Introduction
The Supreme Court faces a real dilemma in Loper Bright Enterprises v. Raimondo,145 F.4th 359 (D.C. Cir. 2022), cert. granted in part sub nom., Loper Bright Enters. v. Raimondo, 143 S. Ct. 2429 (argued Jan. 17, 2024) (mem.). in which the Court will explicitly consider whether to overrule Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc.2467 U.S. 837 (1984); see Petition for Writ of Certiorari, Loper Bright, 143 S. Ct. 2429 (No. 22-451). The dilemma is decades in the making and arises from the interplay of large structural forces, between which the Court is uneasily positioned. On the one hand, the background conditions of the American administrative state, which produce an array of broad and vague delegations to administrative agencies on highly technical subjects, tend to limit the scope of judicial review of agency legal interpretations. The last thing judges want to be forced to do is to decide for themselves what exactly statutes mean when they refer to “unreasonable risk of injury to health or the environment”3Toxic Substances Control Act, 15 U.S.C. § 2603(a)(1)(A)(i)(I). or “best system of emission reduction.”4Clean Air Act, 42 U.S.C. § 7411(a)(1). On the other hand, the fundamental importance of judicial review of agency legal authority as a legitimating mechanism for the administrative state presses judges towards plenary review of agency legal interpretations for reasons Professor Louis Jaffe explored in the middle of the twentieth century. As Jaffe put it, “[t]he availability of judicial review is the necessary condition, psychologically if not logically, of a system of administrative power which purports to be legitimate, or legally valid.”5Louis L. Jaffe, Judicial Control of Administrative Action 320 (1965).
The combination of these two large-scale pressures creates the deference dilemma. In the limit, the deference dilemma threatens to make plenary judicial review of agency legal interpretations both intolerable and indispensable. The problem for the Court is now—as it was before Chevron, at the time of Chevron, and will be in the future—how to navigate between these countervailing pressures.
In what follows, I explain this basic dilemma, explore its causes and sources, and speculate about some possible futures for the Chevron framework in particular and the deference dilemma in general. Although there are many possibilities, too many for any confident predictions, the two most interesting futures are either (1) a distinct narrowing of Chevron’s scope and force, adding preconditions and limitations in much the same way that the Court in Kisor v. Wilkie6139 S. Ct. 2400 (2019). narrowed but also stabilized Auer7Auer v. Robbins, 519 U.S. 452 (1997). deference to agency interpretations of their own rules; or (2) an express overruling of Chevron as a wholesale framework, combined with a reframing of “deference” at retail that preserves much of the content of Chevron under a different label.
On the second possibility, deference will be reframed but not eliminated. The overruling majority will say—along lines indicated by Professor Henry Monaghan decades ago8See Henry P. Monaghan, Marbury and the Administrative State, 83 Colum. L. Rev. 1, 2–3, 6 (1983).—that de novo or plenary judicial review of agency legal interpretations is required by legal sources (either by the Administrative Procedure Act (“APA”), by Article III, or both). Yet the Court will also say that de novo interpretation might of course itself yield the conclusion that, in a given statute, Congress has delegated primary responsibility to agencies to fill in statutory gaps or ambiguities, subject to judicial review to ensure that agencies have remained within the scope of the delegation and chosen policy on reasonable grounds. This second possibility—call it retail Chevron rather than wholesale Chevron—will offer the Justices skeptical of Chevron an attractive resolution of the dilemma; it will allow the majority not only to overrule Chevron but, even more importantly, to be seen to overrule Chevron, while also largely preserving Chevron’s major source of appeal to judges—a way to avoid having to actually do fully independent interpretation of statutory terms that are vague, technical, or both.
Part I explains the two horns of the dilemma. Part II explains some possible futures for Chevron and explores the most interesting ones, focusing especially on the possibility that, even if Chevron is overruled, deference will be reframed but not eliminated.
I. The Dilemma
A. “The Hallmark of the Administrative State”: Why Deference Occurs
1. Background Conditions of the Administrative State
In a famous 1989 article on Chevron, Justice Antonin Scalia wrote that Chevron rests on a global presumption or “benign fiction”9Green v. Bock Laundry Mach. Co., 490 U.S. 504, 528 (1989) (Scalia, J., concurring in the judgment).—an attribution of a kind of general, trans-statutory default intention to Congress.10Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 516 (1989). Although Scalia disclaimed any intent to defend that presumption, he then immediately went on to defend it in the following terms:
Surely, however, it is a more rational presumption today than it would have been thirty years ago—which explains the change in the law. Broad delegation to the Executive is the hallmark of the modern administrative state; agency rulemaking powers are the rule rather than, as they once were, the exception . . . .11Id.
Just as, in Scalia’s view, the conditions of the administrative state made Chevron more plausible in 1989 than before, so too, if anything, the conditions of the administrative state make Scalia’s argument for Chevron even more plausible in 2023 than in 1989. Congress is no more functional now than then; arguably, it is far more dysfunctional. When it legislates, it does not always or even usually delegate in more specific, cabined forms than when Scalia wrote; indeed, major enactments in the intervening years have seen broad, vague, or general delegations to the executive, as well as delegations on technical subjects in complex regulatory domains such as financial regulation12See, e.g., Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. 111-203, 124 Stat. 1376 (2010). and health care.13See, e.g., Patient Protection and Affordable Care Act, Pub. L. 111-148, 124 Stat. 119 (2010).
One obvious reaction to such an environment, on the part of legal libertarians, has been to attempt to revive the nondelegation doctrine as a constitutional matter. And indeed, the libertarian legal movement has pursued that effort in recent decades in parallel to the critique of Chevron.14See, e.g., Cass R. Sunstein & Adrian Vermeule, Libertarian Administrative Law, 82 U. Chi. L. Rev. 393, 415–16 (2015). Conceptually and legally, however, the critique of Chevron and the constitutional critique of delegation present different issues. Many of the broad, vague, or ambiguous statutory grants of authority that, for Scalia, constitute the hallmark of the administrative state and the premise for judicial deference, nonetheless easily survive the standard “intelligible principle” test that is the centerpiece of the Court’s standard nondelegation analysis.15See Gundy v. United States, 139 S. Ct. 2116, 2129 (2019) (plurality opinion) (tracing the intelligible principle test through history and case law). Although Justice Neil Gorsuch has led a campaign to narrow the standard test,16See id. at 2133–42 (Gorsuch, J., dissenting). that campaign seems to have faltered in the past few years with no new cases squarely presenting the validity of the “intelligible principle” test reaching the merits docket—in part because the Court has increasingly turned to the major questions doctrine at a sub-constitutional level.17See Biden v. Nebraska, 143 S. Ct. 2355, 2374 (2023); West Virginia v. EPA, 597 U.S. 697, 728–30 (2022). There is skepticism even within the Court’s six-Justice majority about whether the major questions doctrine is best understood as implementing background constitutional principles of nondelegation.18See Biden, 143 S. Ct. at 2377–78 (Barrett, J., concurring) (arguing that major questions doctrine is best understood as resting, not on nondelegation principles, but on ordinary meaning in context). For skepticism about the coherence of Justice Barrett’s approach, see Adrian Vermeule, Text and “Context,” Yale J. on Regul.: Notice & Comment (July 13, 2023), https://perma.cc/DP98-PJ5D. Loper Bright, furthermore, does not present a constitutional nondelegation issue. Any argument against Chevron that, boiled down, amounts to a more stringent version of constitutional nondelegation analysis, amounts to changing the subject, and the question presented, in Loper Bright.
In this legal and institutional environment, featuring broad, vague, and open-ended grants of authority to the executive that are valid under the still-prevailing constitutional nondelegation test, four fundamental circumstances obtain and, taken together, press judges towards some form of deference to agency legal interpretations. I will call these four circumstances respectively (1) the fundamental logic of Chevron; (2) the managerial imperative; (3) judicial self-knowledge; and (4) Chevron, major questions, and merits avoidance. Let me explain each in turn.
2. The Fundamental Logic of Chevron
The Chevron opinion is analytically flawed in several respects. First, as I have (co-)written elsewhere,19Matthew C. Stephenson & Adrian Vermeule, Chevron Has Only One Step, 95 Va. L. Rev. 597, 597–98 (2009). the famous Chevron two-step framework is unnecessarily complicated. Nothing of substance would change, while much confusion would be avoided (and many law review articles made otiose), if the Court had simply said that agency interpretations must be “reasonable” and that an agency interpretation contrary to clear statutory meaning is of course necessarily unreasonable, de jure. Chevron can, with no loss of content, be understood to have only one step.
Another, more significant confusion about Chevron is that “deference” is in one sense a misnomer for what happens in Chevron cases—an analytically crucial point to which I shall return when considering possible futures for the Chevron doctrine in Part II. The confusion arises from the mistaken notion that “deference” is the antonym of de novo interpretation. Just before Chevron was decided, Monaghan pointed out that as a logical matter, “deference” is entirely consistent with de novo judicial interpretation; the former is just a byproduct of the latter.20Monaghan, supra note 8, at 27–28. When a court interprets a statute de novo, it may decide that the statute itself, rightly understood according to the traditional tools of statutory interpretation, either explicitly or implicitly delegates to an agency the primary or initial power to determine21Adrian Vermeule, Deference and Determination, Ius & Iustitium (Dec. 2, 2020), https://perma.cc/U43T-CW2U. or complete22Jack Goldsmith & John F. Manning, The President’s Completion Power, 115 Yale L.J. 2280, 2282 (2006). the statutory scheme (whether by rulemaking, by adjudication, or by a mix—which of these powers an agency holds is a separate question23See Nat’l Petrol. Refins. Ass’n v. FTC, 482 F.2d 672, 674 (D.C. Cir. 1973).). In such cases, the agency gives concrete specification to statutory terms or fills in statutory gaps or ambiguities, subject of course to judicial review to ensure that the agency’s specification has remained within the scope of the delegation.24Vermeule, supra note 21. A judicial interpretation that fixes the scope of the delegation is, to repeat, itself an entirely de novo interpretation. “Deference” is then just a description of the collateral effect of judicial interpretation that ascertains the scope of the delegation; “deference” is a byproduct of the judicial power to say what the law is.25Monaghan, supra note 8, at 27–28. It is a conceptual error to conceive of this sort of situation as one in which the court formulates its own interpretation but then adopts an agency interpretation that (however reasonable) is incorrect according to the court’s own view. Rather, as Monaghan explains, the court’s own interpretation is the agency has primary authority to determine or specify statutory meaning within a certain domain.26Id. at 25.
All that said, if the doctrinal framework Chevron erected was poorly constructed, and if the language of “deference” to which it has given rise to rests on a damaging misconception, there is a central line of reasoning in Chevron that was forceful when written and whose force is, if anything, even more visible today. The fundamental logic of the Chevron opinion runs this way: sometimes, statutes contain express delegations of gap-filling authority to agencies.27See Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843–44 (1984). See generally Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. 111-203, 124 Stat. 1376 (2010); Patient Protection and Affordable Care Act, Pub. L. 111-148, 124 Stat. 119 (2010). Such delegations may or may not be problematic from a constitutional standpoint, but that is a separate and independent question, and, under current law, typically an easy question.28See Gundy v. United States, 139 S. Ct. 2116, 2129 (2019). To the extent that Justice Gorsuch and a handful of other Justices hope to enforce a more stringent version of the constitutional nondelegation doctrine,29Id. at 2131, 2141 (Gorsuch, J., dissenting). that enterprise has not succeeded and shows every sign of flagging.30See Biden v. Nebraska, 143 S. Ct. 2355, 2377–78 (2023).
Suppose, however, that the statute does not contain any such explicit delegation but merely contains gaps or ambiguities. Should the court take this to represent an implicit delegation? The court may also do so implicitly, and as things currently stand, there is no basis in law for discerning a general clear statement rule against agency gap-filling authority. (As I will subsequently discuss, the major questions “doctrine” is more limited, both by its terms and by its necessarily selective operation).
The fundamental logic of Chevron is that a court then has only two options. It may attempt to decide, for itself, what exactly the statute means, taking on the daunting task of giving specification to statutory terms like “unreasonable risk,” or what exactly counts as a “stationary source” of air pollution. Alternatively, a court may read the statute to leave such authority to the agency, contenting itself with policing the outer boundaries of the scope of the implicit delegation.31See, e.g., Auer v. Robbins, 519 U.S. 452 (1997); Kisor v. Wilkie, 139 S. Ct. 2400 (2019); Gundy v. United States, 139 S. Ct. 2116 (2019). Congress, by hypothesis, is no longer in the picture; there is no one left on the scene but agencies and courts. This, too, is a subsidiary dilemma of judicial review in the administrative state.
In its strongest passage, the Chevron opinion argued that, in this difficult situation, for courts to fill in the statutory scheme for themselves would result in law that is both less informed and less politically accountable than treating the statute as an implicit delegation.32Chevron, 467 U.S. at 865–66. Here, there is no tradeoff or conflict between political responsibility, on the one hand, and the “artificial reason of the law”—the special competence of lawyers and judges to interpret the law—on the other. The relevant terms, gaps, or ambiguities are not ones to which legal training affords any special comparative expertise for generalist judges. Nothing in law school education, legal practice, or legal training equips the judge to give specifying content to “unreasonable risk” or “stationary source,” no matter how many law dictionaries the judge consults. That core logic of Chevron stands unimpaired by the errors and confusions to which the awkwardly crafted Chevron framework and two-step test has otherwise given rise. The core logic of the decision is as powerful today as it was when written.
3. The Managerial Imperative
A second factor is one that Professor Peter Strauss identified long ago: Chevron partially centralizes interpretative authority in agencies, within the bounds of reasonableness.33Peter L. Strauss, One Hundred Fifty Cases Per Year: Some Implications of the Supreme Court’s Limited Resources for Judicial Review of Agency Action, 87 Colum. L. Rev. 1093, 1119–20 (1987). From the Court’s standpoint, the managerial risk inherent in a world of genuinely independent interpretation is that lower courts will reach many more conflicting interpretations of statutes and that curing such conflicts will place large, new burdens on the Court’s collective time and attention.34Id. at 1107–08, 1119–21.
4. Judicial Self-Awareness
A final factor pressing the law towards some version of deference is that, often enough, judges know what they do not know. Justice John Paul Stevens, Chevron’s author, is recorded to have said at the conference on the case, “When I am confused, I go with the agency.”35Mark Sherman, High Court Climate Case Looks at EPA’s Power, Huff Post (Feb. 23, 2014, 3:59 PM), https://perma.cc/SQ6K-5ANA. From one standpoint, this is a confession of judicial failure; when judges are confused, one might think they should work on the case with the traditional tools of interpretation until they become unconfused. From another standpoint, however, Justice Stevens here displays a laudable epistemic humility about the limits of the lawyer’s own competence and skills.
This judicial self-awareness of judges’ limitations is, probably, the single factor or condition that is most likely to sustain “deference” to agencies in some form (although, as indicated in Part II, the nominal label for that deference may possibly change after Loper Bright). Justice Elena Kagan’s plurality opinion in Kisor v. Wilkie appealed to it powerfully and, in doing so, seems to have brought home to her colleagues—enough of them anyway—the daunting concrete practical problems involved in judges committing themselves to interpret, de novo, regulatory terms such as “diagnosed” or “active moiety.”36Kisor, 139 S. Ct. at 2410. By parallel, or indeed a fortiori, the relatively abstract arguments about Article III and the APA, while attracting the lion’s share of academic attention, will recede in importance as the arguments in Loper Bright focus on what independent judicial interpretation requires or demands of judges. The prospect of having to decide what counts as an “unreasonable risk” will concentrate the judicial mind wonderfully.37Cf. James Boswell, The Life of Samuel Johnson 231 (Christopher Hibbert ed., Penguin Books 1986) (1791) (“[W]hen a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully.”).
So-called “Skidmore deference”38See Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944). does not cure the problem of judicial self-awareness to which Stevens’s quip pointed, at least not wholly. Because Skidmore deference is a form of epistemic deference and thus persuasive rather than binding in some sense, it has been expressly compared to the judicial consideration of amicus briefs, law review articles, and other sources39See Bradley Lipton, Accountability, Deference, and the Skidmore Doctrine, 119 Yale L.J. 2096, 2126 (2010) (“The agency’s decision is treated by the court in essentially the same manner as a brief by any other party in litigation.”); see also Jim Rossi, Respecting Deference: Conceptualizing Skidmore Within the Architecture of Chevron, 42 Wm. & Mary L. Rev. 1105, 1110 (2001) (“[H]istorically courts and scholars have paid scant attention to what Skidmore deference means. Few law review articles address the topic.”); Mayburg v. Sec’y of Health & Hum. Servs., 740 F.2d 100, 106 (1st Cir. 1984) (“The simple fact that the agency has a position, in and of itself, is of only marginal significance.”); Christopher M. Pietruszkiewicz, Discarded Deference: Judicial Independence in Informal Agency Guidance, 74 Tenn. L. Rev. 1, 6–7 (2006). that are neither binding law nor, on some views, even direct sources of law at all. Persuasive sources are chronically in disagreement; law professors rarely all write on the same side of a contested question and amicus briefs likewise generally do not all favor one side of a contested Supreme Court case. Such sources, therefore, pose, rather than answer, the question to whom the judge should defer, even in a merely epistemic sense. In deciding that question, the judge must inevitably decide, with limited competence, which expert advisor’s view of (say) “unreasonable risk” is more persuasive; the need to make such a decision thus threatens to merely replicate, at one remove, the judge’s limited epistemic competence. Under Chevron’s authority-based deference, by contrast, Congress itself has designated the primary determiner or specifier of the statute, at least within outer boundaries.
5. Chevron, Major Questions, and Merits Avoidance
Ironically enough, judicial self-awareness of the judges’ own epistemic limitations underpins both Chevron and the burgeoning “major questions doctrine,” often seen as Chevron’s antagonist. To be sure, the two are genuinely antagonistic on one major dimension. The major questions doctrine often operates to override or, more accurately, preempt, a Chevron analysis that might otherwise occur. Yet on another dimension, both doctrines can be seen as in fundamental accord; both allow judges to avoid having to say, independently, what exactly counts as the “best system of emissions reduction”40West Virginia v. EPA, 597 U.S. 697, 709 (2022). or as a “stationary source[ ].”41Id. In West Virginia v. EPA,42597 U.S. 697 (2022). for example, the Court did not have to decide what exactly the key statutory phrase meant throughout the range of possibilities. All it had to decide was that EPA’s regulatory scheme fell outside the “ordinary” boundaries of the delegation and instead posed an “extraordinary” question, to which Congress would be deemed not to have delegated authority at all unless and until Congress spoke more clearly.43Id. at 721. In this sense and on this dimension, far from being antagonists, Chevron and the major questions doctrine both serve a crucial merits-avoidance function. Both allow judges to avoid, at all costs, the independent interpretation that some of them profess to be intrinsic to the judicial office, yet which some also fear is simply beyond the practical capacities of generalist judges, at least as applied to broad, vague, ambiguous, or incompletely specified and technically complex regulatory statutes.
B. “Government of a Bureaucratic Character”: Deference and Legitimation
1. The Legitimating Function of Judicial Review
Now let me turn to the other horn of the dilemma. I have urged that “deference” (however misnamed) can straightforwardly be squared with de novo review as a logical matter. I have also urged that powerful structural conditions of the administrative state underscore the force of Chevron’s core logic, putting to courts an unavoidable choice between some version of deference and independent interpretation of complex, incompletely-specified, and technical statutes—a type of interpretation that, judges are often aware, lies beyond the limits of their capacities.
If all this is so, however, why has skepticism about deference proven such a persistent force in U.S. legal discourse in the past several decades? The phenomenon doubtless has multiple roots and causes, some internal to legal theory and practice, some external. The latter category includes the growth of a well-funded network of libertarian paralegal institutions—think tanks, legal centers, and others—that have consistently urged both critiques of Chevron deference in particular and skepticism about the administrative state in general, and that directly or indirectly underwrite scholarship in those registers.44See Thomas A. Berry & Isaiah McKinney, It’s Time to Overrule Chevron, Cato Inst.: Cato Liberty Blog (Jul. 24, 2023, 12:09 PM), https://perma.cc/5PHR-FDL6; see also GianCarlo Canaparo & Jack Fitzhenry, Chevron Deference, Long Abused by Federal Agencies, on Supreme Court’s Chopping Block?, Heritage Found. (May 5, 2023), https://perma.cc/MMM9-2BMP. That said, however, I will focus on the internal causes.
Let me begin with Louis Jaffe, the great mid-century scholar of administrative law and author of a canonical treatise on judicial review of administrative action.45Jaffe, supra note 5, at 1. Jaffe urged that judicial review by the courts was an indispensable condition of the legitimacy of the administrative state—not necessarily for high reasons of constitutional or legal theory but as a pragmatic observation about Anglo-American legal culture and the public psychology that both affects and is affected by it. If I may be forgiven, a self-quotation:
Jaffe’s most famous sentence, and one of the best-known ideas in administrative law theory, is his pronouncement that “[t]he availability of judicial review is the necessary condition, psychologically if not logically, of a system of administrative power which purports to be legitimate, or legally valid.” Elsewhere Jaffe clarifies the relationship between legitimacy and legal validity; the latter, in his view, is at least a precondition for the former. “The guarantee of legality by an organ independent of the executive,” he writes, “is one of the profoundest, most pervasive premises of our system.” Indeed, anticipating later arguments by Jack Goldsmith with respect to presidential powers, Jaffe suggests that some form of constraint is itself the precondition of expansive executive power: “Indeed I would venture to say that [judicial review] is the very condition which makes possible, which makes so acceptable, the wide freedom of our administrative system, and gives it its remarkable vitality and flexibility.” On psychological grounds, national publics will rebel against “monstrous expressions of administrative power” that are unconstrained by law.46Adrian Vermeule, Bureaucracy and Distrust: Landis, Jaffe, and Kagan on the Administrative State, 130 Harv. L. Rev. 2463, 2472 (2017) (alteration in original) (footnotes omitted).
When Jaffe wrote, the psychological condition he observed had already been elevated to a legal or indeed constitutional principle by one of the leading cases legitimating the administrative state, Crowell v. Benson.47285 U.S. 22 (1932). Crowell announced an elaborate scheme for judicial review of administrative action (in the context of formal adjudication by administrative tribunals; rulemaking, although hardly unknown when Crowell was decided, had yet to become a centerpiece of administrative action).48Id. at 50–51. Judicial review of questions of law was to be entirely de novo.49Id. at 65. As to questions of fact, Crowell announced that the baseline approach would be substantial evidence review (again, in the context of formal adjudication) but with exceptions for constitutional and jurisdictional facts.50Id. at 46, 60, 62–63.
All the elements of the Crowell framework were curious, albeit in different ways. As to de novo review of questions of law, the Court’s stricture was not a fair description of the law even when decided. As the government’s brief and some of the amicus briefs in Loper Bright emphasize, Supreme Court precedents offering deference of some sort to executive construction of ambiguous statutes go back to the very beginnings of the republic.51See Brief for Respondents at 22, Loper Bright Enters. v. Raimondo, 143 S. Ct. 2429 (filed Sept. 15, 2023) (No. 22-451). A leading example is United States v. Vowell529 U.S. (5 Cranch) 368 (1810).—a case illustrating that deference was afforded non only on mandamus, but on direct review as well. In any event, Crowell’s commitment to independent judicial interpretation would, in the view of its proponents, soon be threatened by the Court’s deference jurisprudence of the later 1930s and 1940s and would later be directly threatened by Chevron53See Reuel E. Schiller, The Era of Deference: Courts, Expertise, and the Emergence of New Deal Administrative Law, 106 Mich. L. Rev. 399, 433 (2007); Adrian Vermeule, Neo-?, 133 Harv. L. Rev. F. 103, 106 (2020).—even though, as I have argued, there is, in principle, no logical inconsistency between deference and de novo review.
So too, the exceptions for jurisdictional and constitutional fact were based on a constitutional vision that can only be described as overheated, perhaps even alarmist, and an illustration of the powerful psychological forces to which Jaffe later adverted. To allow deferential judicial review of constitutional and jurisdictional facts, Chief Justice Charles Evans Hughes wrote, “would be to sap the judicial power as it exists under the Federal Constitution, and to establish a government of a bureaucratic character alien to our system.”54Crowell, 285 U.S. at 57. Here, Crowell’s reasoning, although nominally confined to situations in which “fundamental rights depend, as not infrequently they do depend, upon the facts, and finality as to facts becomes in effect finality in law,”55Id. applied straightforwardly also to legal questions; indeed, the Court expressly saw constitutional and jurisdictional “facts” as legal predicates for factual determinations, rather than factual questions in the ordinary sense.56Id. at 54–55.
Overall, Crowell’s vision is one in which the administrative state is legitimated by the possibility of limited judicial deference as to ordinary factual questions only, combined with plenary judicial review of legal questions and legal predicates for factual questions to keep the administrative state within the boundaries of constitutionalism and law—the same vision that Jaffe would later adumbrate. Crowell, then, may be seen as Chevron’s symbolic antonym; while the latter represents the forces that impel courts away from independent judicial interpretation, the former tried to embed within constitutional law Jaffe’s psychological, if not logical, condition for the legitimacy of the administrative state.57See Jaffe, supra note 5, at 320.
2. The Force of the Dilemma
The Court, then, is caught between two large-scale background conditions of the administrative state, both of which are not (merely) external to law but have been internalized within law as background principles and even as doctrines of judicial review of agency action. Two things are true: (1) the nature of statutory delegations in the administrative state makes genuinely independent interpretation by judges pragmatically intolerable across a range of cases involving broad, vague, and technically complex regulatory statutes; and (2) at some stage or another, independent judicial review is, for both legal and cultural reasons, at least presumptively58That is, subject to the Administrative Procedure Act’s exceptions to the presumption of reviewability. 5 U.S.C. § 701. See infra note 67. indispensable to legitimacy. These two truths define the poles between which the Court must navigate.
II. Chevron Futures?
In light of the deference dilemma, consider a range of possible futures for the law of judicial review of agency legal interpretations of statutes—both for the Chevron framework in particular and deference in general. As I have emphasized, these are not the same topic; the Court might modify the Chevron framework or jettison it altogether while retaining, under some doctrinal rubric or another, a version of “deference”—including the Monaghan version of deference that makes deference merely a byproduct of de novo interpretation.59See Monaghan, supra note 8, at 9. I return to that possibility below.
A. Simple Resolutions
The two simplest resolutions of Loper Bright would be (1) a simple overruling of Chevron, combined with a declaration that the Constitution, the APA, or both require the Court to interpret statutes de novo or (2) a disposition that essentially ducks the large questions of the status of Chevron, neither overruling nor reaffirming it, but leaving it in the odd limbo to which it has been implicitly condemned for the past several years. Under the second approach, the Court would merely say that the statute has a clear meaning, such that the question of Chevron’s status does not arise. This is the path of “Chevron avoidance”60See Adrian Vermeule, Our Schmittian Administrative Law, 122 Harv. L. Rev. 1095, 1127 (2009). that the Court has sometimes taken in recent cases—at least when identifying a clear meaning would not commit the Justices to taking on the burdens of, in effect, making consequential policy through independent interpretation in complex regulatory domains.61See, e.g., Am. Hosp. Ass’n v. Becerra, 596 U.S. 724, 736 (2022); Becerra v. Empire Health Found. ex rel. Valley Hosp. Med. Ctr., 597 U.S. 424, 434 (2022).
I take the second approach to be self-explanatory; the Court would essentially postpone the issue of Chevron’s larger status, leaving the fight for another day. Indeed, one might understand the Court, under the second disposition, to say, in essence, that the Court (meaning of course a critical mass or majority of the Justices) is pragmatically content with the status quo of uncertainty about Chevron’s continuing status. This disposition would, of course, raise the question why the possibility of overruling Chevron was taken up in the first place.
The first approach is more complex than might initially appear, however. As I will explain shortly, a ringing declaration of judicial authority, and indeed duty, to interpret statutes de novo could easily, as a logical matter, be combined with an approach to de novo interpretation that effectively recreates deference by another name, along the lines of NLRB v. Hearst Publications.62NLRB v. Hearst Publ’ns, 322 U.S. 111, 130 (1944), superseded by statute, Social Security Act of 1948, ch. 468, § 2(a), 62 Stat. 438, as recognized in Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318 (1992). The second step of that approach might, of course, be executed in the same decision that overrules Chevron or in a later decision. Thus, even if Loper Bright overrules Chevron in express (but possibly nominal) terms, one will still have to wait and see whether, or when, the Court falls back upon that expedient to avoid the practical burdens of global, genuinely independent interpretation.
Kisor-ing Chevron. Another very thinkable disposition would be an opinion, perhaps by plurality, that essentially does for Chevron what Kisor did for Auer deference to agency interpretations of their own rules. Under this approach, the Court could both underscore and strengthen the limitations laid out in United States v. Mead,63533 U.S. 218 (2001). including the preconditions for Chevron to apply at all, and clarify or add new limitations within the Chevron framework. A number of such limitations could be added (or, if one believes them to be already implicit in extant doctrine, clarified). I will not attempt to canvass all the possibilities here, but one conspicuous possibility deserves mention.
In the early case law, the Court sometimes emphasized that it would defer not to any executive interpretation of ambiguous text at all but (only) to longstanding executive constructions. That limitation does not invariably appear but is at least a persistent theme throughout the history of deference in American law.64See Thomas W. Merrill, Judicial Deference to Executive Precedent, 101 Yale L.J. 969, 1019 (1992). A Court inclined to nominal originalism and (less nominal) traditionalism could then conceivably modify Chevron essentially by abandoning the National Cable and Telecommunications Ass’n v. Brand X Internet Services,65545 U.S. 967 (2005). under which agencies have in principle unrestricted discretion to switch their interpretations within the zone of reasonableness of a statutory gap or ambiguity, subject to arbitrariness review for the validity of their new interpretations.66See id. at 981–82. Note of course, however, that under FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515–16 (2009), when new agency policies implicate reliance interests, the agency must give additional explanation to show the reasonableness of the change of policy.
B. De Novo Deference
A final possibility—and perhaps the most theoretically interesting possibility—is that the Court will, in essence, write a Janus-faced opinion (or, over time, a series of opinions) embodying two very different commitments: (1) overruling Chevron in its wholesale form and (2) immediately or eventually reinstating a retail form of Chevron deference by embracing the Monaghan logic. At the second step, whether taken in Loper Bright itself or in a later case, the Court would say that de novo interpretation itself requires courts to ascertain the scope of statutory grants of authority, and that sometimes, at retail, the grant of authority will leave a space within which Congress has delegated primary or initial authority to the agency to determine or specify the statutory standards. Such specification would presumptively be subject to review on the usual grounds (if review is otherwise available under the usual APA presumption of reviewability and exceptions to that presumption67See Administrative Procedure Act, 5 U.S.C. § 701(a)(1)–(2).): to ascertain that the agency has indeed remained within the bounds of the area entrusted to the agency for determination, to ensure that the agency has made its determination on reasonable grounds as opposed to arbitrarily and capriciously, and to ensure that the agency has properly found facts.68See 5 U.S.C. § 706. In essence, the Court on this view would reinstate Hearst Publications by saying that it is for judges to say what the law means, but that sometimes, the law itself means that Congress has entrusted to agencies, not courts, the power in the first instance to make reasonable determinations of vague, ambiguous, or general statutory standards, so long as the agency’s view has a “reasonable basis in law.”69NLRB v. Hearst Publ’ns, 322 U.S. 111, 131 (1944), superseded by statute, Social Security Act of 1948, ch. 468, § 2(a), 62 Stat. 438, as recognized in Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318 (1992).
By so doing, the Court would attempt to square the circle, reconciling—at least nominally—the twin forces I have discussed in Part I. From an internal legal perspective, the Court would boldly proclaim as a matter of high principle that either the Constitution or the APA (or both) require de novo interpretation by judges on all legal questions, while also saying, more quietly, that judges need not solve difficult, policy-laden interpretive problems by themselves (with or without whatever aid Skidmore affords), on the order of what counts as “unreasonable risk,” or what counts as a “stationary source.” Politically speaking, from an external perspective, the Court would thereby accomplish two aims: (1) it would give the libertarian legal movement,70Adrian Vermeule, There Is No Conservative Legal Movement, Wash. Post (July 6, 2022, 12:50 PM), https://perma.cc/KFH8-ATFR. skeptical of the administrative state, a conspicuous public victory, while also sidestepping the severe pragmatic problems involved in generalist judges stepping well beyond the limits of their competence, and (2) the judicial management problems that might arise if every regional or specialized court of appeals could decide for themselves what statutes mean.
The downside of such an approach, of course, would be that while it has a kind of logical coherence, as Monaghan emphasized,71Monaghan, supra note 8, at 9. it also has a kind of pragmatic incoherence. In effect, the Court would simultaneously overrule Chevron while reinstating a version of deference, not so different than Chevron, just under another legal rubric shorn of the awkward and confusing two-step Chevron test. What the Court takes away with one hand, it would give back with the other, whether in the same case or a future case.
While intended to please everyone, such a course might, in the end, please no one. Legal libertarians who look past the dramatic headline, the overruling of Chevron, and understand what has occurred at the level of operative law, might justly feel the promise of their decades-long effort has been betrayed. Even while Chevron has been overruled, deference is here to stay. Conversely, defenders of Chevron on the Court might fear its conspicuous overruling will embolden lower courts, and their own colleagues, to engage in potentially disastrous interpretations of complex and incompletely specified regulatory statutes—disastrous because these interpretations are uninformed, beyond the limits of judicial capacities, and amount to an arrogation to judges of powers that Congress intended to entrust primarily to the executive. Still, although it would be foolish to make predictions, one can at least see the promise of de novo deference to at least nominally resolve the deference dilemma. And that promise may make it an irresistible route of escape for a Court pinioned between large-scale, insistent, and mutually-antagonistic background conditions of legal interpretation in the administrative state.