to agencies, perhaps it would be acceptable as a “traditional tool[] of statutory interpretation” under Chevron.106×106. Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843 n.9 (1984). In that case, it would amount to nothing more than the interpretive principle from Whitman v. American Trucking Associations, 531 U.S. 457 (2001), that “Congress . . . does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions — it does not, one might say, hide elephants in mouseholes.” Id. at 468; cf. Coal. for Responsible Regulation, Inc. v. EPA, No. 09-1322, 2012 WL 6621785, at *18 (D.C. Cir. Dec. 20, 2012) (Kavanaugh, J., dissenting from denial of rehearing en banc). But see Jacob Loshin & Aaron Nielson, Hiding Nondelegation in Mouseholes, 62 Admin. L. Rev. 19, 23 (2010) (“One judge’s mouse is another judge’s elephant . . . .”). But again, the Court often finds that broadly worded statutes grant agencies significant powers (Massachusetts v. EPA is an example); and more importantly, the cases simply do not bear out the idea that the major question exception is a mere tool of statutory interpretation.

Most of the latent concerns in the cases are less about “majorness” as such and more about “big changes” — concerns about the destabilizing effect of an agency’s changing its interpretation, usually in a charged political setting. Brown & Williamson focused on reliance interests: for thirty-five years, the FDA had maintained that the FDCA gave it no authority to regulate tobacco, and both Congress and the tobacco industry had relied on that understanding. When the FDA changed its interpretation — based on new information about how tobacco affects the body — it threatened to undo a kind of tacit agreement among Congress, the industry, and the executive branch over tobacco regulation.107×107. See sources cited supra note 85. King’s move looks like a crude way of preventing a future disruptive change. If the Court had deferred to the IRS’s interpretation as one reasonable possibility under Chevron, the IRS could conceivably have later switched its interpretation to disallow tax credits on federal exchanges, perhaps under a new administration.108×108. When the Chevron framework applies, and a statutory provision is ambiguous, an agency can vary its interpretation over time, so long as it stays within reasonable bounds and adequately explains any interpretive changes. See Chevron, 467 U.S. at 863–64. Chief Justice Roberts seemed particularly worried about this possibility, and its potentially disruptive consequences, at oral argument. Transcript of Oral Argument at 76, King v. Burwell, 135 S. Ct. 2480 (2015) (No. 14-114) (“[I]f you’re right about Chevron, that would indicate that a subsequent administration could change [your] interpretation?”); see also Abigail R. Moncrieff, King, Chevron, and the Age of Textualism, 95 B.U. L. Rev. Annex 1, 3 (2015). UARG is complicated by EPA’s “tailoring” of the licensing requirements in order to accommodate its interpretation of “air pollutants,” but the Court there was also skeptical of EPA’s surprising “newfound authority to regulate millions of small sources — including retail stores, offices, apartment buildings, shopping centers, schools, and churches.”109×109. UARG, 134 S. Ct. 2427, 2446 (2014). Intriguingly, a recent D.C. Circuit case hints at a slight variation on the “prevent big changes” use of the major question exception. See District of Columbia v. Dep’t of Labor, No. 14-5132, 2016 WL 1319453, at *1 (D.C. Cir. Apr. 5, 2016) (citing FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 160 (2000)) (pointing to the “novelty,” rather than the size or significance, of an agency’s changed interpretation as a reason not to defer).

The Court’s apparent concerns about “big changes” are better addressed under § 706(2)(A) of the APA, which instructs reviewing courts to “hold unlawful and set aside agency action, findings, and conclusions found to be arbitrary, capricious, [or] an abuse of discretion.”110×110. 5 U.S.C. § 706(2)(A) (2012). Unfortunately, defending this claim is somewhat tricky, since confusion surrounds the precise relationship between Chevron and arbitrary and capricious review, confusion the Supreme Court has done little to dispel.111×111. For academic commentary on the confusion, see, for example, Kenneth A. Bamberger & Peter L. Strauss, Chevron’s Two Steps, 95 Va. L. Rev. 611, 621 (2009) (claiming to identify an “emerging consensus” that equates Chevron’s second step with review under § 706(2)(A)); Jack M. Beermann, Chevron at the Roberts Court: Still Failing After All These Years, 83 Fordham L. Rev. 731, 732, 743–47 (2014) (condemning the Roberts Court’s failure to “clarify the boundary between Chevron and other standards of review such as arbitrary or capricious review,” id. at 732); Ronald M. Levin, The Anatomy of Chevron: Step Two Reconsidered, 72 Chi.-Kent L. Rev. 1253, 1254 (1997) (describing the D.C. Circuit’s puzzlement over the relationship between Chevron and arbitrariness review); Shane & Walker, supra note 2, at 478–79 (noting the confusion); Stephenson & Vermeule, supra note 15, at 603–04 (explaining how equating Chevron’s second step with arbitrary and capricious review creates unjustifiable redundancy). For the Supreme Court’s contribution to the confusion, see Judulang v. Holder, 132 S. Ct. 476, 483 n.7 (2011) (claiming that analysis under arbitrary and capricious review is the same as that of Chevron’s second step); Beermann, supra, at 745–47 (arguing that the Court’s reasons for equating the two analyses in Judulang are puzzling and unsatisfactory). A brief Note is not the place to attempt to cut through that Gordian knot. But the claim here is modest: whatever the precise relationship between Chevron and arbitrary and capricious review, the latter is better suited to addressing the Court’s concerns in major question cases, because it shifts the emphasis from interpretation to reason-giving in just the kinds of cases where the Court has insisted that sound reason-giving is the proper constraint on agency behavior.

Although in practice there is sometimes “substantial overlap” between Chevron and arbitrariness review,112×112. Patricia M. Wald, Judicial Review in Midpassage: The Uneasy Partnership Between Courts and Agencies Plays On, 32 Tulsa L.J. 221, 244 (1996); see also id. (“An agency which comes up with an impermissible interpretation of a statute will almost automatically be found to have acted in an arbitrary and capricious manner, and the failure to consider factors made relevant by the statute may render an agency’s interpretation unreasonable under Chevron step two as well as exposing its actions as arbitrary and capricious.”). the two differ in emphasis. It may help to think in terms of what the agency has to show in order to win its case: its “construction of the statute [must be] permissible on the merits, in light of the appropriate tools and principles of statutory interpretation,” and its “interpretive choice [must be] the product of reasoned decisionmaking,” not arbitrariness or caprice.113×113. Stephenson & Vermeule, supra note 15, at 598. Under arbitrary and capricious review, the agency has a chance to show the latter. In what will sound reminiscent of the earlier discussion about the distinction between Chevron and the pre-Chevron threshold inquiry into delegation,114×114. See supra text accompanying notes 26–28. a too-robust interpretive analysis under Chevron risks depriving the agency of the opportunity to justify its chosen interpretation among the available options.115×115. As Judge Wald notes:
Chevron was basically meant as a device to enhance the power of agencies vis-a-vis the courts and Congress; the courts were excluded from the policy judgments involved in interpreting ambiguous statutes and Congress was told to speak its mind clearly or risk ceding substantial control over policy to the executive branch. By confusing and collapsing Chevron step two with arbitrary and capricious review, the balance of power between courts and agencies is tilted back somewhat in the courts’ favor. Courts obtain an opening through which to subject agency interpretations to greater scrutiny . . . . Wald, supra note 112, at 244; id. (“[T]here are important reasons to keep the two analyses distinct.”).
Put another way, an aggressive approach to Chevron — such as the Court’s equitable approach in major question cases — may allow courts to resolve cases against agencies without examining their reasons for choosing a particular interpretation. In effect, courts can short-circuit arbitrary and capricious review by beefing up Chevron.

Yet in case after case outside the major question context, the Court has steered claims about reliance interests and related concerns arising from significant interpretive changes into arbitrary and capricious review.116×116. See, e.g., Perez v. Mortgage Bankers Ass’n, 135 S. Ct. 1199, 1209 (2015); FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515–16 (2009); Smiley v. Citibank (S.D.), N.A, 517 U.S. 735, 742 (1996). That, rather than any other administrative law doctrine, is the check on “[s]udden and unexplained change” or “change that does not take account of legitimate reliance on prior interpretation.”117×117. Smiley, 517 U.S. at 742. Just last Term, in a different context, the Court struck down extrastatutory procedural requirements invented by the D.C. Circuit to “prevent[] agencies from unilaterally and unexpectedly altering their interpretation of important regulations.”118×118. Perez, 135 S. Ct. at 1209. While acknowledging the importance of reliance interests, the Court disapproved the D.C. Circuit’s improvised, extra-APA mechanism for addressing them.119×119. Id. The Court expounded a general principle: arbitrary and capricious review already provides a way for courts to consider reliance interests, and courts have no business layering on requirements beyond the sound reason-giving that the APA requires.120×120. Id. (quoting Fox Television Stations, 556 U.S. at 515).

Perhaps it will be objected that the distinction between Chevron and arbitrary and capricious review is merely formal in major question cases — that the Court can easily reach the same outcome under arbitrary and capricious review. That is true, of course, if the Court’s decisionmaking is entirely outcome driven; criticizing the Court’s use of doctrine is pointless if the Court merely molds doctrine to reach predetermined outcomes. But if we assume that the Court takes doctrine seriously, the major question cases might have come out differently under arbitrary and capricious review. Rather than focus almost entirely on the surrounding legislative landscape and the “majorness” of the question, for example, the Brown & Williamson Court would have had to grapple with the FDA’s extensive reason-giving in its rulemaking record.121×121. See Regulations Restricting the Sale and Distribution of Cigarettes and Smokeless Tobacco to Protect Children and Adolescents, 61 Fed. Reg. 44,396 (Aug. 28, 1996); Manning, supra note 5, at 225 (“No doubt because of the importance of the question, or because the agency had previously asserted that it lacked jurisdiction over tobacco, the FDA supported its new position with an unusually detailed factual, policy, and legal analysis, including a separate ‘annex’ on jurisdiction that occupied almost 700 pages in the Federal Register.”). Simply going through that exercise might have made it more difficult for the Court to reject the FDA’s interpretation. Imagining the effect on the other cases is a touch more complicated, since they all turned somewhat on comparisons to Brown & Williamson, but the point is that arbitrary and capricious review would not be an empty exercise in major question cases.122×122. Insisting that courts plod through each step of an articulated evaluation process may help “keep the starch” in the process. Cf. Denver Area Educ. Telecomms. Consortium, Inc. v. FCC, 518 U.S. 727, 774 (1996) (Souter, J., concurring) (“Reviewing speech regulations under fairly strict categorical rules keeps the starch in the standards for those moments when the daily politics cries loudest for limiting what may be said.”).

It makes sense to close by returning to the larger concept of the major question exception as an anti-abuse principle — as an equitable standard that allows Chevron to operate as a simple rule in most cases.123×123. See supra text accompanying notes 81–84. The major question exception is redundant as an equitable tool: arbitrary and capricious review is the APA’s backstop or safety valve, its check on “administrative waywardness,” the standard that allows the simple rule to operate.124×124. Cf. Evan J. Criddle, Fiduciary Foundations of Administrative Law, 54 UCLA L. Rev. 117, 151–52 (2006) (comparing arbitrary and capricious review to equitable standards of review for private fiduciary duties). And the Court’s own understanding of § 706(2)(A) requires it to examine the agency’s reasons, and to treat reliance and similar considerations as factors to be weighed, not Chevron trumps. Because it allows the Court to cut the inquiry off at the “interpretation” stage without considering the agency’s reasons for its decision, the major question exception artificially curtails the ordinary process. Sufficiently “big” questions, by the Court’s own measure, simply evade the reasoned decisionmaking part of the inquiry altogether.

IV. Conclusion

The Court constantly finds itself tidying up administrative law, often to correct lower-court departures from the APA.125×125. E.g., Vt. Yankee Nuclear Power Corp. v. Nat. Res. Def. Council, Inc., 435 U.S. 519 (1978). Sometimes it creates its own messes, though; the major question exception is such a mess. The major question exception as a pre-Chevron threshold matter is unsound: major questions are reliable proxies for neither delegation nor agency opportunism. The major question exception within the Chevron framework short-circuits arbitrary and capricious review, releasing courts from evaluating agencies’ reasons for changed, and perhaps politically charged, interpretations. Yet the Court itself insists that the kinds of concerns that arise in major question cases are best addressed under arbitrary and capricious review. The Court should follow its own guidance and remove the major question excrescence from administrative law.