Introduction

Overruling a prior Supreme Court precedent is a step not to be taken lightly. As the Petitioners in Loper Bright Enterprises v. Raimondo¹45 F.4th 359 (D.C. Cir. 2022), cert. granted in part, 143 S. Ct. 2429 (argued Jan. 17, 2024) (mem.). and Relentless v. Department of Commerce²62 F.4th 621 (1st Cir. 2023), cert. granted in part, 144 S. Ct. 325 (argued Jan. 17, 2024) (mem.). have expressly asked the Supreme Court to “overrule”³Petition for a Writ of Certiorari at i, Loper Bright, 143 S. Ct. 2429 (No. 22-451); Petition for a Writ of Certiorari at i, Relentless, 144 S. Ct. 325 (No. 22-1219). Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc.,⁴467 U.S. 837 (1984). it is worth taking a careful look at the Chevron litigation to understand precisely what the Chevron Court did and did not decide. Importantly, not one party or amicus argued to the Court in Chevron that the first sentence of section 706 of the Administrative Procedure Act (“APA”)—which requires courts to “decide all questions of law”—had any relevance to the case. Indeed, other significant and relevant provisions of the APA were also left entirely unbriefed by the parties and received, at best, only passing mentions by amici.⁵See, e.g., Brief for the Adm’r of the Env’t Prot. Agency at iv-vi, Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984) (No. 82-1005) [hereinafter EPA Brief in Chevron] (setting forth the table of authorities that lists all statutes cited in the government’s brief with no provisions of the Administrative Procedure Act listed). Those failures in the briefing explain the glaring omission in the Chevron opinion of any mention of section 706 or any other section of the APA.

The silence of the Chevron briefs and the ultimate opinion on the APA leads to one concrete point about the case: In terms of stare decisis, Chevron maintains no authority on the meaning of the provisions of the APA that govern judicial review. As the Supreme Court held nearly a century ago: “Questions which merely lurk in the record, neither brought to the attention of the court nor ruled upon, are not to be considered as having been so decided as to constitute precedents.”⁶Webster v. Fall, 266 U.S. 507, 511 (1925). Thus, Chevron does not need to be overruled. Rather, the opinion’s reasoning needs to be disavowed as uninformed by any argumentation or citation of the relevant legal authorities. The strategic choices by the parties about how to brief a case should not cast a permanent shadow over administrative law.

Part I shows the proper analysis of the legal issues in Chevron if the Court and the parties had paid attention to the APA and other relevant statutes. Such a statutory approach reveals multiple flaws in the Chevron analysis, beginning with, but not ending with, the complete failure to discuss the first sentence of section 706. The Court also made unnecessary conjectures about the possibility of “implicit” delegations of power even though the agency had relied on an explicit statutory delegation.⁷See Chevron, 467 U.S. at 844. The Court looked to agency “expertise” and “political” accountability as reasons to infer agency power,⁸Id. at 865. even though other statutes show those reasons are unreliable predictors of congressional delegations to administrative agencies.⁹For example, some agencies with political accountability and expertise (such as the Patent and Trademark Office) have not been given broad delegations of power by Congress. See, e.g., 35 U.S.C. § 2(b)(2)(A) (authorizing the Patent and Trademark Office to promulgate regulations limited to “govern[ing] conduct of proceedings in the Office”). These agencies should not get the level of power that the EPA was afforded in Chevron, and perhaps surprisingly, the courts have never afforded these agencies that degree of power even though the reasoning of Chevron would seem to justify it. See Merck v. Kessler, 80 F. 3d 1543, 1549–50 (Fed. Cir. 1996) (holding that the agency’s rulemaking power “does not grant the Commissioner the authority to issue substantive rules”). And finally, the Court confused the agency’s regulatory definition of “stationary source”—which imposed a “plantwide” or “bubble” concept¹⁰See Chevron, 467 U.S. at 840 (framing the issue in the case as whether a “plantwide definition of the term ‘stationary source’” that “treat[s] all all of the pollution-emitting devices within the same industrial grouping as though they were encased within a single ‘bubble’ is based on a reasonable construction of the statutory term ‘stationary source’”). Elsewhere in its opinion, the Chevron Court expressly stated its view that “the EPA’s definition of the term ‘source’ is a permissible construction of the statute.” Id. at 866.—and the agency’s actual statutory interpretation of the term “stationary source”—which was that the term had no “clear-cut” definition and thus was left for the agency to fill in via a reasonable regulation.¹¹Requirements for Preparation, Adoption, and Submittal of Implementation Plans, 46 Fed. Reg. 16280, 16281 (proposed March 12, 1981) (to be codified at 40 C.F.R. pts. 51–52).

Part II then introduces a rigorous, statutory approach to the central issue in cases like Chevron and the many other cases applying, or attempting to apply, the Chevron framework. Such a statutory approach is straightforward. Its initial step requires courts to examine and to decide de novo, as required by APA section 706, the extent of an agency’s statutory delegations of power. Under the next step, a reviewing court would decide whether the agency has stayed within that scope of delegated power. Finally, if the agency has stayed within its statutory delegations of power, the courts would apply the APA’s statutory arbitrary and capricious test (from section 706(2)(a)) to determine whether the agency had reached a reasonable result through reasoned decisionmaking.

Such a rigorous statutory approach focuses more attention on the actual delegations of power in statutory law and thus accounts for a wider variety of circumstances than the judge-made Chevron framework. For example, some agencies have even more power than the EPA did in Chevron; they have the power to write valid rules and regulations that override clear statutory text.¹²See, e.g., 47 U.S.C. § 203(b)(2) (authorizing the FCC to “modify any requirement made by . . . this [statutory] section”); 26 U.S.C. § 163(i)(5)(A) (authorizing the Secretary of the Treasury to issue “regulations providing for modifications to the provisions of this subsection [of the Internal Revenue Code] . . . where such modifications are appropriate to carry out the purposes of this subsection”); 20 U.S.C. § 1098bb(a)(1) (authorizing the Secretary of Education to “waive or modify any statutory or regulatory provision applicable to the [programs covered by] the Act as the Secretary deems necessary”). Unlike the Chevron framework, a statutory approach recognizes that agencies possessing such “super-rulemaking” powers might have the power to change law even where “Congress has directly spoken to the precise question at issue.”¹³Chevron, 467 U.S. at 842. Also contrary to the reasoning of Chevron, agencies having such super-rulemaking powers include multimember, politically balanced agencies like the Federal Communications Commission (“FCC”)—that is, agencies that have somewhat less democratic accountability than the EPA.¹⁴See supra note 12 for the FCC’s power to “modify” statutory law; see also 47 U.S.C. § 154(b)(5) (imposing a limit on the number of FCC Commissioners who can be “members of the same political party”); id. § 154(c)(1)(A) (granting FCC Commissioners tenure “for a term of 5 years”). Here again, political accountability and expertise are simply not good proxies for the scope of delegated powers.

Finally, Part III of this Article addresses the proper approach to judicial review, as authorized under the APA, of the agency action in the Loper Bright and Relentless cases. The analysis begins, as it always should under section 706, with the reviewing court deciding all questions of law without deference to what might be shifting agency positions now or in the future. The relevant statutory scheme delegates some power, but the delegation is divided and unusual—with most of the discretionary power lodged in regional councils consisting of federal inferior officers and State officials. Under the statute, the Secretary of Commerce (or her delegee) has no general rulemaking power. Rather, the very first section of relevant statutory provisions, which concern the promulgation of fishery management plans, confers on the Secretary power merely to promulgate “guidelines” that Congress expressly stated “shall not have the force and effect of law.”¹⁵See 16 U.S.C. § 1851(b).

The regional councils develop the fishery management plans; the Secretary is then required to publish each council’s proposals as proposed rules for public comment if the council’s proposals are “consistent” with applicable law.¹⁶16 U.S.C. § 1854(b)(1). See also 18 U.S.C. § 1854(a)(1)(A) (limiting the Secretary’s power in reviewing any fishery management plan to an inquiry into whether the plan is “consistent with” the statute and other applicable law). In such a statutory structure, it would be extraordinarily odd to read the regional councils’ powers as broad as the general rulemaking power held by the Administrator of the EPA under the statutory scheme in Chevron. Such a reading would, among other things, read into the powers of the regional councils the very kind of broad power expressly denied to the Secretary in the first section of the statute. The delegated powers in the statute, mainly lodged with inferior officers and circumscribed by an extensive list of the statutory mechanisms for mandatory or permissive inclusion, are too limited to support administrative regulations that impose large, non-statutory funding obligations on industry.¹⁷See infra Part III for a detailed discussion.

Yet whatever the correct outcome in the Loper Bright and Relentless cases, the most important reform the Court can and should accomplish is to require reviewing courts to pay more attention to the text of statutory law in deciding whether the agency posseses the scope of delegated power it claims in the case. The scope of delegation can then be decided without any inquiry into whether the agency has some sort of “Chevron” power that is ungrounded in any statutory provision.

I.      A De Novo Review of Chevron

Any discussion of whether the Supreme Court should abandon the Chevron doctrine should begin with a fresh look—a de novo review—of the original case, Chevron v. NRDC. Such de novo review is critical because while the result of the case seems clearly correct, the reasoning in the Court’s opinion is deeply controversial.

A fresh look at the Chevron case, aided somewhat by subsequent history, reveals at least four major problems with the Court’s opinion and the doctrine it launched. First, and most importantly, the opinion nowhere discusses or even bothers to cite 5 U.S.C. § 706. The omission is both unforgivable and yet, paradoxically, understandable. It’s unforgivable because the Court itself had long ago identified the APA as a “comprehensive” statute designed to “introduce greater uniformity of procedure and standardization of administrative practice.”¹⁸Wong Yang Sung v. McGrath, 339 U.S. 33, 36, 41 (1950). Any Supreme Court opinion seeking to define the scope of judicial review of agency action should start with section 706, which is, after all, entitled “Scope of review.”¹⁹5 U.S.C. § 706. Some scholarship suggests that the Chevron case itself was not subject to section 706 because subsection (d) of section 307 of the Clean Air Act (42 U.S.C. § 7607(d)) contains an explicit exemption making section 706 inapplicable. See, e.g., Jack M. Beermann, Chevron Is a Rorschach Test Ink Blot, 32 J.L. & Pol. 305, 307–08, 307 n.8 (2017) (stating that “[j]udicial review of EPA rules under the Clean Air Act is governed by 42 U.S.C. § 7607(d) and not by APA § 706”). The relevant provision, however, exempts from section 706 only EPA actions listed in section 307(d)(1), which at the relevant time included fourteen categories of EPA actions, none of which covered the EPA rulemaking being reviewed in the Chevron litigation. See 42 U.S.C. § 7607(d)(1)(A)–(N) (1982). In fact, at the time it was promulgating the final rule challenged in Chevron, the EPA itself analyzed the applicability of subsection (d) of section 307 and correctly determined that rules being promulgated were “not subject to Section 307(d).” Requirements for Preparation, Adoption and Submittal of Implementation Plans, 46 Fed. Reg. 50766, 50770 (Oct. 14, 1981) (codified at 40 C.F.R. pts. 51–52).

Yet, the omission is also understandable, given that none of the parties even bothered to cite section 706, and the sole amicus to cite the statute did not refer at all to the requirement in the statute’s first section that reviewing courts “shall decide all relevant questions of law.”²⁰5 U.S.C. § 706 (first sentence). In total, twelve briefs were filed in the Chevron litigation: (i) three opening briefs for the petitioners in the consolidated cases (Chevron; the American Iron and Steel Institute and others; and EPA Administrator William Ruckelshaus); (ii) the brief for the respondents (NRDC and others); (iii) three reply briefs for the petitioners; and (iv) five amicus briefs (filed by Mid-Atlantic Legal Foundation, the Pacific Legal Foundation, the American Gas Association, the United Steelworkers of America, and Pennsylvania along with other states). All those briefs were reviewed for this article. The lone brief mentioning section 706, which was filed by the amicus Mid-Atlantic Legal Foundation, referred to the statute only twice. The first mention cited section 706 in support of the legally inaccurate proposition that “in the absence of a conflict between the policy judgment of an agency administrator charged with responsibility for implementing a federal statute and either the language or purpose of the statute, the agency decision must be sustained.” Brief of the Mid-Atlantic Legal Foundation at 14–15, Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984) (No. 82-1005). (That statement ignores the requirement in section 706 that reviewing courts must hold unlawful and set aside arbitrary and capricious policy judgements by agencies even without a conflict with statutory law.) The second mention quoted the “arbitrary” and “capricious” standard in section 706(2)(A) and argued, with greater accuracy than the quote above, that the EPA’s policy decisions could “be set aside only if the Administrator ‘exceeded his statutory authority or if the regulation is ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’” Id. at 20 (quoting Batterton v. Francis, 432 U.S. 416, 426 (1977) (quoting § 706(2)(A))). All twelve briefs filed in the litigation can be found In the ProQuest Supreme Court Insight electronic library. See https://supremecourt.proquest.com/supremecourt/search/basicsearch (search for the three docket numbers associated with the consolidated cases: 82-1005, representing Chevron’s petition for certiorari; 82-1247, representing the petition filed by the American Iron and Steel Institute and others; and 82-1591, representing the petition filed by the EPA Administrator). Eleven of the twelve briefs in the consolidated cases appear in each of the three dockets, but the amicus brief of the American Gas Association was filed only in docket no. 82-1591 of the consolidated cases. See Motion for Leave to File Brief Amicus Curiae and Brief Amicus Curiae of the American Gas Ass’n, Ruckelshaus v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984) (No. 82-1591). The Anglo-American adjudication system rests on adversarial presentations of the law. Where the adversaries fail to cite the relevant law, judges do not feel obliged to do original legal research into non-jurisdictional issues. Indeed, many judges may think that such sua sponte inquiry into sources not cited by the parties is generally inappropriate.²¹Henderson v. Shinseki, 562 U.S. 428, 434 (2011) (noting that, under “our adversarial system[,] . . . [c]ourts are generally limited to addressing the claims and arguments advanced by the parties”). Nevertheless, the Chevron Court could have, in its discretion, addressed the relevance of section 706 because, “[w]hen an issue or claim is properly before the court, the court is not limited to the particular legal theories advanced by the parties, but rather retains the independent power to identify and apply the proper construction of governing law.” Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991). Yet it did not take that step. Yet, even if the omission of any discussion of section 706 is somewhat understandable, it is certainly a mark against providing the opinion strong stare decisis weight, let alone maintaining it as a cornerstone of federal administrative law.

Second, the Chevron opinion incorrectly embraces the “implicit” delegation theory.²²Chevron, 467 U.S. at 844. This flaw may be slightly less fundamental than the omission of any discussion of section 706, but it is much less understandable. In its merits brief, the United States specifically pointed to the EPA’s rulemaking authority in section 301(a)(1) of the Clean Air Act and quoted that statute prominently in the first section of its brief.²³EPA Brief in Chevron, supra note 5, at 21 (quoting 42 U.S.C. § 7601(a)(1)). Furthermore, the EPA itself cited its rulemaking authority in its orders proposing and promulgating the challenged rules.²⁴Requirements for Preparation, Adoption, and Submittal of Implementation Plans, 46 Fed. Reg. 16280, 16282 (proposed Mar. 12, 1981) (to be codified at 40 C.F.R. pts. 51–52); Requirements for Preparation, Adoption, and Submittal of Implementation Plans 46 Fed. Reg. 50766, 50771 (Oct. 14, 1981) (codified at 40 C.F.R. pts. 51–52). Citing the legal authority for the basis of proposed rules is required by section 553 of the APA,²⁵5 U.S.C. § 553(b)(2) (requiring “reference to the legal authority under which the rule is proposed”). so the Court had no excuse for not knowing that an explicit delegation in the Clean Air Act gave the EPA the power to promulgate the relevant rules. The entire discussion of implicit delegation is nothing more than misleading dicta.

Third, the Supreme Court improperly invoked “expertise” and “political” accountability as reasons for granting deference to agencies.²⁶Chevron, 467 U.S. at 865. Expertise and political accountability are reasons why Congress may want to grant significant policymaking powers to an agency. But there are counterbalancing considerations. Congress may fear the executive branch will overly politicize particular issues. Or Congress may want a resolution of a legal issue to be stable over time, which is more likely if an issue is decided by an institution that follows stare decisis and does not dramatically change perspectives based on the outcome of elections. To determine whether Congress wanted more power in an agency or in the courts, the courts must read the delegations in the statute.

Subsequent case law demonstrates the Supreme Court has not followed those policy rationales in granting or refusing Chevron deference. The Equal Employment Opportunity Commission (“EEOC”) and the Patent and Trademark Office (“PTO”) are politically accountable and have great expertise in their fields, but they generally do not get Chevron deference.²⁷EEOC v. Arabian Am. Oil. Co., 499 U.S. 244, 257–58 (1991) (limiting the EEOC to mere “Skidmore” deference because Congress did not grant the agency any substantive rulemaking power); Merck v. Kessler, 80 F.3d 1543, 1549–50 (Fed. Cir. 1996) (same for the PTO). Or, more accurately, they don’t always get Chevron deference. Lawyers must read the agencies’ statutes carefully to find out when they have sufficiently broad delegations of power to trigger Chevron deference and when they do not.²⁸Thus, for example, the EEOC does now get Chevron deference with respect to its substantive regulations promulgated under the Americans with Disabilities Act (“ADA”) of 1990 because that statute does include a delegation of substantive rulemaking authority. See 42 U.S.C. § 12116; see also Jacobs v. N.C. Admin. Off. of the Cts., 780 F.3d 562, 572–73 (4th Cir. 2015) (granting Chevron deference to an EEOC regulation defining the concept of “disability” for purposes of the ADA). Indeed, Congress went out of its way to enact an additional statute in 2008 clarify that the EEOC did have authority “to issue regulations implementing the definitions of disability” in the ADA and that statute’s amendments. See ADA Amendments Act of 2008, 122 Stat. 3553, 3558 (codified at 42 U.S.C. § 12205a). So too, in United States v. Mead,²⁹533 U.S. 218 (2001). the Customs Service did not get Chevron deference but not because it lacked expertise or political accountability—the agency had both.³⁰The Customs Service was, at the time, a component of the Treasury Department—an executive department fully accountable to the President—and had been given its administrative powers to issue “ruling letters” by a statutorily authorized delegation from the Secretary of the Treasury. See id. at 222 (tracing the powers of the Custom Service to delegations from the Secretary). The agency also had significant expertise as the Court itself noted in the portion of its opinion holding that the agency’s views may be entitled to some Skidmore deference. See id. at 234–35. Rather, Chevron deference was denied because the relevant substantive rulemaking powers granted were not invoked and the procedures required for rulemaking were not followed.³¹Id. at 231–33. The Secretary of the Treasury, the supervising department head for the Customs Service, had rulemaking powers over customs duties under both 19 U.S.C. § 1502(a) and § 1624. See id. at 222 (noting the Secretary’s two rulemaking powers). The Secretary could have resolved the issue in Mead by exercising those rulemaking powers himself or perhaps by delegating a rulemaking power to the Customs Service. Yet neither rulemaking power was invoked nor were the rulemaking procedures of the APA followed in issuing the Customs Service’s ruling letters.

Fourth, the Chevron opinion’s overarching focus on interpretation should have seemed terribly wrong to the Court even in 1984, and subsequent decisions have demonstrated the fallacy of explaining the Chevron doctrine in terms of an agency’s interpretive superiority rather than an agency’s exercise of statutorily delegated lawmaking power.³²See, e.g., id. at 234. To see this, begin with the EPA’s challenged regulation, which reads as follows:

i. “Stationary source” means any building, structure, facility, or installation which emits or may emit any air pollutant subject to regulation under the Act.

ii. “Building, structure, or facility, or installation” means all of the pollutant-emitting activities which belong to the same industrial grouping, are located on one or more contiguous or adjacent properties, and are under the control of the same person (or persons under common control). Pollutant-emitting activities shall be considered as part of the same industrial grouping if they belong to the same “Major Group” (i.e., which have the same two-digit code) as described in the Standard Industrial Classification Manual, 1972, as amended by the 1977 Supplement (U.S. Government Printing Office stock numbers 4101-0066 and 003-005-00176-0, respectively).³³40 C.F.R. § 51.18(j)(1)(i)–(ii) (1983), reproduced in EPA Brief in Chevron, supra note 5, at 1a.

For multiple reasons, the EPA could not claim (and did not claim) that it was merely interpreting the statutory term “stationary source.”³⁴The EPA’s brief at the Supreme Court described the agency as interpreting the Clean Air Act as merely permitting a “plantwide” definition of “stationary source.” EPA Brief in Chevron, supra note 5, at 45 n.57. To adopt that permissible definition as the definition for regulatory purposes, the EPA had to invoke its substantive rulemaking power. Consider initially the sheer complexity of the definition. It relies on numerous distinctions that are not even hinted at in the two-word phrase “stationary source.” The definition distinguishes between (1) facilities owned by one party versus sources owned by different parties; (2) facilities on “contiguous or adjacent” properties versus facilities that are merely near each other; and most dramatically, (3) facilities within one industrial classification and facilities within multiple classifications, applying a particular set of industrial classifications. Such distinctions might be part of a wise way to define the concept of stationary sources for purposes of the Clean Air Act. Still, it is hard to imagine all these complex distinctions and their interrelationships can somehow be teased out of the two-word statutory phrase “stationary sources.”³⁵42 U.S.C. § 7502(b)(6) (1982) (setting forth § 172(b)(6) of the Clean Air Act as amended by the Clean Air Act Amendments of 1977). See also Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 849–50, 849 n.22 (citing and quoting 42 U.S.C. § 7502(b)(6)). The statutory term is plural, though the Chevron Court often quotes it as singular. See id. at 851..

Furthermore, in its rulemaking, the EPA never claimed the complexities in the rule were merely interpretations of the language Congress wrote. Rather, the Agency explicitly used its notice-and-comment substantive rulemaking power, which is the power to make law, not just to find the law through careful parsing of statutory language. Indeed, the Agency explicitly said the proper “definition of ‘source’ is not a clear-cut legal question” because the “statute does not provide an explicit answer, nor is the issue squarely addressed in the legislative history.”³⁶Requirements for Preparation, Adoption, and Submittal of Implementation Plans, 46 Fed. Reg. 16280, 16281 (proposed March 12, 1981) (to be codified at 40 C.F.R. pts. 51–52). “The question,” the Agency correctly noted, “thus involves a judgment as to how to best carry out the Act.”³⁷Id. If the Agency had claimed at the Supreme Court that it was merely engaging in statutory interpretation (and it did not do that), then the agency would have had a serious problem with SEC v. Chenery Corporation³⁸318 U.S. 80 (1943). because it would be varying the basis of the administrative decision made below.³⁹Id. at 95 (holding that “an administrative order cannot be upheld unless the grounds upon which the agency acted in exercising its powers were those upon which its action can be sustained”). In sum, the agency’s actual statutory interpretation of “stationary source” was not the “bubble” definition in the regulation; it was instead that the term was an empty vessel that the Agency was given the power to fill with “a judgment as to how to best carry out the Act.”⁴⁰See id.; see also Requirements for Preparation, Adoption, and Submittal of Implementation Plans, supra note 36.

Indeed, if the Agency had said that its stationary source definition was merely an interpretive rule reflecting the statute’s meaning, it almost certainly would have lost the case for two reasons. First, the challenged rule was modifying a prior substantive rule, and only another substantive rule can change a prior substantive rule.⁴¹See Am. Mining Cong. v. Mine Safety & Health Admin., 995 F.2d 1106, 1109 (D.C. Cir. 1993) (collecting caselaw holding that interpretive rules cannot repudiate or change prior substantive rules). Second, even if the EPA’s plantwide definition had been the first interpretation of “stationary source” and, therefore, the agency did not have to modify a prior rule, it seems extraordinarily unlikely that a court would allow the complex, nuanced rule to be promulgated without notice and comment as a mere interpretive rule. An interpretive rule “derive[s] a proposition from an existing document, . . . whose meaning compels or logically justifies the proposition.”⁴²See, e.g., Nat. Res. Def. Council v. Wheeler, 955 F.3d 68, 83 (D.C. Cir. 2020) (internal quotations and citations omitted). Nothing in the statutory language of the Clean Air Act compelled or logically justified the precise set of nuances included in the agency’s definitions. Those nuances, however reasonable, were the product of the agency’s filling in the statute’s details through delegated lawmaking powers, not merely interpreting the statute’s words to find the best meaning.

The final reason the Chevron doctrine should not be viewed as being about interpretation is that the Supreme Court’s subsequent decision in United States v. Mead has since squarely held that interpretive rules “as a class” do not get Chevron deference.⁴³United States v. Mead Corp., 533 U.S. 215, 232 (2001). As Justice Antonin Scalia said in dissent, that ruling was very much an “avulsive change” to the Chevron doctrine.⁴⁴Id. at 239 (Scalia, J., dissenting). After Mead, the “Chevron–Mead doctrine” is clearly about the statutory delegation of law-making powers, not deference to any superior interpretive prowess of agencies over courts.⁴⁵See id. at 226–27, 229 (majority opinion) (requiring, as a precondition for Chevron deference, that Congress have delegated to the agency authority “to make rules carrying the force of law” or otherwise “to speak with the force of law”).

II.      A Rigorous Statutory Approach vs. the Chevron–Mead Doctrine

After Mead’s dramatic reinterpretation of Chevron, the resulting Chevron–Mead doctrine now bears some resemblance to a rigorous statutory approach, but it is worthwhile defining that statutory approach in detail and comparing it to the Chevron–Mead framework. One initial point, however: The similarity between a statutory approach and the Chevron–Mead framework is not a good reason to try to save the existing framework, wholly or partially. To the contrary, the similarity between the two approaches merely confuses lawyers and courts with pointless questions such as (1) whether the Chevron–Mead framework can somehow be reconciled with section 706’s command that courts decide all questions of law and (2) whether the second step of Chevron is the same or slightly different than the arbitrary-and-capricious analysis required by section 706(2)(A). The maddening similarity between a statutory approach and the Chevron–Mead framework distracts lawyers and judges from the most important inquiry: the precise scope and limits of the powers that Congress delegated to the agency.

Like the Chevron–Mead framework, a rigorous statutory approach has three steps. In the first step, courts would identify, de novo, the precise scope of the powers delegated to the agency. This step can be justified not merely by judicial decisions such as Chevron and Mead but by fundamental principles of administrative law and the APA. Because the U.S. Constitution creates no agencies, agencies are wholly creatures of statute.⁴⁶See, e.g., Nat’l Fed’n of Indep. Bus. v. OSHA, 595 U.S. 109, 117 (2022) (“Administrative agencies are creatures of statute.”). For agencies to do anything, they must find authority in statutory law, and that point is memorialized in section 706(2)(C) of the APA, which requires courts to reverse agency actions that are “in excess of statutory jurisdiction, authority, or limitations, or short of statutory right.”⁴⁷5 U.S.C. § 706(2)(C). And of course, under a statutory approach, the first sentence of section 706 requires reviewing courts to “decide” the scope of the agency’s delegated powers, not merely to accept one reasonable position advanced by an agency in one administration and then to accept a different reasonable position advanced by the agency under a different administration. The command for courts to “decide” legal issues means an actual decision that embraces a position and affords that position the weight of stare decisis.

That first step in a statutory approach may sound similar to the first step (often called “step zero”⁴⁸Cass R. Sunstein, Chevron Step Zero, 92 Va. L. Rev. 187, 191 (2013).) in the Chevron–Mead framework, which requires courts to decide, as a prerequisite to granting deference, whether the agency has been delegated the power to speak with the force of law.⁴⁹Mead, 533 U.S. at 229 (requiring, as a precondition for affording Chevron deference, that agencies to have the power “to speak with the force of law”). That step is also an inquiry into delegated powers, and it is obviously conducted de novo because it is a precondition for granting deference. Nevertheless, there is an immensely important difference. The Chevron–Mead doctrine seems to allow only two possibilities: the agency has been delegated power to speak with the force of law, or it has not. In reality, Congress can choose from several different forms of delegated power.

A statutory approach to deciding agency power casts off the distractions of the Chevron–Mead framework and instead focuses on the basic and straightforward question of how much power Congress granted. Here are just a few of the possible answers to that question, arranged in order of greater power to lesser power:

* Congress could confer a very broad power similar to the one granted 19 U.S.C. § 3004, which authorizes the executive branch to override a list of tariffs originally enacted as statutory law and also provides that “[e]ach modification or change” in the tariff schedule “shall be considered to be statutory provisions of law for all purposes.”⁵⁰19 U.S.C. § 3004(c). The original “Harmonized Tariff Schedule” was enacted as statutory law in 1988. See Omnibus Trade and Competitiveness Act of 1988, Pub. L. No. 100-418, § 1204, 102 Stat. 1107, 1148–49.

* Congress could grant the agency “super-rulemaking” power, which allows the agency to “modify” statutory law—even clear statutory law—but does not provide that the modifications are on an equal footing with statutory law.⁵¹See MCI Telecomms. Corp. v. AT&T Co., 512 U.S. 218, 234 (1994) (holding a power to “[m]odify” statutory law permits the agency to make changes to clear text statutory text, but only on modest issues).

* Congress could grant an ordinary rulemaking power, such as the power given to the EPA in Chevron.⁵²See 42 U.S.C. § 7601(a)(1).

* Congress could also give an agency formal adjudicatory authority, which might authorize an agency to develop its own common-law precedents, filling in statutory language.

* Congress could withhold any general rulemaking power, but nonetheless—as in the Skidmore v. Swift & Co.⁵³323 U.S. 124 (1944). and Mead cases—grant the agency some administrative powers that may justify a reviewing court giving the agency’s positions some degree of weight as the court decides the correct interpretation of a statute.

* Congress could grant the agency little or no delegated power other than the power to “comply with” the statute.⁵⁴See, e.g., 5 U.S.C. § 559 (granting each agency “the authority necessary to comply with the requirements of this subchapter through the issuance of rules or otherwise”) (emphasis added). Examples of such statutes include the APA and other general statutes that treat agencies as the targets, not the tools, of regulation.

The key point in this first step of a statutory approach is that courts should consider the entire range of possible delegations and then examine the relevant statutes delegating power to decide how much power the particulate agency possesses.

The next step of a rigorous statutory approach would require courts to determine whether the agency has stayed within the scope of its delegated power. If an agency has a super-rulemaking power, as the FCC did in MCI Telecommunications Corp. v. AT&T,⁵⁵512 U.S. 218 (1994). the agency may be able to modify clear statutory text provided the change is a relatively small modification rather than a basic change in the operation of the statute.

In rulemaking and formal adjudicatory proceedings, courts may very well afford agencies the scope of power defined by the Chevron–Mead doctrine. But the justification for the power would be the scope of delegation in the statute, not the judge-made case law. A statutory approach grounded in delegation can then easily be reconciled with the first sentence of section 706, which requires reviewing courts to “decide all relevant questions of law.”⁵⁶5 U.S.C. § 706. As Chief Justice John Roberts has noted, courts are “respect[ing]” section 706’s requirements by permitting agencies to fill in statutory ambiguities if “Congress has delegated to the agency the authority” to do just that.⁵⁷City of Arlington v. FCC, 569 U.S. 290, 317 (2013) (Roberts, J., dissenting). The precondition for allowing the agency to do that, however, is Congress must have “in fact delegated to the agency lawmaking power over the ambiguity at issue.”⁵⁸Id. In cases where agencies have less power, the courts would be the ultimate arbiter for filling in statutory gaps.

For example, in cases involving the APA, courts should continue to decide those cases without giving agencies’ views any weight. That’s true even though section 559 of the APA confers a rulemaking power on each agency, but that rulemaking power—framed merely as a power to “comply with” the statute—is best interpreted not as a general delegation from Congress for each agency to make its own version of the APA. Where the agency has some specialized enforcement powers within the statute, a reviewing court would also follow the close sibling of de novo review set forth in Skidmore v. Swift. The court would take account of the agency’s specialized experience, but the agency would have to “persuade” the court that its position is the correct reading of the statute.⁵⁹Skidmore v. Swift & Co., 323 U.S. 124, 140 (1944). The degree to which the Skidmore approach is essentially identical to de novo review was emphasized by candid comments from Chief Justice John Roberts during a Supreme Court oral argument. He commented: “It’s just—maybe it’s that I’ve never understood Skidmore. To me, anyway, as it’s been articulated, it seems to be the principle as you should defer to agencies when you agree with their interpretation.” Transcript of Oral Argument at 15, Advoc. Health Care Network v. Stapleton, 581 U.S. 468 (2017) (No. 16-74). And, consistent with section 706, once an agency persuades the court to reach a particular interpretation, the court has “decide[d]” the question of law and its decision is protected by stare decisis against future changes in the agency’s position.

The final step of a rigorous statutory approach would be the arbitrary-and-capricious test of section 706(2)(a) of the APA. Here again, the confusion caused by the similarity between the Chevron–Mead framework and a rigorous statutory approach is evident. The Court of Appeals for the D.C. Circuit appears to treat the final step of the Chevron framework as identical to arbitrary-and-capricious review.⁶⁰Wedgewood Vill. Pharmacy v. Drug Enf’t Admin., 509 F.3d 541, 549 (2007) (seemingly equating the final step of Chevron analysis with section 706(2)(A)’s arbitrary-and-capricious test). The Supreme Court has said something similar but ultimately quite confusing. In Judulang v. Holder,⁶¹565 U.S. 42 (2011). the Court stated the final step of the Chevron–Mead framework was arbitrary-and-capricious review “in substance.”⁶²Id. at 52 n.7.

One charitable interpretation of Judulang’s assertion is that the final step of Chevron would reverse an agency only if the outcome were unreasonable, whereas the arbitrary-and-capricious test authorizes reversing the agency if the outcome is either unreasonable or lacks reasonable decisionmaking. The result is that the Chevron–Mead doctrine just introduces confusion in its final step. It duplicates at least part of the statutorily required arbitrary and capricious review, but arbitrary-and-capricious review includes everything in the final step of Chevron plus more (the requirement of reasoned decisionmaking, not just a reasonable outcome). The end result is that parties litigate the Chevron framework’s final step in parallel with the statutory arbitrary-and-capricious test. At best, it’s a worthless duplication; at worst, it’s a cause of confusion.

III.      Loper Bright and Relentless: Perfect Illustrations of the Value of a Statutory Approach

The advantages of a statutory approach over the Chevron–Mead framework are not merely theoretical. They are also practical. The statute at issue in the Loper Bright and Relentless, Inc. v. Department of Commerce cases perfectly illustrates what is wrong with the Chevron framework and how the statutory approach is better.

The first step of the Chevron–Mead framework (again, often called “step zero”) asks whether the agency has the authority to speak with the force of law.⁶³Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843–44 (1984). The doctrine appears to be an on-off switch: the agency either has the power or doesn’t. The choice forces courts to try to classify statutes according to judicial preconceptions about what sort of delegations Congress has bestowed. But the statute in Loper Bright and Relentless shows Congress does not always follow standard patterns.

The first section in the statute governing fishing regulations sets forth principles that regional councils must follow in setting fishing limits and regulations.⁶⁴16 U.S.C. § 1851(a). The statute includes a highly unusual restriction: it specifies that the Secretary of Commerce shall have the power to issue only “guidelines” to supplement the principles set forth in the statute.⁶⁵16 U.S.C. § 1851(b). The Secretary’s authority to issue only “guidelines” was in the original version of the statute. See Fishery Conservation and Management Act of 1976, Pub. L. No. 94-265, § 301(b), 90 Stat. 331, 347. Guidelines in administrative law typically refer to agency pronouncements that do not have the force and effect of law.⁶⁶United States v. Mead Corp., 533 U.S. 218, 234 (2001) (holding that “interpretations contained in policy statements, agency manuals, and enforcement guidelines” are “beyond the Chevron pale” (quoting Christensen v. Harris Cnty., 529 U.S. 576, 587 (2000))); see also EEOC v. Arabian Am. Oil Co., 499 U.S. 244, 257 (1991) (holding that agency “guidelines” were not to be afforded Chevron deference because Congress had not delegated to the agency the power to promulgate rules and regulations). Lest there be any doubt on that point, Congress in 1983 went out of its way to amend the statute by adding a parenthetical that explained exactly how limited the Secretary’s authority was: the parenthetical specifies that the Secretary’s guidelines “shall not have the force or effect of law.”⁶⁷Act of Jan. 12, 1983, Pub. L. No. 97-453, § 4, 96 Stat. 2481, 2484 (amending 16 U.S.C. § 1851(b) to its current form).

Subsequent sections in the statute give power to the regional councils to write “fishery management plans” that follow the “national standards” about fishing specified in the first section of the statute.⁶⁸16 U.S.C. § 1853(a) (setting forth the “[r]equired provisions” in any “fishery management plan” and requiring consistency with “the national standards”); id. § 1853(b) (setting forth the “[d]iscretionary provisions” for any such plan and also requiring consistency with “the national standards”). The statute authorizes the regional councils, in writing those fishery management plans, to propose to the Secretary regulations for “implementing a fishery management plan.” The Secretary can reject those plans only if they are inconsistent with statutory law⁶⁹16 U.S.C. § 1854(a)(1)(A) (limiting the Secretary’s review to “determin[ing] whether [the plan] is consistent with the national standards [specified in the statute], the other provisions of this chapter, and any other applicable law”). and can reject the accompanying proposed regulations only if they are inconsistent with statutory law or the fishery management plan that the regulations are supposed to be implementing.⁷⁰16 U.S.C. § 1854(b)(1) (limiting the Secretary’s review of the proposed regulations to “determin[ing] whether they are consistent with the fishery management plan, plan amendment, this chapter and other applicable law”). If a court asks the question dictated by the Chevron–Mead framework—whether the agency has the power to speak with the force of law—the correct answer is unclear. Under the first section of the statute, Congress went out of its way to specify that the Secretary does not have any power to speak with the force of law.⁷¹See 16 U.S.C. § 1851(b); see also supra notes 66 and 67 and accompanying text. Subsequent sections give some limited rulemaking powers to subordinates of the Secretary (or, more accurately, powers to propose regulations to the Secretary),⁷²See 16 U.S.C. § 1853(c). but it seems odd to interpret the statute as granting more power to subordinates of the Secretary than to the Secretary herself.

A statutory approach to the case would have to determine how much power Congress intended to delegate to the agency. The best answer seems to be that the agency has much less delegated power than an agency like the EPA, which possesses a traditional rulemaking power to speak with the force and effect of law. The agency seems more akin to the Department of Labor in Skidmore.⁷³See Skidmore v. Swift & Co., 323 U.S. 124, 139–40 (1944). The best interpretation of the statute seems to be that the regional councils in the agency have some modest discretion in setting rules governing fishing limits, but significant questions (including even medium-sized questions) concerning the implementation of the statute ought to be governed exclusively by a de novo application of the principles set forth in the first section in the statute. Those principles are not subject to administrative alteration or supplementation, even by the Secretary.

Another way of explaining the outcome in the case is to say that an agency with a traditional rulemaking power, such as the EPA, can formulate law up to the limits imposed by the major questions doctrine. By contrast, the administrative structure at issue in Loper Bright and Relentless limits the agency’s power far more than the EPA’s general powers in Chevron.⁷⁴Compare 16 U.S.C. § 1851(b) (specifying that the Secretary of Commerce shall have the power to issue only “guidelines” to supplement the principles set forth in the statute), with 42 U.S.C. § 7601(a)(1) (giving broad power to the EPA Administrator “to prescribe such regulations as are necessary to carry out his functions under this chapter”). For significant but not major questions, courts should look directly to the statutory structure imposed by Congress and should view agency statements concerning the proper reach of the statute as being mere guidance without the force and effect of law.⁷⁵More detailed analysis of the statutory sections at issue in Loper Bright and Relentless is provided in John F. Duffy, The Important Statutory Sections Ignored by the Parties in Loper Bright and Relentless, Yale J. on Regul.: Notice & Comment (Jan. 17, 2024), https://perma.cc/U4WB-QYGJ.

Conclusion

The Chevron or Chevron–Mead framework has never precisely aligned either with the APA or, more generally, with the fundamental structure of laws that grant and restrain agency powers. The Loper Bright and Relentless cases give the Supreme Court the opportunity to turn away from its common law doctrines governing the scope of review of legal questions. The Court should take the opportunity.