Loper Bright Enterprises v. Raimondo, 603 U.S. ___ (2024)

NOTICE: This opinion is subject to formal revision before publication in the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, pio@supremecourt.gov, of any typographical or other formal errors.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 22–451 and 22–1219

_________________

LOPER BRIGHT ENTERPRISES, et al., PETITIONERS

22–451v.

GINA RAIMONDO, SECRETARY OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the district of columbia circuit

RELENTLESS, INC., et al., PETITIONERS

22–1219v.

DEPARTMENT OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the first circuit

[June 28, 2024]

Chief Justice Roberts delivered the opinion of the Court.

Since our decision in Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), we have sometimes required courts to defer to “permissible” agency interpretations of the statutes those agencies administer—even when a reviewing court reads the statute differently. In these cases we consider whether that doctrine should be overruled.

I

Our Chevron doctrine requires courts to use a two-step framework to interpret statutes administered by federal agencies. After determining that a case satisfies the various preconditions we have set for Chevron to apply, a reviewing court must first assess “whether Congress has directly spoken to the precise question at issue.” Id., at 842. If, and only if, congressional intent is “clear,” that is the end of the inquiry. Ibid. But if the court determines that “the statute is silent or ambiguous with respect to the specific issue” at hand, the court must, at Chevron’s second step, defer to the agency’s interpretation if it “is based on a permissible construction of the statute.” Id., at 843. The reviewing courts in each of the cases before us applied Chevron’s framework to resolve in favor of the Government challenges to the same agency rule.

A

Before 1976, unregulated foreign vessels dominated fishing in the international waters off the U. S. coast, which began just 12 nautical miles offshore. See, e.g., S. Rep. No. 94–459, pp. 2–3 (1975). Recognizing the resultant overfishing and the need for sound management of fishery resources, Congress enacted the Magnuson-Stevens Fishery Conservation and Management Act (MSA). See 90 Stat. 331 (codified as amended at 16 U. S. C. §1801 et seq.). The MSA and subsequent amendments extended the jurisdiction of the United States to 200 nautical miles beyond the U. S. territorial sea and claimed “exclusive fishery management authority over all fish” within that area, known as the “exclusive economic zone.” §1811(a); see Presidential Proclamation No. 5030, 3 CFR 22 (1983 Comp.); §§101, 102, 90 Stat. 336. The National Marine Fisheries Service (NMFS) administers the MSA under a delegation from the Secretary of Commerce.

The MSA established eight regional fishery management councils composed of representatives from the coastal States, fishery stakeholders, and NMFS. See 16 U. S. C. §§1852(a), (b). The councils develop fishery management plans, which NMFS approves and promulgates as final regulations. See §§1852(h), 1854(a). In service of the statute’s fishery conservation and management goals, see §1851(a), the MSA requires that certain provisions—such as “a mechanism for specifying annual catch limits . . . at a level such that overfishing does not occur,” §1853(a)(15)—be included in these plans, see §1853(a). The plans may also include additional discretionary provisions. See §1853(b). For example, plans may “prohibit, limit, condition, or require the use of specified types and quantities of fishing gear, fishing vessels, or equipment,” §1853(b)(4); “reserve a portion of the allowable biological catch of the fishery for use in scientific research,” §1853(b)(11); and “prescribe such other measures, requirements, or conditions and restrictions as are determined to be necessary and appropriate for the conservation and management of the fishery,” §1853(b)(14).

Relevant here, a plan may also require that “one or more observers be carried on board” domestic vessels “for the purpose of collecting data necessary for the conservation and management of the fishery.” §1853(b)(8). The MSA specifies three groups that must cover costs associated with observers: (1) foreign fishing vessels operating within the exclusive economic zone (which must carry observers), see §§1821(h)(1)(A), (h)(4), (h)(6); (2) vessels participating in certain limited access privilege programs, which impose quotas permitting fishermen to harvest only specific quantities of a fishery’s total allowable catch, see §§1802(26), 1853a(c)(1)(H), (e)(2), 1854(d)(2); and (3) vessels within the jurisdiction of the North Pacific Council, where many of the largest and most successful commercial fishing enterprises in the Nation operate, see §1862(a). In the latter two cases, the MSA expressly caps the relevant fees at two or three percent of the value of fish harvested on the vessels. See §§1854(d)(2)(B), 1862(b)(2)(E). And in general, it authorizes the Secretary to impose “sanctions” when “any payment required for observer services provided to or contracted by an owner or operator . . . has not been paid.” §1858(g)(1)(D).

The MSA does not contain similar terms addressing whether Atlantic herring fishermen may be required to bear costs associated with any observers a plan may mandate. And at one point, NMFS fully funded the observer coverage the New England Fishery Management Council required in its plan for the Atlantic herring fishery. See 79 Fed. Reg. 8792 (2014). In 2013, however, the council proposed amending its fishery management plans to empower it to require fishermen to pay for observers if federal funding became unavailable. Several years later, NMFS promulgated a rule approving the amendment. See 85 Fed. Reg. 7414 (2020).

With respect to the Atlantic herring fishery, the Rule created an industry funded program that aims to ensure observer coverage on 50 percent of trips undertaken by vessels with certain types of permits. Under that program, vessel representatives must “declare into” a fishery before beginning a trip by notifying NMFS of the trip and announcing the species the vessel intends to harvest. If NMFS determines that an observer is required, but declines to assign a Government-paid one, the vessel must contract with and pay for a Government-certified third-party observer. NMFS estimated that the cost of such an observer would be up to $710 per day, reducing annual returns to the vessel owner by up to 20 percent. See id., at 7417–7418.

B

Petitioners Loper Bright Enterprises, Inc., H&L Axelsson, Inc., Lund Marr Trawlers LLC, and Scombrus One LLC are family businesses that operate in the Atlantic herring fishery. In February 2020, they challenged the Rule under the MSA, 16 U. S. C. §1855(f ), which incorporates the Administrative Procedure Act (APA), 5 U. S. C. §551 et seq. In relevant part, they argued that the MSA does not authorize NMFS to mandate that they pay for observers required by a fishery management plan. The District Court granted summary judgment to the Government. It concluded that the MSA authorized the Rule, but noted that even if these petitioners’ “arguments were enough to raise an ambiguity in the statutory text,” deference to the agency’s interpretation would be warranted under Chevron. 544 F. Supp. 3d 82, 107 (DC 2021); see id., at 103–107.

A divided panel of the D. C. Circuit affirmed. See 45 F. 4th 359 (2022). The majority addressed various provisions of the MSA and concluded that it was not “wholly unambiguous” whether NMFS may require Atlantic herring fishermen to pay for observers. Id., at 366. Because there remained “some question” as to Congress’s intent, id., at 369, the court proceeded to Chevron’s second step and deferred to the agency’s interpretation as a “reasonable” construction of the MSA, 45 F. 4th, at 370. In dissent, Judge Walker concluded that Congress’s silence on industry funded observers for the Atlantic herring fishery—coupled with the express provision for such observers in other fisheries and on foreign vessels—unambiguously indicated that NMFS lacked the authority to “require [Atlantic herring] fishermen to pay the wages of at-sea monitors.” Id., at 375.

C

Petitioners Relentless Inc., Huntress Inc., and Seafreeze Fleet LLC own two vessels that operate in the Atlantic herring fishery: the F/V Relentless and the F/V Persistence.[1] These vessels use small-mesh bottom-trawl gear and can freeze fish at sea, so they can catch more species of fish and take longer trips than other vessels (about 10 to 14 days, as opposed to the more typical 2 to 4). As a result, they generally declare into multiple fisheries per trip so they can catch whatever the ocean offers up. If the vessels declare into the Atlantic herring fishery for a particular trip, they must carry an observer for that trip if NMFS selects the trip for coverage, even if they end up harvesting fewer herring than other vessels—or no herring at all.

This set of petitioners, like those in the D. C. Circuit case, filed a suit challenging the Rule as unauthorized by the MSA. The District Court, like the D. C. Circuit, deferred to NMFS’s contrary interpretation under Chevron and thus granted summary judgment to the Government. See 561 F. Supp. 3d 226, 234–238 (RI 2021).

The First Circuit affirmed. See 62 F. 4th 621 (2023). It relied on a “default norm” that regulated entities must bear compliance costs, as well as the MSA’s sanctions provision, Section 1858(g)(1)(D). See id., at 629–631. And it rejected petitioners’ argument that the express statutory authorization of three industry funding programs demonstrated that NMFS lacked the broad implicit authority it asserted to impose such a program for the Atlantic herring fishery. See id., at 631–633. The court ultimately concluded that the “[a]gency’s interpretation of its authority to require at-sea monitors who are paid for by owners of regulated vessels does not ‘exceed[ ] the bounds of the permissible.’ ” Id., at 633–634 (quoting Barnhart v. Walton, 535 U.S. 212, 218 (2002); alteration in original). In reaching that conclusion, the First Circuit stated that it was applying Chevron’s two-step framework. 62 F. 4th, at 628. But it did not explain which aspects of its analysis were relevant to which of Chevron’s two steps. Similarly, it declined to decide whether the result was “a product of Chevron step one or step two.” Id., at 634.

We granted certiorari in both cases, limited to the question whether Chevron should be overruled or clarified. See 601 U. S. ___ (2023); 598 U. S. ___ (2023).[2]

II

A

Article III of the Constitution assigns to the Federal Judiciary the responsibility and power to adjudicate “Cases” and “Controversies”—concrete disputes with consequences for the parties involved. The Framers appreciated that the laws judges would necessarily apply in resolving those disputes would not always be clear. Cognizant of the limits of human language and foresight, they anticipated that “[a]ll new laws, though penned with the greatest technical skill, and passed on the fullest and most mature deliberation,” would be “more or less obscure and equivocal, until their meaning” was settled “by a series of particular discussions and adjudications.” The Federalist No. 37, p. 236 (J. Cooke ed. 1961) (J. Madison).

The Framers also envisioned that the final “interpretation of the laws” would be “the proper and peculiar province of the courts.” Id., No. 78, at 525 (A. Hamilton). Unlike the political branches, the courts would by design exercise “neither Force nor Will, but merely judgment.” Id., at 523. To ensure the “steady, upright and impartial administration of the laws,” the Framers structured the Constitution to allow judges to exercise that judgment independent of influence from the political branches. Id., at 522; see id., at 522–524; Stern v. Marshall, 564 U.S. 462, 484 (2011).

This Court embraced the Framers’ understanding of the judicial function early on. In the foundational decision of Marbury v. Madison, Chief Justice Marshall famously declared that “[i]t is emphatically the province and duty of the judicial department to say what the law is.” 1 Cranch 137, 177 (1803). And in the following decades, the Court understood “interpret[ing] the laws, in the last resort,” to be a “solemn duty” of the Judiciary. United States v. Dickson, 15 Pet. 141, 162 (1841) (Story, J., for the Court). When the meaning of a statute was at issue, the judicial role was to “interpret the act of Congress, in order to ascertain the rights of the parties.” Decatur v. Paulding, 14 Pet. 497, 515 (1840).

The Court also recognized from the outset, though, that exercising independent judgment often included according due respect to Executive Branch interpretations of federal statutes. For example, in Edwards’ Lessee v. Darby, 12 Wheat. 206 (1827), the Court explained that “[i]n the construction of a doubtful and ambiguous law, the contemporaneous construction of those who were called upon to act under the law, and were appointed to carry its provisions into effect, is entitled to very great respect.” Id., at 210; see also United States v. Vowell, 5 Cranch 368, 372 (1809) (Marshall, C. J., for the Court).

Such respect was thought especially warranted when an Executive Branch interpretation was issued roughly contemporaneously with enactment of the statute and remained consistent over time. See Dickson, 15 Pet., at 161; United States v. Alabama Great Southern R. Co., 142 U.S. 615, 621 (1892); National Lead Co. v. United States, 252 U.S. 140, 145–146 (1920). That is because “the longstanding ‘practice of the government’ ”—like any other interpretive aid—“can inform [a court’s] determination of ‘what the law is.’ ” NLRB v. Noel Canning, 573 U.S. 513, 525 (2014) (first quoting McCulloch v. Maryland, 4 Wheat. 316, 401 (1819); then quoting Marbury, 1 Cranch, at 177). The Court also gave “the most respectful consideration” to Executive Branch interpretations simply because “[t]he officers concerned [were] usually able men, and masters of the subject,” who were “[n]ot unfrequently . . . the draftsmen of the laws they [were] afterwards called upon to interpret.” United States v. Moore, 95 U.S. 760, 763 (1878); see also Jacobs v. Prichard, 223 U.S. 200, 214 (1912).

“Respect,” though, was just that. The views of the Executive Branch could inform the judgment of the Judiciary, but did not supersede it. Whatever respect an Executive Branch interpretation was due, a judge “certainly would not be bound to adopt the construction given by the head of a department.” Decatur, 14 Pet., at 515; see also Burnet v. Chicago Portrait Co., 285 U.S. 1, 16 (1932). Otherwise, judicial judgment would not be independent at all. As Justice Story put it, “in cases where [a court’s] own judgment . . . differ[ed] from that of other high functionaries,” the court was “not at liberty to surrender, or to waive it.” Dickson, 15 Pet., at 162.

B

The New Deal ushered in a “rapid expansion of the administrative process.” United States v. Morton Salt Co., 338 U.S. 632, 644 (1950). But as new agencies with new powers proliferated, the Court continued to adhere to the traditional understanding that questions of law were for courts to decide, exercising independent judgment.

During this period, the Court often treated agency determinations of fact as binding on the courts, provided that there was “evidence to support the findings.” St. Joseph Stock Yards Co. v. United States, 298 U.S. 38, 51 (1936). “When the legislature itself acts within the broad field of legislative discretion,” the Court reasoned, “its determinations are conclusive.” Ibid. Congress could therefore “appoint[ ] an agent to act within that sphere of legislative authority” and “endow the agent with power to make findings of fact which are conclusive, provided the requirements of due process which are specially applicable to such an agency are met, as in according a fair hearing and acting upon evidence and not arbitrarily.” Ibid. (emphasis added).

But the Court did not extend similar deference to agency resolutions of questions of law. It instead made clear, repeatedly, that “[t]he interpretation of the meaning of statutes, as applied to justiciable controversies,” was “exclusively a judicial function.” United States v. American Trucking Assns., Inc., 310 U.S. 534, 544 (1940); see also Social Security Bd. v. Nierotko, 327 U.S. 358, 369 (1946); Medo Photo Supply Corp. v. NLRB, 321 U.S. 678, 681–682, n. 1 (1944). The Court understood, in the words of Justice Brandeis, that “[t]he supremacy of law demands that there shall be opportunity to have some court decide whether an erroneous rule of law was applied.” St. Joseph Stock Yards, 298 U. S., at 84 (concurring opinion). It also continued to note, as it long had, that the informed judgment of the Executive Branch—especially in the form of an interpretation issued contemporaneously with the enactment of the statute—could be entitled to “great weight.” American Trucking Assns., 310 U. S., at 549.

Perhaps most notably along those lines, in Skidmore v. Swift & Co., 323 U.S. 134 (1944), the Court explained that the “interpretations and opinions” of the relevant agency, “made in pursuance of official duty” and “based upon . . . specialized experience,” “constitute[d] a body of experience and informed judgment to which courts and litigants [could] properly resort for guidance,” even on legal questions. Id., at 139–140. “The weight of such a judgment in a particular case,” the Court observed, would “depend upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control.” Id., at 140.

On occasion, to be sure, the Court applied deferential review upon concluding that a particular statute empowered an agency to decide how a broad statutory term applied to specific facts found by the agency. For example, in Gray v. Powell, 314 U.S. 402 (1941), the Court deferred to an administrative conclusion that a coal-burning railroad that had arrangements with several coal mines was not a coal “producer” under the Bituminous Coal Act of 1937. Congress had “specifically” granted the agency the authority to make that determination. Id., at 411. The Court thus reasoned that “[w]here, as here, a determination has been left to an administrative body, this delegation will be respected and the administrative conclusion left untouched” so long as the agency’s decision constituted “a sensible exercise of judgment.” Id., at 412–413. Similarly, in NLRB v. Hearst Publications, Inc., 322 U.S. 111 (1944), the Court deferred to the determination of the National Labor Relations Board that newsboys were “employee[s]” within the meaning of the National Labor Relations Act. The Act had, in the Court’s judgment, “assigned primarily” to the Board the task of marking a “definitive limitation around the term ‘employee.’ ” Id., at 130. The Court accordingly viewed its own role as “limited” to assessing whether the Board’s determination had a “ ‘warrant in the record’ and a reasonable basis in law.” Id., at 131.

Such deferential review, though, was cabined to factbound determinations like those at issue in Gray and Hearst. Neither Gray nor Hearst purported to refashion the longstanding judicial approach to questions of law. In Gray, after deferring to the agency’s determination that a particular entity was not a “producer” of coal, the Court went on to discern, based on its own reading of the text, whether another statutory term—“other disposal” of coal—encompassed a transaction lacking a transfer of title. See 314 U. S., at 416–417. The Court evidently perceived no basis for deference to the agency with respect to that pure legal question. And in Hearst, the Court proclaimed that “[u]ndoubtedly questions of statutory interpretation . . . are for the courts to resolve, giving appropriate weight to the judgment of those whose special duty is to administer the questioned statute.” 322 U. S., at 130–131. At least with respect to questions it regarded as involving “statutory interpretation,” the Court thus did not disturb the traditional rule. It merely thought that a different approach should apply where application of a statutory term was sufficiently intertwined with the agency’s factfinding.

In any event, the Court was far from consistent in reviewing deferentially even such factbound statutory determinations. Often the Court simply interpreted and applied the statute before it. See K. Davis, Administrative Law §248, p. 893 (1951) (“The one statement that can be made with confidence about applicability of the doctrine of Gray v. Powell is that sometimes the Supreme Court applies it and sometimes it does not.”); B. Schwartz, Gray vs. Powell and the Scope of Review, 54 Mich. L. Rev. 1, 68 (1955) (noting an “embarrassingly large number of Supreme Court decisions that do not adhere to the doctrine of Gray v. Powell”). In one illustrative example, the Court rejected the U. S. Price Administrator’s determination that a particular warehouse was a “public utility” entitled to an exemption from the Administrator’s General Maximum Price Regulation. Despite the striking resemblance of that administrative determination to those that triggered deference in Gray and Hearst, the Court declined to “accept the Administrator’s view in deference to administrative construction.” Davies Warehouse Co. v. Bowles, 321 U.S. 144, 156 (1944). The Administrator’s view, the Court explained, had “hardly seasoned or broadened into a settled administrative practice,” and thus did not “overweigh the considerations” the Court had “set forth as to the proper construction of the statute.” Ibid.

Nothing in the New Deal era or before it thus resembled the deference rule the Court would begin applying decades later to all varieties of agency interpretations of statutes. Instead, just five years after Gray and two after Hearst, Congress codified the opposite rule: the traditional understanding that courts must “decide all relevant questions of law.” 5 U. S. C. §706.[3]

C

Congress in 1946 enacted the APA “as a check upon administrators whose zeal might otherwise have carried them to excesses not contemplated in legislation creating their offices.” Morton Salt, 338 U. S., at 644. It was the culmination of a “comprehensive rethinking of the place of administrative agencies in a regime of separate and divided powers.” Bowen v. Michigan Academy of Family Physicians, 476 U.S. 667, 670–671 (1986).

In addition to prescribing procedures for agency action, the APA delineates the basic contours of judicial review of such action. As relevant here, Section 706 directs that “[t]o the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action.” 5 U. S. C. §706. It further requires courts to “hold unlawful and set aside agency action, findings, and conclusions found to be . . . not in accordance with law.” §706(2)(A).

The APA thus codifies for agency cases the unremarkable, yet elemental proposition reflected by judicial practice dating back to Marbury: that courts decide legal questions by applying their own judgment. It specifies that courts, not agencies, will decide “all relevant questions of law” arising on review of agency action, §706 (emphasis added)—even those involving ambiguous laws—and set aside any such action inconsistent with the law as they interpret it. And it prescribes no deferential standard for courts to employ in answering those legal questions. That omission is telling, because Section 706 does mandate that judicial review of agency policymaking and factfinding be deferential. See §706(2)(A) (agency action to be set aside if “arbitrary, capricious, [or] an abuse of discretion”); §706(2)(E) (agency factfinding in formal proceedings to be set aside if “unsupported by substantial evidence”).

In a statute designed to “serve as the fundamental charter of the administrative state,” Kisor v. Wilkie, 588 U.S. 558, 580 (2019) (plurality opinion) (internal quotation marks omitted), Congress surely would have articulated a similarly deferential standard applicable to questions of law had it intended to depart from the settled pre-APA understanding that deciding such questions was “exclusively a judicial function,” American Trucking Assns., 310 U. S., at 544. But nothing in the APA hints at such a dramatic departure. On the contrary, by directing courts to “interpret constitutional and statutory provisions” without differentiating between the two, Section 706 makes clear that agency interpretations of statutes—like agency interpretations of the Constitution—are not entitled to deference. Under the APA, it thus “remains the responsibility of the court to decide whether the law means what the agency says.” Perez v. Mortgage Bankers Assn., 575 U.S. 92, 109 (2015) (Scalia, J., concurring in judgment).[4]

The text of the APA means what it says. And a look at its history if anything only underscores that plain meaning. According to both the House and Senate Reports on the legislation, Section 706 “provide[d] that questions of law are for courts rather than agencies to decide in the last analysis.” H. R. Rep. No. 1980, 79th Cong., 2d Sess., 44 (1946) (emphasis added); accord, S. Rep. No. 752, 79th Cong., 1st Sess., 28 (1945). Some of the legislation’s most prominent supporters articulated the same view. See 92 Cong. Rec. 5654 (1946) (statement of Rep. Walter); P. McCarran, Improving “Administrative Justice”: Hearings and Evidence; Scope of Judicial Review, 32 A. B. A. J. 827, 831 (1946). Even the Department of Justice—an agency with every incentive to endorse a view of the APA favorable to the Executive Branch—opined after its enactment that Section 706 merely “restate[d] the present law as to the scope of judicial review.” Dept. of Justice, Attorney General’s Manual on the Administrative Procedure Act 108 (1947); see also Kisor, 588 U. S., at 582 (plurality opinion) (same). That “present law,” as we have described, adhered to the traditional conception of the judicial function. See supra, at 9–13.

Various respected commentators contemporaneously maintained that the APA required reviewing courts to exercise independent judgment on questions of law. Professor John Dickinson, for example, read the APA to “impose a clear mandate that all [questions of law] shall be decided by the reviewing Court itself, and in the exercise of its own independent judgment.” Administrative Procedure Act: Scope and Grounds of Broadened Judicial Review, 33 A. B. A. J. 434, 516 (1947). Professor Bernard Schwartz noted that §706 “would seem . . . to be merely a legislative restatement of the familiar review principle that questions of law are for the reviewing court, at the same time leaving to the courts the task of determining in each case what are questions of law.” Mixed Questions of Law and Fact and the Administrative Procedure Act, 19 Ford. L. Rev. 73, 84–85 (1950). And Professor Louis Jaffe, who had served in several agencies at the advent of the New Deal, thought that §706 leaves it up to the reviewing “court” to “decide as a ‘question of law’ whether there is ‘discretion’ in the premises”—that is, whether the statute at issue delegates particular discretionary authority to an agency. Judicial Control of Administrative Action 570 (1965).

The APA, in short, incorporates the traditional understanding of the judicial function, under which courts must exercise independent judgment in determining the meaning of statutory provisions. In exercising such judgment, though, courts may—as they have from the start—seek aid from the interpretations of those responsible for implementing particular statutes. Such interpretations “constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance” consistent with the APA. Skidmore, 323 U. S., at 140. And interpretations issued contemporaneously with the statute at issue, and which have remained consistent over time, may be especially useful in determining the statute’s meaning. See ibid.; American Trucking Assns., 310 U. S., at 549.

In a case involving an agency, of course, the statute’s meaning may well be that the agency is authorized to exercise a degree of discretion. Congress has often enacted such statutes. For example, some statutes “expressly delegate[ ]” to an agency the authority to give meaning to a particular statutory term. Batterton v. Francis, 432 U.S. 416, 425 (1977) (emphasis deleted).[5] Others empower an agency to prescribe rules to “fill up the details” of a statutory scheme, Wayman v. Southard, 10 Wheat. 1, 43 (1825), or to regulate subject to the limits imposed by a term or phrase that “leaves agencies with flexibility,” Michigan v. EPA, 576 U.S. 743, 752 (2015), such as “appropriate” or “reasonable.”[6]

When the best reading of a statute is that it delegates discretionary authority to an agency, the role of the reviewing court under the APA is, as always, to independently interpret the statute and effectuate the will of Congress subject to constitutional limits. The court fulfills that role by recognizing constitutional delegations, “fix[ing] the boundaries of [the] delegated authority,” H. Monaghan, Marbury and the Administrative State, 83 Colum. L. Rev. 1, 27 (1983), and ensuring the agency has engaged in “ ‘reasoned decisionmaking’ ” within those boundaries, Michigan, 576 U. S., at 750 (quoting Allentown Mack Sales & Service, Inc. v. NLRB, 522 U.S. 359, 374 (1998)); see also Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29 (1983). By doing so, a court upholds the traditional conception of the judicial function that the APA adopts.

III

The deference that Chevron requires of courts reviewing agency action cannot be squared with the APA.

A

In the decades between the enactment of the APA and this Court’s decision in Chevron, courts generally continued to review agency interpretations of the statutes they administer by independently examining each statute to determine its meaning. Cf. T. Merrill, Judicial Deference to Executive Precedent, 101 Yale L. J. 969, 972–975 (1992). As an early proponent (and later critic) of Chevron recounted, courts during this period thus identified delegations of discretionary authority to agencies on a “statute-by-statute basis.” A. Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L. J. 511, 516.

Chevron, decided in 1984 by a bare quorum of six Justices, triggered a marked departure from the traditional approach. The question in the case was whether an EPA regulation “allow[ing] States to treat all of the pollution-emitting devices within the same industrial grouping as though they were encased within a single ‘bubble’ ” was consistent with the term “stationary source” as used in the Clean Air Act. 467 U. S., at 840. To answer that question of statutory interpretation, the Court articulated and employed a now familiar two-step approach broadly applicable to review of agency action.

The first step was to discern “whether Congress ha[d] directly spoken to the precise question at issue.” Id., at 842. The Court explained that “[i]f the intent of Congress is clear, that is the end of the matter,” ibid., and courts were therefore to “reject administrative constructions which are contrary to clear congressional intent,” id., at 843, n. 9. To discern such intent, the Court noted, a reviewing court was to “employ[ ] traditional tools of statutory construction.” Ibid.

Without mentioning the APA, or acknowledging any doctrinal shift, the Court articulated a second step applicable when “Congress ha[d] not directly addressed the precise question at issue.” Id., at 843. In such a case—that is, a case in which “the statute [was] silent or ambiguous with respect to the specific issue” at hand—a reviewing court could not “simply impose its own construction on the statute, as would be necessary in the absence of an administrative interpretation.” Ibid. (footnote omitted). A court instead had to set aside the traditional interpretive tools and defer to the agency if it had offered “a permissible construction of the statute,” ibid., even if not “the reading the court would have reached if the question initially had arisen in a judicial proceeding,” ibid., n. 11. That directive was justified, according to the Court, by the understanding that administering statutes “requires the formulation of policy” to fill statutory “gap[s]”; by the long judicial tradition of according “considerable weight” to Executive Branch interpretations; and by a host of other considerations, including the complexity of the regulatory scheme, EPA’s “detailed and reasoned” consideration, the policy-laden nature of the judgment supposedly required, and the agency’s indirect accountability to the people through the President. Id., at 843, 844, and n. 14, 865.

Employing this new test, the Court concluded that Congress had not addressed the question at issue with the necessary “level of specificity” and that EPA’s interpretation was “entitled to deference.” Id., at 865. It did not matter why Congress, as the Court saw it, had not squarely addressed the question, see ibid., or that “the agency ha[d] from time to time changed its interpretation,” id., at 863. The latest EPA interpretation was a permissible reading of the Clean Air Act, so under the Court’s new rule, that reading controlled.

Initially, Chevron “seemed destined to obscurity.” T. Merrill, The Story of Chevron: The Making of an Accidental Landmark, 66 Admin. L. Rev. 253, 276 (2014). The Court did not at first treat it as the watershed decision it was fated to become; it was hardly cited in cases involving statutory questions of agency authority. See ibid. But within a few years, both this Court and the courts of appeals were routinely invoking its two-step framework as the governing standard in such cases. See id., at 276–277. As the Court did so, it revisited the doctrine’s justifications. Eventually, the Court decided that Chevron rested on “a presumption that Congress, when it left ambiguity in a statute meant for implementation by an agency, understood that the ambiguity would be resolved, first and foremost, by the agency, and desired the agency (rather than the courts) to possess whatever degree of discretion the ambiguity allows.” Smiley v. Citibank (South Dakota), N. A., 517 U.S. 735, 740–741 (1996); see also, e.g., Cuozzo Speed Technologies, LLC v. Lee, 579 U.S. 261, 276–277 (2016); Utility Air Regulatory Group v. EPA, 573 U.S. 302, 315 (2014); National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967, 982 (2005).

B

Neither Chevron nor any subsequent decision of this Court attempted to reconcile its framework with the APA. The “law of deference” that this Court has built on the foundation laid in Chevron has instead been “[h]eedless of the original design” of the APA. Perez, 575 U. S., at 109 (Scalia, J., concurring in judgment).

1

Chevron defies the command of the APA that “the reviewing court”—not the agency whose action it reviews—is to “decide all relevant questions of law” and “interpret . . . statutory provisions.” §706 (emphasis added). It requires a court to ignore, not follow, “the reading the court would have reached” had it exercised its independent judgment as required by the APA. Chevron, 467 U. S., at 843, n. 11. And although exercising independent judgment is consistent with the “respect” historically given to Executive Branch interpretations, see, e.g., Edwards’ Lessee, 12 Wheat., at 210; Skidmore, 323 U. S., at 140, Chevron insists on much more. It demands that courts mechanically afford binding deference to agency interpretations, including those that have been inconsistent over time. See 467 U. S., at 863. Still worse, it forces courts to do so even when a pre-existing judicial precedent holds that the statute means something else—unless the prior court happened to also say that the statute is “unambiguous.” Brand X, 545 U. S., at 982. That regime is the antithesis of the time honored approach the APA prescribes. In fretting over the prospect of “allow[ing]” a judicial interpretation of a statute “to override an agency’s” in a dispute before a court, ibid., Chevron turns the statutory scheme for judicial review of agency action upside down.

Chevron cannot be reconciled with the APA, as the Government and the dissent contend, by presuming that statutory ambiguities are implicit delegations to agencies. See Brief for Respondents in No. 22–1219, pp. 13, 37–38; post, at 4–15 (opinion of Kagan, J.). Presumptions have their place in statutory interpretation, but only to the extent that they approximate reality. Chevron’s presumption does not, because “[a]n ambiguity is simply not a delegation of law-interpreting power. Chevron confuses the two.” C. Sunstein, Interpreting Statutes in the Regulatory State, 103 Harv. L. Rev. 405, 445 (1989). As Chevron itself noted, ambiguities may result from an inability on the part of Congress to squarely answer the question at hand, or from a failure to even “consider the question” with the requisite precision. 467 U. S., at 865. In neither case does an ambiguity necessarily reflect a congressional intent that an agency, as opposed to a court, resolve the resulting interpretive question. And many or perhaps most statutory ambiguities may be unintentional. As the Framers recognized, ambiguities will inevitably follow from “the complexity of objects, . . . the imperfection of the human faculties,” and the simple fact that “no language is so copious as to supply words and phrases for every complex idea.” The Federalist No. 37, at 236.

Courts, after all, routinely confront statutory ambiguities in cases having nothing to do with Chevron—cases that do not involve agency interpretations or delegations of authority. Of course, when faced with a statutory ambiguity in such a case, the ambiguity is not a delegation to anybody, and a court is not somehow relieved of its obligation to independently interpret the statute. Courts in that situation do not throw up their hands because “Congress’s instructions have” supposedly “run out,” leaving a statutory “gap.” Post, at 2 (opinion of Kagan, J.). Courts instead understand that such statutes, no matter how impenetrable, do—in fact, must—have a single, best meaning. That is the whole point of having written statutes; “every statute’s meaning is fixed at the time of enactment.” Wisconsin Central Ltd. v. United States, 585 U.S. 274, 284 (2018) (emphasis deleted). So instead of declaring a particular party’s reading “permissible” in such a case, courts use every tool at their disposal to determine the best reading of the statute and resolve the ambiguity.

In an agency case as in any other, though, even if some judges might (or might not) consider the statute ambiguous, there is a best reading all the same—“the reading the court would have reached” if no agency were involved. Chevron, 467 U. S., at 843, n. 11. It therefore makes no sense to speak of a “permissible” interpretation that is not the one the court, after applying all relevant interpretive tools, concludes is best. In the business of statutory interpretation, if it is not the best, it is not permissible.

Perhaps most fundamentally, Chevron’s presumption is misguided because agencies have no special competence in resolving statutory ambiguities. Courts do. The Framers, as noted, anticipated that courts would often confront statutory ambiguities and expected that courts would resolve them by exercising independent legal judgment. And even Chevron itself reaffirmed that “[t]he judiciary is the final authority on issues of statutory construction” and recognized that “in the absence of an administrative interpretation,” it is “necessary” for a court to “impose its own construction on the statute.” Id., at 843, and n. 9. Chevron gravely erred, though, in concluding that the inquiry is fundamentally different just because an administrative interpretation is in play. The very point of the traditional tools of statutory construction—the tools courts use every day—is to resolve statutory ambiguities. That is no less true when the ambiguity is about the scope of an agency’s own power—perhaps the occasion on which abdication in favor of the agency is least appropriate.

2

The Government responds that Congress must generally intend for agencies to resolve statutory ambiguities because agencies have subject matter expertise regarding the statutes they administer; because deferring to agencies purportedly promotes the uniform construction of federal law; and because resolving statutory ambiguities can involve policymaking best left to political actors, rather than courts. See Brief for Respondents in No. 22–1219, pp. 16–19. The dissent offers more of the same. See post, at 9–14. But none of these considerations justifies Chevron’s sweeping presumption of congressional intent.

Beginning with expertise, we recently noted that interpretive issues arising in connection with a regulatory scheme often “may fall more naturally into a judge’s bailiwick” than an agency’s. Kisor, 588 U. S., at 578 (opinion of the Court). We thus observed that “[w]hen the agency has no comparative expertise in resolving a regulatory ambiguity, Congress presumably would not grant it that authority.” Ibid. Chevron’s broad rule of deference, though, demands that courts presume just the opposite. Under that rule, ambiguities of all stripes trigger deference. Indeed, the Government and, seemingly, the dissent continue to defend the proposition that Chevron applies even in cases having little to do with an agency’s technical subject matter expertise. See Brief for Respondents in No. 22–1219, p. 17; post, at 10.

But even when an ambiguity happens to implicate a technical matter, it does not follow that Congress has taken the power to authoritatively interpret the statute from the courts and given it to the agency. Congress expects courts to handle technical statutory questions. “[M]any statutory cases” call upon “courts [to] interpret the mass of technical detail that is the ordinary diet of the law,” Egelhoff v. Egelhoff, 532 U.S. 141, 161 (2001) (Breyer, J., dissenting), and courts did so without issue in agency cases before Chevron, see post, at 30 (Gorsuch, J., concurring). Courts, after all, do not decide such questions blindly. The parties and amici in such cases are steeped in the subject matter, and reviewing courts have the benefit of their perspectives. In an agency case in particular, the court will go about its task with the agency’s “body of experience and informed judgment,” among other information, at its disposal. Skidmore, 323 U. S., at 140. And although an agency’s interpretation of a statute “cannot bind a court,” it may be especially informative “to the extent it rests on factual premises within [the agency’s] expertise.” Bureau of Alcohol, Tobacco and Firearms v. FLRA, 464 U.S. 89, 98, n. 8 (1983). Such expertise has always been one of the factors which may give an Executive Branch interpretation particular “power to persuade, if lacking power to control.” Skidmore, 323 U. S., at 140; see, e.g., County of Maui v. Hawaii Wildlife Fund, 590 U.S. 165, 180 (2020); Moore, 95 U. S., at 763.

For those reasons, delegating ultimate interpretive authority to agencies is simply not necessary to ensure that the resolution of statutory ambiguities is well informed by subject matter expertise. The better presumption is therefore that Congress expects courts to do their ordinary job of interpreting statutes, with due respect for the views of the Executive Branch. And to the extent that Congress and the Executive Branch may disagree with how the courts have performed that job in a particular case, they are of course always free to act by revising the statute.

Nor does a desire for the uniform construction of federal law justify Chevron. Given inconsistencies in how judges apply Chevron, see infra, at 30–33, it is unclear how much the doctrine as a whole (as opposed to its highly deferential second step) actually promotes such uniformity. In any event, there is little value in imposing a uniform interpretation of a statute if that interpretation is wrong. We see no reason to presume that Congress prefers uniformity for uniformity’s sake over the correct interpretation of the laws it enacts.

The view that interpretation of ambiguous statutory provisions amounts to policymaking suited for political actors rather than courts is especially mistaken, for it rests on a profound misconception of the judicial role. It is reasonable to assume that Congress intends to leave policymaking to political actors. But resolution of statutory ambiguities involves legal interpretation. That task does not suddenly become policymaking just because a court has an “agency to fall back on.” Kisor, 588 U. S., at 575 (opinion of the Court). Courts interpret statutes, no matter the context, based on the traditional tools of statutory construction, not individual policy preferences. Indeed, the Framers crafted the Constitution to ensure that federal judges could exercise judgment free from the influence of the political branches. See The Federalist, No. 78, at 522–525. They were to construe the law with “[c]lear heads . . . and honest hearts,” not with an eye to policy preferences that had not made it into the statute. 1 Works of James Wilson 363 (J. Andrews ed. 1896).

That is not to say that Congress cannot or does not confer discretionary authority on agencies. Congress may do so, subject to constitutional limits, and it often has. But to stay out of discretionary policymaking left to the political branches, judges need only fulfill their obligations under the APA to independently identify and respect such delegations of authority, police the outer statutory boundaries of those delegations, and ensure that agencies exercise their discretion consistent with the APA. By forcing courts to instead pretend that ambiguities are necessarily delegations, Chevron does not prevent judges from making policy. It prevents them from judging.

3

In truth, Chevron’s justifying presumption is, as Members of this Court have often recognized, a fiction. See Buffington v. McDonough, 598 U. S. ___, ___ (2022) (Gorsuch, J., dissenting from denial of certiorari) (slip op., at 11); Cuozzo, 579 U. S., at 286 (Thomas, J., concurring); Scalia, 1989 Duke L. J., at 517; see also post, at 15 (opinion of Kagan, J.). So we have spent the better part of four decades imposing one limitation on Chevron after another, pruning its presumption on the understanding that “where it is in doubt that Congress actually intended to delegate particular interpretive authority to an agency, Chevron is ‘inapplicable.’ ” United States v. Mead Corp., 533 U.S. 218, 230 (2001) (quoting Christensen v. Harris County, 529 U.S. 576, 597 (2000) (Breyer, J., dissenting)); see also Adams Fruit Co. v. Barrett, 494 U.S. 638, 649 (1990).

Consider the many refinements we have made in an effort to match Chevron’s presumption to reality. We have said that Chevron applies only “when it appears that Congress delegated authority to the agency generally to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of that authority.” Mead, 533 U. S., at 226–227. In practice, that threshold requirement—sometimes called Chevron “step zero”—largely limits Chevron to “the fruits of notice-and-comment rulemaking or formal adjudication.” 533 U. S., at 230. But even when those processes are used, deference is still not warranted “where the regulation is ‘procedurally defective’—that is, where the agency errs by failing to follow the correct procedures in issuing the regulation.” Encino Motorcars, LLC v. Navarro, 579 U.S. 211, 220 (2016) (quoting Mead, 533 U. S., at 227).

Even where those procedural hurdles are cleared, substantive ones remain. Most notably, Chevron does not apply if the question at issue is one of “deep ‘economic and political significance.’ ” King v. Burwell, 576 U.S. 473, 486 (2015). We have instead expected Congress to delegate such authority “expressly” if at all, ibid., for “[e]xtraordinary grants of regulatory authority are rarely accomplished through ‘modest words,’ ‘vague terms,’ or ‘subtle device[s],’ ” West Virginia v. EPA, 597 U.S. 697, 723 (2022) (quoting Whitman v. American Trucking Assns., Inc., 531 U.S. 457, 468 (2001); alteration in original). Nor have we applied Chevron to agency interpretations of judicial review provisions, see Adams Fruit Co., 494 U. S., at 649–650, or to statutory schemes not administered by the agency seeking deference, see Epic Systems Corp. v. Lewis, 584 U.S. 497, 519–520 (2018). And we have sent mixed signals on whether Chevron applies when a statute has criminal applications. Compare Abramski v. United States, 573 U.S. 169, 191 (2014), with Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U.S. 687, 704, n. 18 (1995).

Confronted with this byzantine set of preconditions and exceptions, some courts have simply bypassed Chevron, saying it makes no difference for one reason or another.[7] And even when they do invoke Chevron, courts do not always heed the various steps and nuances of that evolving doctrine. In one of the cases before us today, for example, the First Circuit both skipped “step zero,” see 62 F. 4th, at 628, and refused to “classify [its] conclusion as a product of Chevron step one or step two”—though it ultimately appears to have deferred under step two, id., at 634.

This Court, for its part, has not deferred to an agency interpretation under Chevron since 2016. See Cuozzo, 579 U. S., at 280 (most recent occasion). But Chevron remains on the books. So litigants must continue to wrestle with it, and lower courts—bound by even our crumbling precedents, see Agostini v. Felton, 521 U.S. 203, 238 (1997)—understandably continue to apply it.

The experience of the last 40 years has thus done little to rehabilitate Chevron. It has only made clear that Chevron’s fictional presumption of congressional intent was always unmoored from the APA’s demand that courts exercise independent judgment in construing statutes administered by agencies. At best, our intricate Chevron doctrine has been nothing more than a distraction from the question that matters: Does the statute authorize the challenged agency action? And at worst, it has required courts to violate the APA by yielding to an agency the express responsibility, vested in “the reviewing court,” to “decide all relevant questions of law” and “interpret . . . statutory provisions.” §706 (emphasis added).

IV

The only question left is whether stare decisis, the doctrine governing judicial adherence to precedent, requires us to persist in the Chevron project. It does not. Stare decisis is not an “inexorable command,” Payne v. Tennessee, 501 U.S. 808, 828 (1991), and the stare decisis considerations most relevant here—“the quality of [the precedent’s] reasoning, the workability of the rule it established, . . . and reliance on the decision,” Knick v. Township of Scott, 588 U.S. 180, 203 (2019) (quoting Janus v. State, County, and Municipal Employees, 585 U.S. 878, 917 (2018))—all weigh in favor of letting Chevron go.

Chevron has proved to be fundamentally misguided. Despite reshaping judicial review of agency action, neither it nor any case of ours applying it grappled with the APA—the statute that lays out how such review works. Its flaws were nonetheless apparent from the start, prompting this Court to revise its foundations and continually limit its application. It has launched and sustained a cottage industry of scholars attempting to decipher its basis and meaning. And Members of this Court have long questioned its premises. See, e.g., Pereira v. Sessions, 585 U.S. 198, 219–221 (2018) (Kennedy, J., concurring); Michigan, 576 U. S., at 760–764 (Thomas, J., concurring); Buffington, 598 U. S. ___ (opinion of Gorsuch, J.); B. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2150–2154 (2016). Even Justice Scalia, an early champion of Chevron, came to seriously doubt whether it could be reconciled with the APA. See Perez, 575 U. S., at 109–110 (opinion concurring in judgment). For its entire existence, Chevron has been a “rule in search of a justification,” Knick, 588 U. S., at 204, if it was ever coherent enough to be called a rule at all.

Experience has also shown that Chevron is unworkable. The defining feature of its framework is the identification of statutory ambiguity, which requires deference at the doctrine’s second step. But the concept of ambiguity has always evaded meaningful definition. As Justice Scalia put the dilemma just five years after Chevron was decided: “How clear is clear?” 1989 Duke L. J., at 521.

We are no closer to an answer to that question than we were four decades ago. “ ‘[A]mbiguity’ is a term that may have different meanings for different judges.” Exxon Mobil Corp. v. Allapattah Services, Inc., 545 U.S. 546, 572 (2005) (Stevens, J., dissenting). One judge might see ambiguity everywhere; another might never encounter it. Compare L. Silberman, Chevron—The Intersection of Law & Policy, 58 Geo. Wash. L. Rev. 821, 822 (1990), with R. Kethledge, Ambiguities and Agency Cases: Reflections After (Almost) Ten Years on the Bench, 70 Vand. L. Rev. En Banc 315, 323 (2017). A rule of law that is so wholly “in the eye of the beholder,” Exxon Mobil Corp., 545 U. S., at 572 (Stevens, J., dissenting), invites different results in like cases and is therefore “arbitrary in practice,” Gulfstream Aerospace Corp. v. Mayacamas Corp., 485 U.S. 271, 283 (1988). Such an impressionistic and malleable concept “cannot stand as an every-day test for allocating” interpretive authority between courts and agencies. Swift & Co. v. Wickham, 382 U.S. 111, 125 (1965).

The dissent proves the point. It tells us that a court should reach Chevron’s second step when it finds, “at the end of its interpretive work,” that “Congress has left an ambiguity or gap.” Post, at 1–2. (The Government offers a similar test. See Brief for Respondents in No. 22–1219, pp. 7, 10, 14; Tr. of Oral Arg. 113–114, 116.) That is no guide at all. Once more, the basic nature and meaning of a statute does not change when an agency happens to be involved. Nor does it change just because the agency has happened to offer its interpretation through the sort of procedures necessary to obtain deference, or because the other preconditions for Chevron happen to be satisfied. The statute still has a best meaning, necessarily discernible by a court deploying its full interpretive toolkit. So for the dissent’s test to have any meaning, it must think that in an agency case (unlike in any other), a court should give up on its “interpretive work” before it has identified that best meaning. But how does a court know when to do so? On that point, the dissent leaves a gap of its own. It protests only that some other interpretive tools—all with pedigrees more robust than Chevron’s, and all designed to help courts identify the meaning of a text rather than allow the Executive Branch to displace it—also apply to ambiguous texts. See post, at 27. That this is all the dissent can come up with, after four decades of judicial experience attempting to identify ambiguity under Chevron, reveals the futility of the exercise.[8]

Because Chevron in its original, two-step form was so indeterminate and sweeping, we have instead been forced to clarify the doctrine again and again. Our attempts to do so have only added to Chevron’s unworkability, transforming the original two-step into a dizzying breakdance. See Adams Fruit Co., 494 U. S., at 649–650; Mead, 533 U. S., at 226–227; King, 576 U. S., at 486; Encino Motorcars, 579 U. S., at 220; Epic Systems, 584 U. S., at 519–520; on and on. And the doctrine continues to spawn difficult threshold questions that promise to further complicate the inquiry should Chevron be retained. See, e.g., Cargill v. Garland, 57 F. 4th 447, 465–468 (CA5 2023) (plurality opinion) (May the Government waive reliance on Chevron? Does Chevron apply to agency interpretations of statutes imposing criminal penalties? Does Chevron displace the rule of lenity?), aff ’d, 602 U. S. ___ (2024).

Four decades after its inception, Chevron has thus become an impediment, rather than an aid, to accomplishing the basic judicial task of “say[ing] what the law is.” Marbury, 1 Cranch, at 177. And its continuing import is far from clear. Courts have often declined to engage with the doctrine, saying it makes no difference. See n. 7, supra. And as noted, we have avoided deferring under Chevron since 2016. That trend is nothing new; for decades, we have often declined to invoke Chevron even in those cases where it might appear to be applicable. See W. Eskridge & L. Baer, The Continuum of Deference: Supreme Court Treatment of Agency Statutory Interpretations From Chevron to Hamdan, 96 Geo. L. J. 1083, 1125 (2008). At this point, all that remains of Chevron is a decaying husk with bold pretensions.

Nor has Chevron been the sort of “ ‘stable background’ rule” that fosters meaningful reliance. Post, at 8, n. 1 (opinion of Kagan, J.) (quoting Morrison v. National Australia Bank Ltd., 561 U.S. 247, 261 (2010)). Given our constant tinkering with and eventual turn away from Chevron, and its inconsistent application by the lower courts, it instead is hard to see how anyone—Congress included—could reasonably expect a court to rely on Chevron in any particular case. And even if it were possible to predict accurately when courts will apply Chevron, the doctrine “does not provide ‘a clear or easily applicable standard, so arguments for reliance based on its clarity are misplaced.’ ” Janus, 585 U. S., at 927 (quoting South Dakota v. Wayfair, Inc., 585 U.S. 162, 186 (2018)). To plan on Chevron yielding a particular result is to gamble not only that the doctrine will be invoked, but also that it will produce readily foreseeable outcomes and the stability that comes with them. History has proved neither bet to be a winning proposition.

Rather than safeguarding reliance interests, Chevron affirmatively destroys them. Under Chevron, a statutory ambiguity, no matter why it is there, becomes a license authorizing an agency to change positions as much as it likes, with “[u]nexplained inconsistency” being “at most . . . a reason for holding an interpretation to be . . . arbitrary and capricious.” Brand X, 545 U. S., at 981. But statutory ambiguity, as we have explained, is not a reliable indicator of actual delegation of discretionary authority to agencies. Chevron thus allows agencies to change course even when Congress has given them no power to do so. By its sheer breadth, Chevron fosters unwarranted instability in the law, leaving those attempting to plan around agency action in an eternal fog of uncertainty.

Chevron accordingly has undermined the very “rule of law” values that stare decisis exists to secure. Michigan v. Bay Mills Indian Community, 572 U.S. 782, 798 (2014). And it cannot be constrained by admonishing courts to be extra careful, or by tacking on a new batch of conditions. We would need to once again “revis[e] its theoretical basis . . . in order to cure its practical deficiencies.” Montejo v. Louisiana, 556 U.S. 778, 792 (2009). Stare decisis does not require us to do so, especially because any refinements we might make would only point courts back to their duties under the APA to “decide all relevant questions of law” and “interpret . . . statutory provisions.” §706. Nor is there any reason to wait helplessly for Congress to correct our mistake. The Court has jettisoned many precedents that Congress likewise could have legislatively overruled. See, e.g., Patterson v. McLean Credit Union, 485 U.S. 617, 618 (1988) (per curiam) (collecting cases). And part of “judicial humility,” post, at 3, 25 (opinion of Kagan, J.,), is admitting and in certain cases correcting our own mistakes, especially when those mistakes are serious, see post, at 8–9 (opinion of Gorsuch, J.).

This is one of those cases. Chevron was a judicial invention that required judges to disregard their statutory duties. And the only way to “ensure that the law will not merely change erratically, but will develop in a principled and intelligible fashion,” Vasquez v. Hillery, 474 U.S. 254, 265 (1986), is for us to leave Chevron behind.

By doing so, however, we do not call into question prior cases that relied on the Chevron framework. The holdings of those cases that specific agency actions are lawful—including the Clean Air Act holding of Chevron itself—are still subject to statutory stare decisis despite our change in interpretive methodology. See CBOCS West, Inc. v. Humphries, 553 U.S. 442, 457 (2008). Mere reliance on Chevron cannot constitute a “ ‘special justification’ ” for overruling such a holding, because to say a precedent relied on Chevron is, at best, “just an argument that the precedent was wrongly decided.” Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258, 266 (2014) (quoting Dickerson v. United States, 530 U.S. 428, 443 (2000)). That is not enough to justify overruling a statutory precedent.

*  *  *

The dissent ends by quoting Chevron: “ ‘Judges are not experts in the field.’ ” Post, at 31 (quoting 467 U. S., at 865). That depends, of course, on what the “field” is. If it is legal interpretation, that has been, “emphatically,” “the province and duty of the judicial department” for at least 221 years. Marbury, 1 Cranch, at 177. The rest of the dissent’s selected epigraph is that judges “ ‘are not part of either political branch.’ ” Post, at 31 (quoting Chevron, 467 U. S., at 865). Indeed. Judges have always been expected to apply their “judgment” independent of the political branches when interpreting the laws those branches enact. The Federalist No. 78, at 523. And one of those laws, the APA, bars judges from disregarding that responsibility just because an Executive Branch agency views a statute differently.

Chevron is overruled. Courts must exercise their independent judgment in deciding whether an agency has acted within its statutory authority, as the APA requires. Careful attention to the judgment of the Executive Branch may help inform that inquiry. And when a particular statute delegates authority to an agency consistent with constitutional limits, courts must respect the delegation, while ensuring that the agency acts within it. But courts need not and under the APA may not defer to an agency interpretation of the law simply because a statute is ambiguous.

Because the D. C. and First Circuits relied on Chevron in deciding whether to uphold the Rule, their judgments are vacated, and the cases are remanded for further proceedings consistent with this opinion.

It is so ordered.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 22–451 and 22–1219

_________________

LOPER BRIGHT ENTERPRISES, et al., PETITIONERS

22–451v.

GINA RAIMONDO, SECRETARY OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the district of columbia circuit

RELENTLESS, INC., et al., PETITIONERS

22–1219v.

DEPARTMENT OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the first circuit

[June 28, 2024]

Justice Thomas, concurring.

I join the Court’s opinion in full because it correctly concludes that Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), must finally be overruled. Under Chevron, a judge was required to adopt an agency’s interpretation of an ambiguous statute, so long as the agency had a “permissible construction of the statute.” See id., at 843. As the Court explains, that deference does not comport with the Administrative Procedure Act, which requires judges to decide “all relevant questions of law” and “interpret constitutional and statutory provisions” when reviewing an agency action. 5 U. S. C. §706; see also ante, at 18–23; Baldwin v. United States, 589 U. S. ___, ___–___ (2020) (Thomas, J., dissenting from denial of certiorari) (slip op., at 4–5).

I write separately to underscore a more fundamental problem: Chevron deference also violates our Constitution’s separation of powers, as I have previously explained at length. See Baldwin, 589 U. S., at ___–___ (dissenting opinion) (slip op., at 2–4); Michigan v. EPA, 576 U.S. 743, 761–763 (2015) (concurring opinion); see also Perez v. Mortgage Bankers Assn., 575 U.S. 92, 115–118 (2015) (opinion concurring in judgment). And, I agree with Justice Gorsuch that we should not overlook Chevron’s constitutional defects in overruling it.[1]* Post, at 15–20 (concurring opinion). To provide “practical and real protections for individual liberty,” the Framers drafted a Constitution that divides the legislative, executive, and judicial powers between three branches of Government. Perez, 575 U. S., at 118 (opinion of Thomas, J.). Chevron deference compromises this separation of powers in two ways. It curbs the judicial power afforded to courts, and simultaneously expands agencies’ executive power beyond constitutional limits.

Chevron compels judges to abdicate their Article III “judicial Power.” §1. “[T]he judicial power, as originally understood, requires a court to exercise its independent judgment in interpreting and expounding upon the laws.” Perez, 575 U. S., at 119 (opinion of Thomas, J.); accord, post, at 17–18 (opinion of Gorsuch, J.). The Framers understood that “legal texts . . . often contain ambiguities,” and that the judicial power included “the power to resolve these ambiguities over time.” Perez, 575 U. S., at 119 (opinion of Thomas, J.); accord, ante, at 7–9. But, under Chevron, a judge must accept an agency’s interpretation of an ambiguous law, even if he thinks another interpretation is correct. Ante, at 19. Chevron deference thus prevents judges from exercising their independent judgment to resolve ambiguities. Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (slip op., at 3); see also Michigan, 576 U. S., at 761 (opinion of Thomas, J.); see also Perez, 575 U. S., at 123 (opinion of Thomas, J.). By tying a judge’s hands, Chevron prevents the Judiciary from serving as a constitutional check on the Executive. It allows “the Executive . . . to dictate the outcome of cases through erroneous interpretations.” Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (slip op., at 4); Michigan, 576 U. S., at 763, n. 1 (opinion of Thomas, J.); see also Perez, 575 U. S., at 124 (opinion of Thomas, J.). Because the judicial power requires judges to exercise their independent judgment, the deference that Chevron requires contravenes Article III’s mandate.

Chevron deference also permits the Executive Branch to exercise powers not given to it. “When the Government is called upon to perform a function that requires an exercise of legislative, executive, or judicial power, only the vested recipient of that power can perform it.” Department of Transportation v. Association of American Railroads, 575 U.S. 43, 68 (2015) (Thomas, J., concurring in judgment). Because the Constitution gives the Executive Branch only “[t]he executive Power,” executive agencies may constitutionally exercise only that power. Art. II, §1, cl. 1. But, Chevron gives agencies license to exercise judicial power. By allowing agencies to definitively interpret laws so long as they are ambiguous, Chevron “transfer[s]” the Judiciary’s “interpretive judgment to the agency.” Perez, 575 U. S., at 124 (opinion of Thomas, J.); see also Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (slip op., at 4); Michigan, 576 U. S., at 761–762 (opinion of Thomas, J.); post, at 18 (Gorsuch, J., concurring).

Chevron deference “cannot be salvaged” by recasting it as deference to an agency’s “formulation of policy.” Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (internal quotation marks omitted) (slip op., at 3). If that were true, Chevron would mean that “agencies are unconstitutionally exercising ‘legislative Powers’ vested in Congress.” Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (slip op., at 3) (quoting Art. I, §1). By “giv[ing] the force of law to agency pronouncements on matters of private conduct as to which Congress did not actually have an intent,” Chevron “permit[s] a body other than Congress to perform a function that requires an exercise of legislative power.” Michigan, 576 U. S., at 762 (opinion of Thomas, J.) (internal quotation marks omitted). No matter the gloss put on it, Chevron expands agencies’ power beyond the bounds of Article II by permitting them to exercise powers reserved to another branch of Government.

Chevron deference was “not a harmless transfer of power.” Baldwin, 589 U. S., at ___ (opinion of Thomas, J.) (slip op., at 3). “The Constitution carefully imposes structural constraints on all three branches, and the exercise of power free of those accompanying restraints subverts the design of the Constitution’s ratifiers.” Ibid. In particular, the Founders envisioned that “the courts [would] check the Executive by applying the correct interpretation of the law.” Id., at ___ (slip op., at 4). Chevron was thus a fundamental disruption of our separation of powers. It improperly strips courts of judicial power by simultaneously increasing the power of executive agencies. By overruling Chevron, we restore this aspect of our separation of powers. To safeguard individual liberty, “[s]tructure is everything.” A. Scalia, Foreword: The Importance of Structure in Constitutional Interpretation, 83 Notre Dame L. Rev. 1417, 1418 (2008). Although the Court finally ends our 40-year misadventure with Chevron deference, its more profound problems should not be overlooked. Regardless of what a statute says, the type of deference required by Chevron violates the Constitution.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 22–451 and 22–1219

_________________

LOPER BRIGHT ENTERPRISES, et al., PETITIONERS

22–451v.

GINA RAIMONDO, SECRETARY OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the district of columbia circuit

RELENTLESS, INC., et al., PETITIONERS

22–1219v.

DEPARTMENT OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the first circuit

[June 28, 2024]

Justice Gorsuch, concurring.

In disputes between individuals and the government about the meaning of a federal law, federal courts have traditionally sought to offer independent judgments about “what the law is” without favor to either side. Marbury v. Madison, 1 Cranch 137, 177 (1803). Beginning in the mid-1980s, however, this Court experimented with a radically different approach. Applying Chevron deference, judges began deferring to the views of executive agency officials about the meaning of federal statutes. See Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). With time, the error of this approach became widely appreciated. So much so that this Court has refused to apply Chevron deference since 2016. Today, the Court places a tombstone on Chevron no one can miss. In doing so, the Court returns judges to interpretive rules that have guided federal courts since the Nation’s founding. I write separately to address why the proper application of the doctrine of stare decisis supports that course.

I

A

Today, the phrase “common law judge” may call to mind a judicial titan of the past who brilliantly devised new legal rules on his own. The phrase “stare decisis” might conjure up a sense that judges who come later in time are strictly bound to follow the work of their predecessors. But neither of those intuitions fairly describes the traditional common-law understanding of the judge’s role or the doctrine of stare decisis.

At common law, a judge’s charge to decide cases was not usually understood as a license to make new law. For much of England’s early history, different rulers and different legal systems prevailed in different regions. As England consolidated into a single kingdom governed by a single legal system, the judge’s task was to examine those pre-existing legal traditions and apply in the disputes that came to him those legal rules that were “common to the whole land and to all Englishmen.” F. Maitland, Equity, Also the Forms of Action at Common Law 2 (1929). That was “common law” judging.

This view of the judge’s role had consequences for the authority due judicial decisions. Because a judge’s job was to find and apply the law, not make it, the “opinion of the judge” and “the law” were not considered “one and the same thing.” 1 W. Blackstone, Commentaries on the Laws of England 71 (1765) (Blackstone) (emphasis deleted). A judge’s decision might bind the parties to the case at hand. M. Hale, The History and Analysis of the Common Law of England 68 (1713) (Hale). But none of that meant the judge had the power to “make a Law properly so called” for society at large, “for that only the King and Parliament can do.” Ibid.

Other consequences followed for the role precedent played in future judicial proceedings. Because past decisions represented something “less than a Law,” they did not bind future judges. Ibid. At the same time, as Matthew Hale put it, a future judge could give a past decision “Weight” as “Evidence” of the law. Ibid. Expressing the same idea, William Blackstone conceived of judicial precedents as “evidence” of “the common law.” 1 Blackstone 69, 71. And much like other forms of evidence, precedents at common law were thought to vary in the weight due them. Some past decisions might supply future courts with considerable guidance. But others might be entitled to lesser weight, not least because judges are no less prone to error than anyone else and they may sometimes “mistake” what the law demands. Id., at 71 (emphasis deleted). In cases like that, both men thought, a future judge should not rotely repeat a past mistake but instead “vindicate” the law “from misrepresentation.” Id., at 70.

When examining past decisions as evidence of the law, common law judges did not, broadly speaking, afford overwhelming weight to any “single precedent.” J. Baker, An Introduction to English Legal History 209–210 (5th ed. 2019). Instead, a prior decision’s persuasive force depended in large measure on its “Consonancy and Congruity with Resolutions and Decisions of former Times.” Hale 68. An individual decision might reflect the views of one court at one moment in time, but a consistent line of decisions representing the wisdom of many minds across many generations was generally considered stronger evidence of the law’s meaning. Ibid.

With this conception of precedent in mind, Lord Mansfield cautioned against elevating “particular cases” above the “general principles” that “run through the cases, and govern the decision of them.” Rust v. Cooper, 2 Cowp. 629, 632, 98 Eng. Rep. 1277, 1279 (K. B. 1777). By discarding aberrational rulings and pursuing instead the mainstream of past decisions, he observed, the common law tended over time to “wor[k] itself pure.” Omychund v. Barker, 1 Atk. 22, 33, 26 Eng. Rep. 15, 23 (Ch. 1744) (emphasis deleted). Reflecting similar thinking, Edmund Burke offered five principles for the evaluation of past judicial decisions: “They ought to be shewn; first, to be numerous and not scattered here and there;—secondly, concurrent and not contradictory and mutually destructive;—thirdly, to be made in good and constitutional times;—fourthly, not to be made to serve an occasion;—and fifthly, to be agreeable to the general tenor of legal principles.” Speech of Dec. 23, 1790, in 3 The Speeches of the Right Honourable Edmund Burke 513 (1816).

Not only did different decisions carry different weight, so did different language within a decision. An opinion’s holding and the reasoning essential to it (the ratio decidendi) merited careful attention. Dicta, stray remarks, and digressions warranted less weight. See N. Duxbury, The Intricacies of Dicta and Dissent 19–24 (2021) (Duxbury). These were no more than “the vapours and fumes of law.” F. Bacon, The Lord Keeper’s Speech in the Exchequer (1617), in 2 The Works of Francis Bacon 478 (B. Montagu ed. 1887) (Bacon).

That is not to say those “vapours” were worthless. Often dicta might provide the parties to a particular dispute a “fuller understanding of the court’s decisional path or related areas of concern.” B. Garner et al., The Law of Judicial Precedent 65 (2016) (Precedent). Dicta might also provide future courts with a source of “thoughtful advice.” Ibid. But future courts had to be careful not to treat every “hasty expression . . . as a serious and deliberate opinion.” Steel v. Houghton, 1 Bl. H. 51, 53, 126 Eng. Rep. 32, 33 (C. P. 1788). To do so would work an “injustice to [the] memory” of their predecessors who could not expect judicial remarks issued in one context to apply perfectly in others, perhaps especially ones they could not foresee. Ibid. Also, the limits of the adversarial process, a distinctive feature of English law, had to be borne in mind. When a single judge or a small panel reached a decision in a case, they did so based on the factual record and legal arguments the parties at hand have chosen to develop. Attuned to those constraints, future judges had to proceed with an open mind to the possibility that different facts and different legal arguments might dictate different outcomes in later disputes. See Duxbury 19–24.

B

Necessarily, this represents just a quick sketch of traditional common-law understandings of the judge’s role and the place of precedent in it. It focuses, too, on the horizontal, not vertical, force of judicial precedents. But there are good reasons to think that the common law’s understandings of judges and precedent outlined above crossed the Atlantic and informed the nature of the “judicial Power” the Constitution vests in federal courts. Art. III, §1.

Not only was the Constitution adopted against the backdrop of these understandings and, in light of that alone, they may provide evidence of what the framers meant when they spoke of the “judicial Power.” Many other, more specific provisions in the Constitution reflect much the same distinction between lawmaking and lawfinding functions the common law did. The Constitution provides that its terms may be amended only through certain prescribed democratic processes. Art. V. It vests the power to enact federal legislation exclusively in the people’s elected representatives in Congress. Art. I, §1. Meanwhile, the Constitution describes the judicial power as the power to resolve cases and controversies. Art. III, §2, cl. 1. As well, it delegates that authority to life-tenured judges, see §1, an assignment that would have made little sense if judges could usurp lawmaking powers vested in periodically elected representatives. But one that makes perfect sense if what is sought is a neutral party “to interpret and apply” the law without fear or favor in a dispute between others. 2 The Works of James Wilson 161 (J. Andrews ed. 1896) (Wilson); see Osborn v. Bank of United States, 9 Wheat. 738, 866 (1824).

The constrained view of the judicial power that runs through our Constitution carries with it familiar implications, ones the framers readily acknowledged. James Madison, for example, proclaimed that it would be a “fallacy” to suggest that judges or their precedents could “repeal or alter” the Constitution or the laws of the United States. Letter to N. Trist (Dec. 1831), in 9 The Writings of James Madison 477 (G. Hunt ed. 1910). A court’s opinion, James Wilson added, may be thought of as “effective la[w]” “[a]s to the parties.” Wilson 160–161. But as in England, Wilson said, a prior judicial decision could serve in a future dispute only as “evidence” of the law’s proper construction. Id., at 160; accord, 1 J. Kent, Commentaries on American Law 442–443 (1826).

The framers also recognized that the judicial power described in our Constitution implies, as the judicial power did in England, a power (and duty) of discrimination when it comes to assessing the “evidence” embodied in past decisions. So, for example, Madison observed that judicial rulings “repeatedly confirmed ” may supply better evidence of the law’s meaning than isolated or aberrant ones. Letter to C. Ingersoll (June 1831), in 4 Letters and Other Writings of James Madison 184 (1867) (emphasis added). Extending the thought, Thomas Jefferson believed it would often take “numerous decisions” for the meaning of new statutes to become truly “settled.” Letter to S. Jones (July 1809), in 12 The Writings of Thomas Jefferson 299 (A. Bergh ed. 1907).

From the start, too, American courts recognized that not everything found in a prior decision was entitled to equal weight. As Chief Justice Marshall warned, “It is a maxim not to be disregarded, that general expressions, in every opinion, are to be taken in connection with the case in which those expressions are used.” Cohens v. Virginia, 6 Wheat. 264, 399 (1821). To the extent a past court offered views “beyond the case,” those expressions “may be respected” in a later case “but ought not to control the judgment.” Ibid. One “obvious” reason for this, Marshall continued, had to do with the limits of the adversarial process we inherited from England: Only “[t]he question actually before the Court is investigated with care, and considered in its full extent. Other principles which may serve to illustrate it, are considered in their relation to the case decided, but their possible bearing on all other cases is seldom completely investigated.” Id., at 399–400.

Abraham Lincoln championed these traditional understandings in his debates with Stephen Douglas. Douglas took the view that a single decision of this Court—no matter how flawed—could definitively resolve a contested issue for everyone and all time. Those who thought otherwise, he said, “aim[ed] a deadly blow to our whole Republican system of government.” Speech at Springfield, Ill. (June 26, 1857), in 2 The Collected Works of Abraham Lincoln 401 (R. Basler ed. 1953) (Lincoln Speech). But Lincoln knew better. While accepting that judicial decisions “absolutely determine” the rights of the parties to a court’s judgment, he refused to accept that any single judicial decision could “fully settl[e]” an issue, particularly when that decision departs from the Constitution. Id., at 400–401. In cases such as these, Lincoln explained, “it is not resistance, it is not factious, it is not even disrespectful, to treat [the decision] as not having yet quite established a settled doctrine for the country.” Id., at 401.

After the Civil War, the Court echoed some of these same points. It stressed that every statement in a judicial opinion “must be taken in connection with its immediate context,” In re Ayers, 123 U.S. 443, 488 (1887), and stray “remarks” must not be elevated above the written law, see The Belfast, 7 Wall. 624, 641 (1869); see also, e.g., Trebilcock v. Wilson, 12 Wall. 687, 692–693 (1872); Mason v. Eldred, 6 Wall. 231, 236–238 (1868). During Chief Justice Chase’s tenure, it seems a Justice writing the Court’s majority opinion would generally work alone and present his work orally and in summary form to his colleagues at conference, which meant that other Justices often did not even review the opinion prior to publication. 6 C. Fairman, History of the Supreme Court of the United States 69–70 (1971). The Court could proceed in this way because it understood that a single judicial opinion may resolve a “case or controversy,” and in so doing it may make “effective law” for the parties, but it does not legislate for the whole of the country and is not to be confused with laws that do.

C

From all this, I see at least three lessons about the doctrine of stare decisis relevant to the decision before us today. Each concerns a form of judicial humility.

First, a past decision may bind the parties to a dispute, but it provides this Court no authority in future cases to depart from what the Constitution or laws of the United States ordain. Instead, the Constitution promises, the American people are sovereign and they alone may, through democratically responsive processes, amend our foundational charter or revise federal legislation. Unelected judges enjoy no such power. Part I–B, supra.

Recognizing as much, this Court has often said that stare decisis is not an “ ‘inexorable command.’ ” State Oil Co. v. Khan, 522 U.S. 3, 20 (1997). And from time to time it has found it necessary to correct its past mistakes. When it comes to correcting errors of constitutional interpretation, the Court has stressed the importance of doing so, for they can be corrected otherwise only through the amendment process. See, e.g., Franchise Tax Bd. of Cal. v. Hyatt, 587 U.S. 230, 248 (2019). When it comes to fixing errors of statutory interpretation, the Court has proceeded perhaps more circumspectly. But in that field, too, it has overruled even longstanding but “flawed” decisions. See, e.g., Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877, 904, 907 (2007).

Recent history illustrates all this. During the tenures of Chief Justices Warren and Burger, it seems this Court overruled an average of around three cases per Term, including roughly 50 statutory precedents between the 1960s and 1980s alone. See W. Eskridge, Overruling Statutory Precedents, 76 Geo. L. J. 1361, 1427–1434 (1988) (collecting cases). Many of these decisions came in settings no less consequential than today’s. In recent years, we have not approached the pace set by our predecessors, overruling an average of just one or two prior decisions each Term.[1] But the point remains: Judicial decisions inconsistent with the written law do not inexorably control.

Second, another lesson tempers the first. While judicial decisions may not supersede or revise the Constitution or federal statutory law, they merit our “respect as embodying the considered views of those who have come before.” Ramos v. Louisiana, 590 U.S. 83, 105 (2020). As a matter of professional responsibility, a judge must not only avoid confusing his writings with the law. When a case comes before him, he must also weigh his view of what the law demands against the thoughtful views of his predecessors. After all, “[p]recedent is a way of accumulating and passing down the learning of past generations, a font of established wisdom richer than what can be found in any single judge or panel of judges.” Precedent 9.

Doubtless, past judicial decisions may, as they always have, command “greater or less authority as precedents, according to circumstances.” Lincoln Speech 401. But, like English judges before us, we have long turned to familiar considerations to guide our assessment of the weight due a past decision. So, for example, as this Court has put it, the weight due a precedent may depend on the quality of its reasoning, its consistency with related decisions, its workability, and reliance interests that have formed around it. See Ramos, 590 U. S., at 106. The first factor recognizes that the primary power of any precedent lies in its power to persuade—and poorly reasoned decisions may not provide reliable evidence of the law’s meaning. The second factor reflects the fact that a precedent is more likely to be correct and worthy of respect when it reflects the time-tested wisdom of generations than when it sits “unmoored” from surrounding law. Ibid. The remaining factors, like workability and reliance, do not often supply reason enough on their own to abide a flawed decision, for almost any past decision is likely to benefit some group eager to keep things as they are and content with how things work. See, e.g., id., at 108. But these factors can sometimes serve functions similar to the others, by pointing to clues that may suggest a past decision is right in ways not immediately obvious to the individual judge.

When asking whether to follow or depart from a precedent, some judges deploy adverbs. They speak of whether or not a precedent qualifies as “demonstrably erroneous,” Gamble v. United States, 587 U.S. 678, 711 (2019) (Thomas, J., concurring), or “egregiously wrong,” Ramos, 590 U. S., at 121 (Kavanaugh, J., concurring in part). But the emphasis the adverb imparts is not meant for dramatic effect. It seeks to serve instead as a reminder of a more substantive lesson. The lesson that, in assessing the weight due a past decision, a judge is not to be guided by his own impression alone, but must self-consciously test his views against those who have come before, open to the possibility that a precedent might be correct in ways not initially apparent to him.

Third, it would be a mistake to read judicial opinions like statutes. Adopted through a robust and democratic process, statutes often apply in all their particulars to all persons. By contrast, when judges reach a decision in our adversarial system, they render a judgment based only on the factual record and legal arguments the parties at hand have chosen to develop. A later court assessing a past decision must therefore appreciate the possibility that different facts and different legal arguments may dictate a different outcome. They must appreciate, too, that, like anyone else, judges are “innately digressive,” and their opinions may sometimes offer stray asides about a wider topic that may sound nearly like legislative commands. Duxbury 4. Often, enterprising counsel seek to exploit such statements to maximum effect. See id., at 25. But while these digressions may sometimes contain valuable counsel, they remain “vapours and fumes of law,” Bacon 478, and cannot “control the judgment in a subsequent suit,” Cohens, 6 Wheat., at 399.

These principles, too, have long guided this Court and others. As Judge Easterbrook has put it, an “opinion is not a comprehensive code; it is just an explanation for the Court’s disposition. Judicial opinions must not be confused with statutes, and general expressions must be read in light of the subject under consideration.” United States v. Skoien, 614 F.3d 638, 640 (CA7 2010) (en banc); see also Reiter v. Sonotone Corp., 442 U.S. 330, 341 (1979) (stressing that an opinion is not “a statute,” and its language should not “be parsed” as if it were); Nevada v. Hicks, 533 U.S. 353, 372 (2001) (same). If stare decisis counsels respect for the thinking of those who have come before, it also counsels against doing an “injustice to [their] memory” by overreliance on their every word. Steel, 1 Bl. H., at 53, 126 Eng. Rep., at 33. As judges, “[w]e neither expect nor hope that our successors will comb” through our opinions, searching for delphic answers to matters we never fully explored. Brown v. Davenport, 596 U.S. 118, 141 (2022). To proceed otherwise risks “turn[ing] stare decisis from a tool of judicial humility into one of judicial hubris.” Ibid.

II

Turning now directly to the question what stare decisis effect Chevron deference warrants, each of these lessons seem to me to weigh firmly in favor of the course the Court charts today: Lesson 1, because Chevron deference contravenes the law Congress prescribed in the Administrative Procedure Act. Lesson 2, because Chevron deference runs against mainstream currents in our law regarding the separation of powers, due process, and centuries-old interpretive rules that fortify those constitutional commitments. And Lesson 3, because to hold otherwise would effectively require us to endow stray statements in Chevron with the authority of statutory language, all while ignoring more considered language in that same decision and the teachings of experience.

A

Start with Lesson 1. The Administrative Procedure Act of 1946 (APA) directs a “reviewing court” to “decide all relevant questions of law” and “interpret” relevant “constitutional and statutory provisions.” 5 U. S. C. §706. When applying Chevron deference, reviewing courts do not interpret all relevant statutory provisions and decide all relevant questions of law. Instead, judges abdicate a large measure of that responsibility in favor of agency officials. Their interpretations of “ambiguous” laws control even when those interpretations are at odds with the fairest reading of the law an independent “reviewing court” can muster. Agency officials, too, may change their minds about the law’s meaning at any time, even when Congress has not amended the relevant statutory language in any way. National Cable & Telecommunications Assn. v. Brand X Internet Services, 545 U.S. 967, 982–983 (2005). And those officials may even disagree with and effectively overrule not only their own past interpretations of a law but a court’s past interpretation as well. Ibid. None of that is consistent with the APA’s clear mandate.

The hard fact is Chevron “did not even bother to cite” the APA, let alone seek to apply its terms. United States v. Mead Corp., 533 U.S. 218, 241 (2001) (Scalia, J., dissenting). Instead, as even its most ardent defenders have conceded, Chevron deference rests upon a “fictionalized statement of legislative desire,” namely, a judicial supposition that Congress implicitly wishes judges to defer to executive agencies’ interpretations of the law even when it has said nothing of the kind. D. Barron & E. Kagan, Chevron’s Nondelegation Doctrine, 2001 S. Ct. Rev. 201, 212 (Kagan) (emphasis added). As proponents see it, that fiction represents a “policy judgmen[t] about what . . . make[s] for good government.” Ibid.[2] But in our democracy unelected judges possess no authority to elevate their own fictions over the laws adopted by the Nation’s elected representatives. Some might think the legal directive Congress provided in the APA unwise; some might think a different arrangement preferable. See, e.g., post, at 9–11 (Kagan, J., dissenting). But it is Congress’s view of “good government,” not ours, that controls.

Much more could be said about Chevron’s inconsistency with the APA. But I have said it in the past. See Buffington v. McDonough, 598 U. S. ___, ___–___ (2022) (opinion dissenting from denial of certiorari) (slip op., at 5–6); Gutierrez-Brizuela v. Lynch, 834 F.3d 1142, 1151–1153 (CA10 2016) (concurring opinion). And the Court makes many of the same points at length today. See ante, at 18–22. For present purposes, the short of it is that continuing to abide Chevron deference would require us to transgress the first lesson of stare decisis—the humility required of judges to recognize that our decisions must yield to the laws adopted by the people’s elected representatives.[3]

B

Lesson 2 cannot rescue Chevron deference. If stare decisis calls for judicial humility in the face of the written law, it also cautions us to test our present conclusions carefully against the work of our predecessors. At the same time and as we have seen, this second form of humility counsels us to remember that precedents that have won the endorsement of judges across many generations, demonstrated coherence with our broader law, and weathered the tests of time and experience are entitled to greater consideration than those that have not. See Part I, supra. Viewed by each of these lights, the case for Chevron deference only grows weaker still.

1

Start with a look to how our predecessors traditionally understood the judicial role in disputes over a law’s meaning. From the Nation’s founding, they considered “[t]he interpretation of the laws” in cases and controversies “the proper and peculiar province of the courts.” The Federalist No. 78, p. 467 (C. Rossiter ed. 1961) (A. Hamilton). Perhaps the Court’s most famous early decision reflected exactly that view. There, Chief Justice Marshall declared it “emphatically the province and duty of the judicial department to say what the law is.” Marbury, 1 Cranch, at 177. For judges “have neither FORCE nor WILL but merely judgment”—and an obligation to exercise that judgment independently. The Federalist No. 78, at 465. No matter how “disagreeable that duty may be,” this Court has said, a judge “is not at liberty to surrender, or to waive it.” United States v. Dickson, 15 Pet. 141, 162 (1841) (Story, J.). This duty of independent judgment is perhaps “the defining characteristi[c] of Article III judges.” Stern v. Marshall, 564 U.S. 462, 483 (2011).

To be sure, this Court has also long extended “great respect” to the “contemporaneous” and consistent views of the coordinate branches about the meaning of a statute’s terms. Edwards’ Lessee v. Darby, 12 Wheat. 206, 210 (1827); see also McCulloch v. Maryland, 4 Wheat. 316, 401 (1819); Stuart v. Laird, 1 Cranch 299, 309 (1803).[4] But traditionally, that did not mean a court had to “defer” to any “reasonable” construction of an “ambiguous” law that an executive agency might offer. It did not mean that the government could propound a “reasonable” view of the law’s meaning one day, a different one the next, and bind the judiciary always to its latest word. Nor did it mean the executive could displace a pre-existing judicial construction of a statute’s terms, replace it with its own, and effectively overrule a judicial precedent in the process. Put simply, this Court was “not bound” by any and all reasonable “administrative construction[s]” of ambiguous statutes when resolving cases and controversies. Burnet v. Chicago Portrait Co., 285 U.S. 1, 16 (1932). While the executive’s consistent and contemporaneous views warranted respect, they “by no means control[led] the action or the opinion of this court in expounding the law with reference to the rights of parties litigant before them.” Irvine v. Marshall, 20 How. 558, 567 (1858); see also A. Bamzai, The Origins of Judicial Deference to Executive Interpretation, 126 Yale L. J. 908, 987 (2017).

Sensing how jarringly inconsistent Chevron is with this Court’s many longstanding precedents discussing the nature of the judicial role in disputes over the law’s meaning, the government and dissent struggle for a response. The best they can muster is a handful of cases from the early 1940s in which, they say, this Court first “put [deference] principles into action.” Post, at 21 (Kagan, J., dissenting). And, admittedly, for a period this Court toyed with a form of deference akin to Chevron, at least for so-called mixed questions of law and fact. See, e.g., Gray v. Powell, 314 U.S. 402, 411–412 (1941); NLRB v. Hearst Publications, Inc., 322 U.S. 111, 131 (1944). But, as the Court details, even that limited experiment did not last. See ante, at 10–12. Justice Roberts, in his Gray dissent, decried these decisions for “abdicat[ing our] function as a court of review” and “complete[ly] revers[ing] . . . the normal and usual method of construing a statute.” 314 U. S., at 420–421. And just a few years later, in Skidmore v. Swift & Co., 323 U.S. 134 (1944), the Court returned to its time-worn path.

Echoing themes that had run throughout our law from its start, Justice Robert H. Jackson wrote for the Court in Skidmore. There, he said, courts may extend respectful consideration to another branch’s interpretation of the law, but the weight due those interpretations must always “depend upon the[ir] thoroughness . . . , the validity of [their] reasoning, [their] consistency with earlier and later pronouncements, and all those factors which give [them] power to persuade.” Id., at 140. In another case the same year, and again writing for the Court, Justice Jackson expressly rejected a call for a judge-made doctrine of deference much like Chevron, offering that, “[i]f Congress had deemed it necessary or even appropriate” for courts to “defe[r] to administrative construction[,] . . . it would not have been at a loss for words to say so.” Davies Warehouse Co. v. Bowles, 321 U.S. 144, 156 (1944).

To the extent proper respect for precedent demands, as it always has, special respect for longstanding and mainstream decisions, Chevron scores badly. It represented not a continuation of a long line of decisions but a break from them. Worse, it did not merely depart from our precedents. More nearly, Chevron defied them.

2

Consider next how uneasily Chevron deference sits alongside so many other settled aspects of our law. Having witnessed first-hand King George’s efforts to gain influence and control over colonial judges, see Declaration of Independence ¶ 11, the framers made a considered judgment to build judicial independence into the Constitution’s design. They vested the judicial power in decisionmakers with life tenure. Art. III, §1. They placed the judicial salary beyond political control during a judge’s tenure. Ibid. And they rejected any proposal that would subject judicial decisions to review by political actors. The Federalist No. 81, at 482; United States v. Hansen, 599 U.S. 762, 786–791 (2023) (Thomas, J., concurring). All of this served to ensure the same thing: “A fair trial in a fair tribunal.” In re Murchison, 349 U.S. 133, 136 (1955). One in which impartial judges, not those currently wielding power in the political branches, would “say what the law is” in cases coming to court. Marbury, 1 Cranch, at 177.

Chevron deference undermines all that. It precludes courts from exercising the judicial power vested in them by Article III to say what the law is. It forces judges to abandon the best reading of the law in favor of views of those presently holding the reins of the Executive Branch. It requires judges to change, and change again, their interpretations of the law as and when the government demands. And that transfer of power has exactly the sort of consequences one might expect. Rather than insulate adjudication from power and politics to ensure a fair hearing “without respect to persons” as the federal judicial oath demands, 28 U. S. C. §453, Chevron deference requires courts to “place a finger on the scales of justice in favor of the most powerful of litigants, the federal government.” Buffington, 598 U. S., at ___ (slip op., at 9). Along the way, Chevron deference guarantees “systematic bias” in favor of whichever political party currently holds the levers of executive power. P. Hamburger, Chevron Bias, 84 Geo. Wash. L. Rev. 1187, 1212 (2016).

Chevron deference undermines other aspects of our settled law, too. In this country, we often boast that the Constitution’s promise of due process of law, see Amdts. 5, 14, means that “ ‘no man can be a judge in his own case.’ ” Williams v. Pennsylvania, 579 U.S. 1, 8–9 (2016); Calder v. Bull, 3 Dall. 386, 388 (1798) (opinion of Chase, J.). That principle, of course, has even deeper roots, tracing far back into the common law where it was known by the Latin maxim nemo iudex in causa sua. See 1 E. Coke, Institutes of the Laws of England §212, *141a. Yet, under the Chevron regime, all that means little, for executive agencies may effectively judge the scope of their own lawful powers. See, e.g., Arlington v. FCC, 569 U.S. 290, 296–297 (2013).

Traditionally, as well, courts have sought to construe statutes as a reasonable reader would “when the law was made.” Blackstone 59; see United States v. Fisher, 2 Cranch 358, 386 (1805). Today, some call this “textualism.” But really it’s a very old idea, one that constrains judges to a lawfinding rather than lawmaking role by focusing their work on the statutory text, its linguistic context, and various canons of construction. In that way, textualism serves as an essential guardian of the due process promise of fair notice. If a judge could discard an old meaning and assign a new one to a law’s terms, all without any legislative revision, how could people ever be sure of the rules that bind them? New Prime Inc. v. Oliveira, 586 U.S. 105, 113 (2019). Were the rules otherwise, Blackstone warned, the people would be rendered “slaves to their magistrates.” 4 Blackstone 371.

Yet, replace “magistrates” with “bureaucrats,” and Blackstone’s fear becomes reality when courts employ Chevron deference. Whenever we confront an ambiguity in the law, judges do not seek to resolve it impartially according to the best evidence of the law’s original meaning. Instead, we resort to a far cruder heuristic: “The reasonable bureaucrat always wins.” And because the reasonable bureaucrat may change his mind year-to-year and election-to-election, the people can never know with certainty what new “interpretations” might be used against them. This “fluid” approach to statutory interpretation is “as much a trap for the innocent as the ancient laws of Caligula,” which were posted so high up on the walls and in print so small that ordinary people could never be sure what they required. United States v. Cardiff, 344 U.S. 174, 176 (1952).

The ancient rule of lenity is still another of Chevron’s victims. Since the founding, American courts have construed ambiguities in penal laws against the government and with lenity toward affected persons. Wooden v. United States, 595 U.S. 360, 388–390 (2022) (Gorsuch, J., concurring in judgment). That principle upholds due process by safeguarding individual liberty in the face of ambiguous laws. Ibid. And it fortifies the separation of powers by keeping the power of punishment firmly “ ‘in the legislative, not in the judicial department.’ ” Id., at 391 (quoting United States v. Wiltberger, 5 Wheat. 76, 95 (1820)). But power begets power. And pressing Chevron deference as far as it can go, the government has sometimes managed to leverage “ambiguities” in the written law to penalize conduct Congress never clearly proscribed. Compare Guedes v. ATF, 920 F.3d 1, 27–28, 31 (CADC 2019), with Garland v. Cargill, 602 U.S. 604 (2024).

In all these ways, Chevron’s fiction has led us to a strange place. One where authorities long thought reserved for Article III are transferred to Article II, where the scales of justice are tilted systematically in favor of the most powerful, where legal demands can change with every election even though the laws do not, and where the people are left to guess about their legal rights and responsibilities. So much tension with so many foundational features of our legal order is surely one more sign that we have “taken a wrong turn along the way.” Kisor v. Wilkie, 588 U.S. 558, 607 (2019) (Gorsuch, J., concurring in judgment).[5]

3

Finally, consider workability and reliance. If, as I have sought to suggest, these factors may sometimes serve as useful proxies for the question whether a precedent comports with the historic tide of judicial practice or represents an aberrational mistake, see Part I–C, supra, they certainly do here.

Take Chevron’s “workability.” Throughout its short life, this Court has been forced to supplement and revise Chevron so many times that no one can agree on how many “steps” it requires, nor even what each of those “steps” entails. Some suggest that the analysis begins with “step zero” (perhaps itself a tell), an innovation that traces to United States v. Mead Corp., 533 U.S. 218. Mead held that, before even considering whether Chevron applies, a court must determine whether Congress meant to delegate to the agency authority to interpret the law in a given field. 533 U. S., at 226–227. But that exercise faces an immediate challenge: Because Chevron depends on a judicially implied, rather than a legislatively expressed, delegation of interpretive authority to an executive agency, Part II–A, supra, when should the fiction apply and when not? Mead fashioned a multifactor test for judges to use. 533 U. S., at 229–231. But that test has proved as indeterminate in application as it was contrived in origin. Perhaps for these reasons, perhaps for others, this Court has sometimes applied Mead and often ignored it. See Brand X, 545 U. S., at 1014, n. 8 (Scalia, J., dissenting).

Things do not improve as we move up the Chevron ladder. At “step one,” a judge must defer to an executive official’s interpretation when the statute at hand is “ambiguous.” But even today, Chevron’s principal beneficiary—the federal government—still cannot say when a statute is sufficiently ambiguous to trigger deference. See, e.g., Tr. of Oral Arg. in American Hospital Assn. v. Becerra, O. T. 2021, No. 20–1114, pp. 71–72. Perhaps thanks to this particular confusion, the search for ambiguity has devolved into a sort of Snark hunt: Some judges claim to spot it almost everywhere, while other equally fine judges claim never to have seen it. Compare L. Silberman, Chevron—The Intersection of Law & Policy, 58 Geo. Wash. L. Rev. 821, 826 (1990), with R. Kethledge, Ambiguities and Agency Cases: Reflections After (Almost) Ten Years on the Bench, 70 Vand. L. Rev. En Banc 315, 323 (2017).

Nor do courts agree when it comes to “step two.” There, a judge must assess whether an executive agency’s interpretation of an ambiguous statute is “reasonable.” But what does that inquiry demand? Some courts engage in a comparatively searching review; others almost reflexively defer to an agency’s views. Here again, courts have pursued “wildly different” approaches and reached wildly different conclusions in similar cases. See B. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2152 (2016) (Kavanaugh).

Today’s cases exemplify some of these problems. We have before us two circuit decisions, three opinions, and at least as many interpretive options on the Chevron menu. On the one hand, we have the D. C. Circuit majority, which deemed the Magnuson-Stevens Act “ambiguous” and upheld the agency’s regulation as “ ‘permissible.’ ” 45 F. 4th 359, 365 (2022). On the other hand, we have the D. C. Circuit dissent, which argues the statute is “unambiguou[s]” and that it plainly forecloses the agency’s new rule. Id., at 372 (opinion of Walker, J.). And on yet a third hand, we have the First Circuit, which claimed to have identified “clear textual support” for the regulation, yet refused to say whether it would “classify [its] conclusion as a product of Chevron step one or step two.” 62 F. 4th 621, 631, 634 (2023). As these cases illustrate, Chevron has turned statutory interpretation into a game of bingo under blindfold, with parties guessing at how many boxes there are and which one their case might ultimately fall in.

Turn now from workability to reliance. Far from engendering reliance interests, the whole point of Chevron deference is to upset them. Under Chevron, executive officials can replace one “reasonable” interpretation with another at any time, all without any change in the law itself. The result: Affected individuals “can never be sure of their legal rights and duties.” Buffington, 598 U. S., at ___ (slip op., at 12).

How bad is the problem? Take just one example. Brand X concerned a law regulating broadband internet services. There, the Court upheld an agency rule adopted by the administration of President George W. Bush because it was premised on a “reasonable” interpretation of the statute. Later, President Barack Obama’s administration rescinded the rule and replaced it with another. Later still, during President Donald J. Trump’s administration, officials replaced that rule with a different one, all before President Joseph R. Biden, Jr.’s administration declared its intention to reverse course for yet a fourth time. See Safeguarding and Securing the Open Internet, 88 Fed. Reg. 76048 (2023); Brand X, 545 U. S., at 981–982. Each time, the government claimed its new rule was just as “reasonable” as the last. Rather than promoting reliance by fixing the meaning of the law, Chevron deference engenders constant uncertainty and convulsive change even when the statute at issue itself remains unchanged.

Nor are these antireliance harms distributed equally. Sophisticated entities and their lawyers may be able to keep pace with rule changes affecting their rights and responsibilities. They may be able to lobby for new “ ‘reasonable’ ” agency interpretations and even capture the agencies that issue them. Buffington, 598 U. S., at ___, ___ (slip op., at 8, 13). But ordinary people can do none of those things. They are the ones who suffer the worst kind of regulatory whiplash Chevron invites.

Consider a couple of examples. Thomas Buffington, a veteran of the U. S. Air Force, was injured in the line of duty. For a time after he left the Air Force, the Department of Veterans Affairs (VA) paid disability benefits due him by law. But later the government called on Mr. Buffington to reenter active service. During that period, everyone agreed, the VA could (as it did) suspend his disability payments. After he left active service for a second time, however, the VA turned his patriotism against him. By law, Congress permitted the VA to suspend disability pay only “for any period for which [a servicemember] receives active service pay.” 38 U. S. C. §5304(c). But the VA had adopted a self-serving regulation requiring veterans to file a form asking for the resumption of their disability pay after a second (or subsequent) stint in active service. 38 CFR §3.654(b)(2) (2021). Unaware of the regulation, Mr. Buffington failed to reapply immediately. When he finally figured out what had happened and reapplied, the VA agreed to resume payments going forward but refused to give Mr. Buffington all of the past disability payments it had withheld. Buffington, 598 U. S., at ___–___ (slip op., at 1–4).

Mr. Buffington challenged the agency’s action as inconsistent with Congress’s direction that the VA may suspend disability payments only for those periods when a veteran returns to active service. But armed with Chevron, the agency defeated Mr. Buffington’s claim. Maybe the self-serving regulation the VA cited as justification for its action was not premised on the best reading of the law, courts said, but it represented a “ ‘permissible’ ” one. 598 U. S., at ___ (slip op., at 7). In that way, the Executive Branch was able to evade Congress’s promises to someone who took the field repeatedly in the Nation’s defense.

In another case, one which I heard as a court of appeals judge, De Niz Robles v. Lynch, 803 F.3d 1165 (CA10 2015), the Board of Immigration Appeals invoked Chevron to overrule a judicial precedent on which many immigrants had relied, see In re Briones, 24 I. & N. Dec. 355, 370 (BIA 2007) (purporting to overrule Padilla–Caldera v. Gonzales, 426 F.3d 1294 (CA10 2005)). The agency then sought to apply its new interpretation retroactively to punish those immigrants—including Alfonzo De Niz Robles, who had relied on that judicial precedent as authority to remain in this country with his U. S. wife and four children. See 803 F. 3d, at 1168–1169. Our court ruled that this retrospective application of the BIA’s new interpretation of the law violated Mr. De Niz Robles’s due process rights. Id., at 1172. But as a lower court, we could treat only the symptom, not the disease. So Chevron permitted the agency going forward to overrule a judicial decision about the best reading of the law with its own different “reasonable” one and in that way deny relief to countless future immigrants.

Those are just two stories among so many that federal judges could tell (and have told) about what Chevron deference has meant for ordinary people interacting with the federal government. See, e.g., Lambert v. Saul, 980 F.3d 1266, 1268–1276 (CA9 2020); Valent v. Commissioner of Social Security, 918 F.3d 516, 525–527 (CA6 2019) (Kethledge, J., dissenting); Gonzalez v. United States Atty. Gen., 820 F.3d 399, 402–405 (CA11 2016) (per curiam).

What does the federal government have to say about this? It acknowledges that Chevron sits as a heavy weight on the scale in favor of the government, “oppositional” to many “categories of individuals.” Tr. of Oral Arg. in No. 22–1219, p. 133 (Relentless Tr.). But, according to the government, Chevron deference is too important an innovation to undo. In its brief reign, the government says, it has become a “fundamenta[l] . . . ground rul[e] for how all three branches of the government are operating together.” Relentless Tr. 102. But, in truth, the Constitution, the APA, and our longstanding precedents set those ground rules some time ago. And under them, agencies cannot invoke a judge-made fiction to unsettle our Nation’s promise to individuals that they are entitled to make their arguments about the law’s demands on them in a fair hearing, one in which they stand on equal footing with the government before an independent judge.

C

How could a Court, guided for 200 years by Chief Justice Marshall’s example, come to embrace a counter-Marbury revolution, one at war with the APA, time honored precedents, and so much surrounding law? To answer these questions, turn to Lesson 3 and witness the temptation to endow a stray passage in a judicial decision with extraordinary authority. Call it “power quoting.”

Chevron was an unlikely place for a revolution to begin. The case concerned the Clean Air Act’s requirement that States regulate “stationary sources” of air pollution in their borders. See 42 U. S. C. §7401 et seq. At the time, it was an open question whether entire industrial plants or their constituent polluting parts counted as “stationary sources.” The Environmental Protection Agency had defined entire plants as sources, an approach that allowed companies to replace individual plant parts without automatically triggering the permitting requirements that apply to new sources. Chevron, 467 U. S., at 840.

This Court upheld the EPA’s definition as consistent with the governing statute. Id., at 866. The decision, issued by a bare quorum of the Court, without concurrence or dissent, purported to apply “well-settled principles.” Id., at 845. “If a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue,” Chevron provided, then “that intention is the law and must be given effect.” Id., at 843, n. 9. Many of the cases Chevron cited to support its judgment stood for the traditional proposition that courts afford respectful consideration, not deference, to executive interpretations of the law. See, e.g., Burnet, 285 U. S., at 16; United States v. Moore, 95 U.S. 760, 763 (1878). And the decision’s sole citation to legal scholarship was to Roscoe Pound, who long championed de novo judicial review. 467 U. S., at 843, n. 10; see R. Pound, The Place of the Judiciary in a Democratic Polity, 27 A. B. A. J. 133, 136–137 (1941).

At the same time, of course, the opinion contained bits and pieces that spoke differently. The decision also said that, “if [a] statute is silent or ambiguous with respect to [a] specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute.” 467 U. S., at 843. But it seems the government didn’t advance this formulation in its brief, so there was no adversarial engagement on it. T. Merrill, The Story of Chevron: The Making of an Accidental Landmark, 66 Admin. L. Rev. 253, 268 (2014) (Merrill). As we have seen, too, the Court did not pause to consider (or even mention) the APA. See Part II–A, supra. It did not discuss contrary precedents issued by the Court since the founding, let alone purport to overrule any of them. See Part II–B–1, supra. Nor did the Court seek to address how its novel rule of deference might be squared with so much surrounding law. See Part II–B–2, supra. As even its defenders have acknowledged, “Chevron barely bothered to justify its rule of deference, and the few brief passages on this matter pointed in disparate directions.” Kagan 212–213. “[T]he quality of the reasoning,” they acknowledge, “was not high,” C. Sunstein, Chevron as Law, 107 Geo. L. J. 1613, 1669 (2019).

If Chevron meant to usher in a revolution in how judges interpret laws, no one appears to have realized it at the time. Chevron’s author, Justice Stevens, characterized the decision as a “simpl[e] . . . restatement of existing law, nothing more or less.” Merrill 255, 275. In the “19 argued cases” in the following Term “that presented some kind of question about whether the Court should defer to an agency interpretation of statutory law,” this Court cited Chevron just once. Merrill 276. By some accounts, the decision seemed “destined to obscurity.” Ibid.

It was only three years later when Justice Scalia wrote a concurrence that a revolution began to take shape. Buffington, 598 U. S., at ___ (slip op., at 8). There, he argued for a new rule requiring courts to defer to executive agency interpretations of the law whenever a “ ‘statute is silent or ambiguous.’ ” NLRB v. Food & Commercial Workers, 484 U.S. 112, 133–134 (1987) (opinion of Scalia, J.). Eventually, a majority of the Court followed his lead. Buffington, 598 U. S., at ___ (slip op., at 8). But from the start, Justice Scalia made no secret about the scope of his ambitions. See Judicial Deference to Administrative Interpretations of Law, 1989 Duke L. J. 511, 521 (1989) (Scalia). The rule he advocated for represented such a sharp break from prior practice, he explained, that many judges of his day didn’t yet “understand” the “old criteria” were “no longer relevant.” Ibid. Still, he said, overthrowing the past was worth it because a new deferential rule would be “easier to follow.” Ibid.

Events proved otherwise. As the years wore on and the Court’s new and aggressive reading of Chevron gradually exposed itself as unworkable, unfair, and at odds with our separation of powers, Justice Scalia could have doubled down on the project. But he didn’t. He appreciated that stare decisis is not a rule of “if I thought it yesterday, I must think it tomorrow.” And rather than cling to the pride of personal precedent, the Justice began to express doubts over the very project that he had worked to build. See Perez v. Mortgage Bankers Assn., 575 U.S. 92, 109–110 (2015) (opinion concurring in judgment); cf. Decker v. Northwest Environmental Defense Center, 568 U.S. 597, 617–618, 621 (2013) (opinion concurring in part and dissenting in part). If Chevron’s ascent is a testament to the Justice’s ingenuity, its demise is an even greater tribute to his humility.[6]

Justice Scalia was not alone in his reconsideration. After years spent laboring under Chevron, trying to make sense of it and make it work, Member after Member of this Court came to question the project. See, e.g., Pereira v. Sessions, 585 U.S. 198, 219–221 (2018) (Kennedy, J., concurring); Michigan v. EPA, 576 U.S. 743, 760–764 (2015) (Thomas, J., concurring); Kisor, 588 U. S., at 591 (Roberts, C. J., concurring in part); Gutierrez-Brizuela, 834 F. 3d, at 1153; Buffington, 598 U. S., at ___–___ (slip op., at 14–15); Kavanaugh 2150–2154. Ultimately, the Court gave up. Despite repeated invitations, it has not applied Chevron deference since 2016. Relentless Tr. 81; App. to Brief for Respondents in No. 22–1219, p. 68a. So an experiment that began only in the mid-1980s effectively ended eight years ago. Along the way, an unusually large number of federal appellate judges voiced their own thoughtful and extensive criticisms of Chevron. Buffington, 598 U. S., at ___–___ (slip op., at 14–15) (collecting examples). A number of state courts did, too, refusing to import Chevron deference into their own administrative law jurisprudence. See 598 U. S., at ___ (slip op., at 15).

Even if all that and everything else laid out above is true, the government suggests we should retain Chevron deference because judges simply cannot live without it; some statutes are just too “technical” for courts to interpret “intelligently.” Post, at 9, 32 (dissenting opinion). But that objection is no answer to Chevron’s inconsistency with Congress’s directions in the APA, so much surrounding law, or the challenges its multistep regime have posed in practice. Nor does history counsel such defeatism. Surely, it would be a mistake to suggest our predecessors before Chevron’s rise in the mid-1980s were unable to make their way intelligently through technical statutory disputes. Following their lead, over the past eight years this Court has managed to resolve even highly complex cases without Chevron deference, and done so even when the government sought deference. Nor, as far as I am aware, did any Member of the Court suggest Chevron deference was necessary to an intelligent resolution of any of those matters.[7] If anything, by affording Chevron deference a period of repose before addressing whether it should be retained, the Court has enabled its Members to test the propriety of that precedent and reflect more deeply on how well it fits into the broader architecture of our law. Others may see things differently, see post, at 26–27 (dissenting opinion), but the caution the Court has exhibited before overruling Chevron may illustrate one of the reasons why the current Court has been slower to overrule precedents than some of its predecessors, see Part I–C, supra.

None of this, of course, discharges any Member of this Court from the task of deciding for himself or herself today whether Chevron deference itself warrants deference. But when so many past and current judicial colleagues in this Court and across the country tell us our doctrine is misguided, and when we ourselves managed without Chevron for centuries and manage to do so today, the humility at the core of stare decisis compels us to pause and reflect carefully on the wisdom embodied in that experience. And, in the end, to my mind the lessons of experience counsel wisely against continued reliance on Chevron’s stray and unconsidered digression. This Court’s opinions fill over 500 volumes, and perhaps “some printed judicial word may be found to support almost any plausible proposition.” R. Jackson, Decisional Law and Stare Decisis, 30 A. B. A. J. 334 (1944). It is not for us to pick and choose passages we happen to like and demand total obedience to them in perpetuity. That would turn stare decisis from a doctrine of humility into a tool for judicial opportunism. Brown, 596 U. S., at 141.

III

Proper respect for precedent helps “keep the scale of justice even and steady,” by reinforcing decisional rules consistent with the law upon which all can rely. 1 Blackstone 69. But that respect does not require, nor does it readily tolerate, a steadfast refusal to correct mistakes. As early as 1810, this Court had already overruled one of its cases. See Hudson v. Guestier, 6 Cranch 281, 284 (overruling Rose v. Himely, 4 Cranch 241 (1808)). In recent years, the Court may have overruled precedents less frequently than it did during the Warren and Burger Courts. See Part I–C, supra. But the job of reconsidering past decisions remains one every Member of this Court faces from time to time.[8]

Justice William O. Douglas served longer on this Court than any other person in the Nation’s history. During his tenure, he observed how a new colleague might be inclined initially to “revere” every word written in an opinion issued before he arrived. W. Douglas, Stare Decisis, 49 Colum. L. Rev. 735, 736 (1949). But, over time, Justice Douglas reflected, his new colleague would “remembe[r] . . . that it is the Constitution which he swore to support and defend, not the gloss which his predecessors may have put on it.” Ibid. And “[s]o he [would] com[e] to formulate his own views, rejecting some earlier ones as false and embracing others.” Ibid. This process of reexamination, Justice Douglas explained, is a “necessary consequence of our system” in which each judge takes an oath—both “personal” and binding—to discern the law’s meaning for himself and apply it faithfully in the cases that come before him. Id., at 736–737.

Justice Douglas saw, too, how appeals to precedent could be overstated and sometimes even overwrought. Judges, he reflected, would sometimes first issue “new and startling decision[s],” and then later spin around and “acquire an acute conservatism” in their aggressive defense of “their new status quo.” Id., at 737. In that way, even the most novel and unlikely decisions became “coveted anchorage[s],” defended heatedly, if ironically, under the banner of “stare decisis.” Ibid.; see also Edwards v. Vannoy, 593 U.S. 255, 294, n. 7 (2021) (Gorsuch, J., concurring).

That is Chevron’s story: A revolution masquerading as the status quo. And the defense of it follows the same course Justice Douglas described. Though our dissenting colleagues have not hesitated to question other precedents in the past, they today manifest what Justice Douglas called an “acute conservatism” for Chevron’s “startling” development, insisting that if this “coveted anchorage” is abandoned the heavens will fall. But the Nation managed to live with busy executive agencies of all sorts long before the Chevron revolution began to take shape in the mid-1980s. And all today’s decision means is that, going forward, federal courts will do exactly as this Court has since 2016, exactly as it did before the mid-1980s, and exactly as it had done since the founding: resolve cases and controversies without any systemic bias in the government’s favor.

Proper respect for precedent does not begin to suggest otherwise. Instead, it counsels respect for the written law, adherence to consistent teachings over aberrations, and resistance to the temptation of treating our own stray remarks as if they were statutes. And each of those lessons points toward the same conclusion today: Chevron deference is inconsistent with the directions Congress gave us in the APA. It represents a grave anomaly when viewed against the sweep of historic judicial practice. The decision undermines core rule-of-law values ranging from the promise of fair notice to the promise of a fair hearing. Even on its own terms, it has proved unworkable and operated to undermine rather than advance reliance interests, often to the detriment of ordinary Americans. And from the start, the whole project has relied on the overaggressive use of snippets and stray remarks from an opinion that carried mixed messages. Stare decisis’s true lesson today is not that we are bound to respect Chevron’s “startling development,” but bound to inter it.

SUPREME COURT OF THE UNITED STATES

_________________

Nos. 22–451 and 22–1219

_________________

LOPER BRIGHT ENTERPRISES, et al., PETITIONERS

22–451v.

GINA RAIMONDO, SECRETARY OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the district of columbia circuit

RELENTLESS, INC., et al., PETITIONERS

22–1219v.

DEPARTMENT OF COMMERCE, et al.

on writ of certiorari to the united states court of appeals for the first circuit

[June 28, 2024]

Justice Kagan, with whom Justice Sotomayor and Justice Jackson join,[1]* dissenting.

For 40 years, Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), has served as a cornerstone of administrative law, allocating responsibility for statutory construction between courts and agencies. Under Chevron, a court uses all its normal interpretive tools to determine whether Congress has spoken to an issue. If the court finds Congress has done so, that is the end of the matter; the agency’s views make no difference. But if the court finds, at the end of its interpretive work, that Congress has left an ambiguity or gap, then a choice must be made. Who should give content to a statute when Congress’s instructions have run out? Should it be a court? Or should it be the agency Congress has charged with administering the statute? The answer Chevron gives is that it should usually be the agency, within the bounds of reasonableness. That rule has formed the backdrop against which Congress, courts, and agencies—as well as regulated parties and the public—all have operated for decades. It has been applied in thousands of judicial decisions. It has become part of the warp and woof of modern government, supporting regulatory efforts of all kinds—to name a few, keeping air and water clean, food and drugs safe, and financial markets honest.

And the rule is right. This Court has long understood Chevron deference to reflect what Congress would want, and so to be rooted in a presumption of legislative intent. Congress knows that it does not—in fact cannot—write perfectly complete regulatory statutes. It knows that those statutes will inevitably contain ambiguities that some other actor will have to resolve, and gaps that some other actor will have to fill. And it would usually prefer that actor to be the responsible agency, not a court. Some interpretive issues arising in the regulatory context involve scientific or technical subject matter. Agencies have expertise in those areas; courts do not. Some demand a detailed understanding of complex and interdependent regulatory programs. Agencies know those programs inside-out; again, courts do not. And some present policy choices, including trade-offs between competing goods. Agencies report to a President, who in turn answers to the public for his policy calls; courts have no such accountability and no proper basis for making policy. And of course Congress has conferred on that expert, experienced, and politically accountable agency the authority to administer—to make rules about and otherwise implement—the statute giving rise to the ambiguity or gap. Put all that together and deference to the agency is the almost obvious choice, based on an implicit congressional delegation of interpretive authority. We defer, the Court has explained, “because of a presumption that Congress” would have “desired the agency (rather than the courts)” to exercise “whatever degree of discretion” the statute allows. Smiley v. Citibank (South Dakota), N. A., 517 U.S. 735, 740–741 (1996).

Today, the Court flips the script: It is now “the courts (rather than the agency)” that will wield power when Congress has left an area of interpretive discretion. A rule of judicial humility gives way to a rule of judicial hubris. In recent years, this Court has too often taken for itself decision-making authority Congress assigned to agencies. The Court has substituted its own judgment on workplace health for that of the Occupational Safety and Health Administration; its own judgment on climate change for that of the Environmental Protection Agency; and its own judgment on student loans for that of the Department of Education. See, e.g., National Federation of Independent Business v. OSHA, 595 U.S. 109 (2022); West Virginia v. EPA, 597 U.S. 697 (2022); Biden v. Nebraska, 600 U.S. 477 (2023). But evidently that was, for this Court, all too piecemeal. In one fell swoop, the majority today gives itself exclusive power over every open issue—no matter how expertise-driven or policy-laden—involving the meaning of regulatory law. As if it did not have enough on its plate, the majority turns itself into the country’s administrative czar. It defends that move as one (suddenly) required by the (nearly 80-year-old) Administrative Procedure Act. But the Act makes no such demand. Today’s decision is not one Congress directed. It is entirely the majority’s choice.

And the majority cannot destroy one doctrine of judicial humility without making a laughing-stock of a second. (If opinions had titles, a good candidate for today’s would be Hubris Squared.) Stare decisis is, among other things, a way to remind judges that wisdom often lies in what prior judges have done. It is a brake on the urge to convert “every new judge’s opinion” into a new legal rule or regime. Dobbs v. Jackson Women’s Health Organization, 597 U.S. 215, 388 (2022) (joint opinion of Breyer, Sotomayor, and Kagan, JJ., dissenting) (quoting 1 W. Blackstone, Commentaries on the Laws of England 69 (7th ed. 1775)). Chevron is entrenched precedent, entitled to the protection of stare decisis, as even the majority acknowledges. In fact, Chevron is entitled to the supercharged version of that doctrine because Congress could always overrule the decision, and because so many governmental and private actors have relied on it for so long. Because that is so, the majority needs a “particularly special justification” for its action. Kisor v. Wilkie, 588 U.S. 558, 588 (2019) (opinion of the Court). But the majority has nothing that would qualify. It barely tries to advance the usual factors this Court invokes for overruling precedent. Its justification comes down, in the end, to this: Courts must have more say over regulation—over the provision of health care, the protection of the environment, the safety of consumer products, the efficacy of transportation systems, and so on. A longstanding precedent at the crux of administrative governance thus falls victim to a bald assertion of judicial authority. The majority disdains restraint, and grasps for power.

I

Begin with the problem that gave rise to Chevron (and also to its older precursors): The regulatory statutes Congress passes often contain ambiguities and gaps. Sometimes they are intentional. Perhaps Congress “consciously desired” the administering agency to fill in aspects of the legislative scheme, believing that regulatory experts would be “in a better position” than legislators to do so. Chevron, 467 U. S., at 865. Or “perhaps Congress was unable to forge a coalition on either side” of a question, and the contending parties “decided to take their chances with” the agency’s resolution. Ibid. Sometimes, though, the gaps or ambiguities are what might be thought of as predictable accidents. They may be the result of sloppy drafting, a not infrequent legislative occurrence. Or they may arise from the well-known limits of language or foresight. Accord, ante, at 7, 22. “The subject matter” of a statutory provision may be too “specialized and varying” to “capture in its every detail.” Kisor, 588 U. S., at 566 (plurality opinion). Or the provision may give rise, years or decades down the road, to an issue the enacting Congress could not have anticipated. Whichever the case—whatever the reason—the result is to create uncertainty about some aspect of a provision’s meaning.

Consider a few examples from the caselaw. They will help show what a typical Chevron question looks like—or really, what a typical Chevron question is. Because when choosing whether to send some class of questions mainly to a court, or mainly to an agency, abstract analysis can only go so far; indeed, it may obscure what matters most. So I begin with the concrete:

Under the Public Health Service Act, the Food and Drug Administration (FDA) regulates “biological product[s],” including “protein[s].” 42 U. S. C. §262(i)(1). When does an alpha amino acid polymer qualify as such a “protein”? Must it have a specific, defined sequence of amino acids? See Teva Pharmaceuticals USA, Inc. v. FDA, 514 F. Supp. 3d 66, 79–80, 93–106 (DC 2020).

Under the Endangered Species Act, the Fish and Wildlife Service must designate endangered “vertebrate fish or wildlife” species, including “distinct population segment[s]” of those species. 16 U. S. C. §1532(16); see §1533. What makes one population segment “distinct” from another? Must the Service treat the Washington State population of western gray squirrels as “distinct”

00because it is geographically separated from other western gray squirrels? Or can the Service take into account that the genetic makeup of the Washington population does not differ markedly from the rest? See Northwest Ecosystem Alliance v. United States Fish and Wildlife Serv., 475 F.3d 1136, 1140–1145, 1149 (CA9 2007).

Under the Medicare program, reimbursements to hospitals are adjusted to reflect “differences in hospital wage levels” across “geographic area[s].” 42 U. S. C. §1395ww(d)(3)(E)(i). How should the Department of Health and Human Services measure a “geographic area”? By city? By county? By metropolitan area? See Bellevue Hospital Center v. Leavitt, 443 F.3d 163, 174–176 (CA2 2006).

Congress directed the Department of the Interior and the Federal Aviation Administration to reduce noise from aircraft flying over Grand Canyon National Park—specifically, to “provide for substantial restoration of the natural quiet.” §3(b)(1), 101 Stat. 676; see §3(b)(2). How much noise is consistent with “the natural quiet”? And how much of the park, for how many hours a day, must be that quiet for the “substantial restoration” requirement to be met? See Grand Canyon Air Tour Coalition v. FAA, 154 F.3d 455, 466–467, 474–475 (CADC 1998).

Or take Chevron itself. In amendments to the Clean Air Act, Congress told States to require permits for modifying or constructing “stationary sources” of air pollution. 42 U. S. C. §7502(c)(5). Does the term “stationary source[ ]” refer to each pollution-emitting piece of equipment within a plant? Or does it refer to the entire plant, and thus allow escape from the permitting requirement when increased emissions from one piece of equipment are offset by reductions from another? See 467 U. S., at 857, 859.

In each case, a statutory phrase has more than one reasonable reading. And Congress has not chosen among them: It has not, in any real-world sense, “fixed” the “single, best meaning” at “the time of enactment” (to use the majority’s phrase). Ante, at 22. A question thus arises: Who decides which of the possible readings should govern?

This Court has long thought that the choice should usually fall to agencies, with courts broadly deferring to their judgments. For the last 40 years, that doctrine has gone by the name of Chevron deference, after the 1984 decision that formalized and canonized it. In Chevron, the Court set out a simple two-part framework for reviewing an agency’s interpretation of a statute that it administers. First, the reviewing court must determine whether Congress has “directly spoken to the precise question at issue.” 467 U. S., at 842. That inquiry is rigorous: A court must exhaust all the “traditional tools of statutory construction” to divine statutory meaning. Id., at 843, n. 9. And when it can find that meaning—a “single right answer”—that is “the end of the matter”: The court cannot defer because it “must give effect to the unambiguously expressed intent of Congress.” Kisor, 588 U. S., at 575 (opinion of the Court); Chevron, 467 U. S., at 842–843. But if the court, after using its whole legal toolkit, concludes that “the statute is silent or ambiguous with respect to the specific issue” in dispute—for any of the not-uncommon reasons discussed above—then the court must cede the primary interpretive role. Ibid.; see supra, at 4–5. At that second step, the court asks only whether the agency construction is within the sphere of “reasonable” readings. Chevron, 467 U. S., at 844. If it is, the agency’s interpretation of the statute that it every day implements will control.

That rule, the Court has long explained, rests on a presumption about legislative intent—about what Congress wants when a statute it has charged an agency with implementing contains an ambiguity or a gap. See id., at 843–845; Smiley, 517 U. S., at 740–741. An enacting Congress, as noted above, knows those uncertainties will arise, even if it does not know what they will turn out to be. See supra, at 4–5. And every once in a while, Congress provides an explicit instruction for dealing with that contingency—assigning primary responsibility to the courts, or else to an agency. But much more often, Congress does not say. Thus arises the need for a presumption—really, a default rule—for what should happen in that event. Does a statutory silence or ambiguity then go to a court for resolution? Or to an agency? This Court has long thought Congress would choose an agency, with courts serving only as a backstop to make sure the agency makes a reasonable choice among the possible readings. Or said otherwise, Congress would select the agency it has put in control of a regulatory scheme to exercise the “degree of discretion” that the statute’s lack of clarity or completeness allows. Smiley, 517 U. S., at 741. Of course, Congress can always refute that presumptive choice—can say that, really, it would prefer courts to wield that discretionary power. But until then, the presumption cuts in the agency’s favor.[2] The next question is why.

For one, because agencies often know things about a statute’s subject matter that courts could not hope to. The point is especially stark when the statute is of a “scientific or technical nature.” Kisor, 588 U. S., at 571 (plurality opinion). Agencies are staffed with “experts in the field” who can bring their training and knowledge to bear on open statutory questions. Chevron, 467 U. S., at 865. Consider, for example, the first bulleted case above. When does an alpha amino acid polymer qualify as a “protein”? See supra, at 5. I don’t know many judges who would feel confident resolving that issue. (First question: What even is an alpha amino acid polymer?) But the FDA likely has scores of scientists on staff who can think intelligently about it, maybe collaborate with each other on its finer points, and arrive at a sensible answer. Or take the perhaps more accessible-sounding second case, involving the Endangered Species Act. See supra, at 5–6. Deciding when one squirrel population is “distinct” from another (and thus warrants protection) requires knowing about species more than it does consulting a dictionary. How much variation of what kind—geographic, genetic, morphological, or behavioral—should be required? A court could, if forced to, muddle through that issue and announce a result. But wouldn’t the Fish and Wildlife Service, with all its specialized expertise, do a better job of the task—of saying what, in the context of species protection, the open-ended term “distinct” means? One idea behind the Chevron presumption is that Congress—the same Congress that charged the Service with implementing the Act—would answer that question with a resounding “yes.”

A second idea is that Congress would value the agency’s experience with how a complex regulatory regime functions, and with what is needed to make it effective. Let’s stick with squirrels for a moment, except broaden the lens. In construing a term like “distinct” in a case about squirrels, the Service likely would benefit from its “historical familiarity” with how the term has covered the population segments of other species. Martin v. Occupational Safety and Health Review Comm’n, 499 U.S. 144, 153 (1991); see, e.g., Center for Biological Diversity v. Zinke, 900 F.3d 1053, 1060–1062 (CA9 2018) (arctic grayling); Center for Biological Diversity v. Zinke, 868 F.3d 1054, 1056 (CA9 2017) (desert eagle). Just as a common-law court makes better decisions as it sees multiple variations on a theme, an agency’s construction of a statutory term benefits from its unique exposure to all the related ways the term comes into play. Or consider, for another way regulatory familiarity matters, the example about adjusting Medicare reimbursement for geographic wage differences. See supra, at 6. According to a dictionary, the term “geographic area” could be as large as a multi-state region or as small as a census tract. How to choose? It would make sense to gather hard information about what reimbursement levels each approach will produce, to explore the ease of administering each on a nationwide basis, to survey how regulators have dealt with similar questions in the past, and to confer with the hospitals themselves about what makes sense. See Kisor, 588 U. S., at 571 (plurality opinion) (noting that agencies are able to “conduct factual investigations” and “consult with affected parties”). Congress knows the Department of Health and Human Services can do all those things—and that courts cannot.

Still more, Chevron’s presumption reflects that resolving statutory ambiguities, as Congress well knows, is “often more a question of policy than of law.” Pauley v. BethEnergy Mines, Inc., 501 U.S. 680, 696 (1991). The task is less one of construing a text than of balancing competing goals and values. Consider the statutory directive to achieve “substantial restoration of the [Grand Canyon’s] natural quiet.” See supra, at 6. Someone is going to have to decide exactly what that statute means for air traffic over the canyon. How many flights, in what places and at what times, are consistent with restoring enough natural quiet on the ground? That is a policy trade-off of a kind familiar to agencies—but peculiarly unsuited to judges. Or consider Chevron itself. As the Court there understood, the choice between defining a “stationary source” as a whole plant or as a pollution-emitting device is a choice about how to “reconcile” two “manifestly competing interests.” 467 U. S., at 865. The plantwide definition relaxes the permitting requirement in the interest of promoting economic growth; the device-specific definition strengthens that requirement to better reduce air pollution. See id., at 851, 863, 866. Again, that is a choice a judge should not be making, but one an agency properly can. Agencies are “subject to the supervision of the President, who in turn answers to the public.” Kisor, 588 U. S., at 571–572 (plurality opinion). So when faced with a statutory ambiguity, “an agency to which Congress has delegated policymaking responsibilities” may rely on an accountable actor’s “views of wise policy to inform its judgments.” Chevron, 467 U. S., at 865.

None of this is to say that deference to agencies is always appropriate. The Court over time has fine-tuned the Chevron regime to deny deference in classes of cases in which Congress has no reason to prefer an agency to a court. The majority treats those “refinements” as a flaw in the scheme, ante, at 27, but they are anything but. Consider the rule that an agency gets no deference when construing a statute it is not responsible for administering. See Epic Systems Corp. v. Lewis, 584 U.S. 497, 519–520 (2018). Well, of course not—if Congress has not put an agency in charge of implementing a statute, Congress would not have given the agency a special role in its construction. Or take the rule that an agency will not receive deference if it has reached its decision without using—or without using properly—its rulemaking or adjudicatory authority. See United States v. Mead Corp., 533 U.S. 218, 226–227 (2001); Encino Motorcars, LLC v. Navarro, 579 U.S. 211, 220 (2016). Again, that should not be surprising: Congress expects that authoritative pronouncements on a law’s meaning will come from the procedures it has enacted to foster “fairness and deliberation” in agency decision-making. Mead, 533 U. S., at 230. Or finally, think of the “extraordinary cases” involving questions of vast “economic and political significance” in which the Court has declined to defer. King v. Burwell, 576 U.S. 473, 485–486 (2015). The theory is that Congress would not have left matters of such import to an agency, but would instead have insisted on maintaining control. So the Chevron refinements proceed from the same place as the original doctrine. Taken together, they give interpretive primacy to the agency when—but only when—it is acting, as Congress specified, in the heartland of its delegated authority.

That carefully calibrated framework “reflects a sensitivity to the proper roles of the political and judicial branches.” Pauley, 501 U. S., at 696. Where Congress has spoken, Congress has spoken; only its judgments matter. And courts alone determine when that has happened: Using all their normal interpretive tools, they decide whether Congress has addressed a given issue. But when courts have decided that Congress has not done so, a choice arises. Absent a legislative directive, either the administering agency or a court must take the lead. And the matter is more fit for the agency. The decision is likely to involve the agency’s subject-matter expertise; to fall within its sphere of regulatory experience; and to involve policy choices, including cost-benefit assessments and trade-offs between conflicting values. So a court without relevant expertise or experience, and without warrant to make policy calls, appropriately steps back. The court still has a role to play: It polices the agency to ensure that it acts within the zone of reasonable options. But the court does not insert itself into an agency’s expertise-driven, policy-laden functions. That is the arrangement best suited to keep every actor in its proper lane. And it is the one best suited to ensure that Congress’s statutes work in the way Congress intended.

The majority makes two points in reply, neither convincing. First, it insists that “agencies have no special competence” in filling gaps or resolving ambiguities in regulatory statutes; rather, “[c]ourts do.” Ante, at 23. Score one for self-confidence; maybe not so high for self-reflection or -knowledge. Of course courts often construe legal texts, hopefully well. And Chevron’s first step takes full advantage of that talent: There, a court tries to divine what Congress meant, even in the most complicated or abstruse statutory schemes. The deference comes in only if the court cannot do so—if the court must admit that standard legal tools will not avail to fill a statutory silence or give content to an ambiguous term. That is when the issues look like the ones I started off with: When does an alpha amino acid polymer qualify as a “protein”? How distinct is “distinct” for squirrel populations? What size “geographic area” will ensure appropriate hospital reimbursement? As between two equally feasible understandings of “stationary source,” should one choose the one more protective of the environment or the one more favorable to economic growth? The idea that courts have “special competence” in deciding such questions whereas agencies have “no[ne]” is, if I may say, malarkey. Answering those questions right does not mainly demand the interpretive skills courts possess. Instead, it demands one or more of: subject-matter expertise, long engagement with a regulatory scheme, and policy choice. It is courts (not agencies) that “have no special competence”—or even legitimacy—when those are the things a decision calls for.

Second, the majority complains that an ambiguity or gap does not “necessarily reflect a congressional intent that an agency” should have primary interpretive authority. Ante, at 22. On that score, I’ll agree with the premise: It doesn’t “necessarily” do so. Chevron is built on a presumption. The decision does not maintain that Congress in every case wants the agency, rather than a court, to fill in gaps. The decision maintains that when Congress does not expressly pick one or the other, we need a default rule; and the best default rule—agency or court?—is the one we think Congress would generally want. As to why Congress would generally want the agency: The answer lies in everything said above about Congress’s delegation of regulatory power to the agency and the agency’s special competencies. See supra, at 9–11. The majority appears to think it is a showstopping rejoinder to note that many statutory gaps and ambiguities are “unintentional.” Ante, at 22. But to begin, many are not; the ratio between the two is uncertain. See supra, at 4–5. And to end, why should that matter in any event? Congress may not have deliberately introduced a gap or ambiguity into the statute; but it knows that pretty much everything it drafts will someday be found to contain such a “flaw.” Given that knowledge, Chevron asks, what would Congress want? The presumed answer is again the same (for the same reasons): The agency. And as with any default rule, if Congress decides otherwise, all it need do is say.

In that respect, the proof really is in the pudding: Congress basically never says otherwise, suggesting that Chevron chose the presumption aligning with legislative intent (or, in the majority’s words, “approximat[ing] reality,” ante, at 22). Over the last four decades, Congress has authorized or reauthorized hundreds of statutes. The drafters of those statutes knew all about Chevron. See A. Gluck & L. Bressman, Statutory Interpretation From the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 928 (fig. 2), 994 (2013). So if they had wanted a different assignment of interpretive responsibility, they would have inserted a provision to that effect. With just a pair of exceptions I know of, they did not. See 12 U. S. C. §25b(b)(5)(A) (exception #1); 15 U. S. C. §8302(c)(3)(A) (exception #2). Similarly, Congress has declined to enact proposed legislation that would abolish Chevron across the board. See S. 909, 116th Cong., 1st Sess., §2 (2019) (still a bill, not a law); H. R. 5, 115th Cong., 1st Sess., §202 (2017) (same). So to the extent the majority is worried that the Chevron presumption is “fiction[al],” ante, at 26—as all legal presumptions in some sense are—it has gotten less and less so every day for 40 years. The congressional reaction shows as well as anything could that the Chevron Court read Congress right.

II

The majority’s principal arguments are in a different vein. Around 80 years after the APA was enacted and 40 years after Chevron, the majority has decided that the former precludes the latter. The APA’s Section 706, the majority says, “makes clear” that agency interpretations of statutes “are not entitled to deference.” Ante, at 14–15 (emphasis in original). And that provision, the majority continues, codified the contemporaneous law, which likewise did not allow for deference. See ante, at 9–13, 15–16. But neither the APA nor the pre-APA state of the law does the work that the majority claims. Both are perfectly compatible with Chevron deference.

Section 706, enacted with the rest of the APA in 1946, provides for judicial review of agency action. It states: “To the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action.” 5 U. S. C. §706.

That text, contra the majority, “does not resolve the Chevron question.” C. Sunstein, Chevron As Law, 107 Geo. L. J. 1613, 1642 (2019) (Sunstein). Or said a bit differently, Section 706 is “generally indeterminate” on the matter of deference. A. Vermeule, Judging Under Uncertainty 207 (2006) (Vermeule). The majority highlights the phrase “decide all relevant questions of law” (italicizing the “all”), and notes that the provision “prescribes no deferential standard” for answering those questions. Ante, at 14. But just as the provision does not prescribe a deferential standard of review, so too it does not prescribe a de novo standard of review (in which the court starts from scratch, without giving deference). In point of fact, Section 706 does not specify any standard of review for construing statutes. See Kisor, 588 U. S., at 581 (plurality opinion). And when a court uses a deferential standard—here, by deciding whether an agency reading is reasonable—it just as much “decide[s]” a “relevant question[ ] of law” as when it uses a de novo standard. §706. The deferring court then conforms to Section 706 “by determining whether the agency has stayed within the bounds of its assigned discretion—that is, whether the agency has construed [the statute it administers] reasonably.” J. Manning, Chevron and the Reasonable Legislator, 128 Harv. L. Rev. 457, 459 (2014); see Arlington v. FCC, 569 U.S. 290, 317 (2013) (Roberts, C. J., dissenting) (“We do not ignore [Section 706’s] command when we afford an agency’s statutory interpretation Chevron deference; we respect it”).[3]

Section 706’s references to standards of review in other contexts only further undercut the majority’s argument. The majority notes that Section 706 requires deferential review for agency fact-finding and policy-making (under, respectively, a substantial-evidence standard and an arbitrary-and-capricious standard). See ante, at 14. Congress, the majority claims, “surely would have articulated a similarly deferential standard applicable to questions of law had it intended to depart” from de novo review. Ibid. Surely? In another part of Section 706, Congress explicitly referred to de novo review. §706(2)(F). With all those references to standards of review—both deferential and not—running around Section 706, what is “telling” (ante, at 14) is the absence of any standard for reviewing an agency’s statutory constructions. That silence left the matter, as noted above, “generally indeterminate”: Section 706 neither mandates nor forbids Chevron-style deference. Vermeule 207.[4]

And contra the majority, most “respected commentators” understood Section 706 in that way—as allowing, even if not requiring, deference. Ante, at 16. The finest administrative law scholars of the time (call them that generation’s Manning, Sunstein, and Vermeule) certainly did. Professor Louis Jaffe described something very like the Chevron two-step as the preferred method of reviewing agency interpretations under the APA. A court, he said, first “must decide as a ‘question of law’ whether there is ‘discretion’ in the premises.” Judicial Control of Administrative Action 570 (1965). That is akin to step 1: Did Congress speak to the issue, or did it leave openness? And if the latter, Jaffe continued, the agency’s view “if ‘reasonable’ is free of control.” Ibid. That of course looks like step 2: defer if reasonable. And just in case that description was too complicated, Jaffe conveyed his main point this way: The argument that courts “must decide all questions of law”—as if there were no agency in the picture—“is, in my opinion, unsound.” Id., at 569. Similarly, Professor Kenneth Culp Davis, author of the then-preeminent treatise on administrative law, noted with approval that “reasonableness” review of agency interpretations—in which courts “refused to substitute judgment”—had “survived the APA.” Administrative Law 880, 883, 885 (1951) (Davis). Other contemporaneous scholars and experts agreed. See R. Levin, The APA and the Assault on Deference, 106 Minn. L. Rev. 125, 181–183 (2021) (Levin) (listing many of them). They did not see in their own time what the majority finds there today.[5]

Nor, evidently, did the Supreme Court. In the years after the APA was enacted, the Court “never indicated that section 706 rejected the idea that courts might defer to agency interpretations of law.” Sunstein 1654. Indeed, not a single Justice so much as floated that view of the APA. To the contrary, the Court issued a number of decisions in those years deferring to an agency’s statutory interpretation. See, e.g., Unemployment Compensation Comm’n of Alaska v. Aragon, 329 U.S. 143, 153–154 (1946); NLRB v. E. C. Atkins & Co., 331 U.S. 398, 403 (1947); Cardillo v. Liberty Mut. Ins. Co., 330 U.S. 469, 478–479 (1947). And that continued right up until Chevron. See, e.g., Mitchell v. Budd, 350 U.S. 473, 480 (1956); Zenith Radio Corp. v. United States, 437 U.S. 443, 450 (1978). To be clear: Deference in those years was not always given to interpretations that would receive it under Chevron. The practice then was more inconsistent and less fully elaborated than it later became. The point here is only that the Court came nowhere close to accepting the majority’s view of the APA. Take the language from Section 706 that the majority most relies on: “decide all relevant questions of law.” See ante, at 14. In the decade after the APA’s enactment, those words were used only four times in Supreme Court opinions (all in footnotes)—and never to suggest that courts could not defer to agency interpretations. See Sunstein 1656.

The majority’s view of Section 706 likewise gets no support from how judicial review operated in the years leading up to the APA. That prior history matters: As the majority recognizes, Section 706 was generally understood to “restate[ ] the present law as to the scope of judicial review.” Dept. of Justice, Attorney General’s Manual on the Administrative Procedure Act 108 (1947); ante, at 15–16. The problem for the majority is that in the years preceding the APA, courts became ever more deferential to agencies. New Deal administrative programs had by that point come into their own. And this Court and others, in a fairly short time, had abandoned their initial resistance and gotten on board. Justice Breyer, wearing his administrative-law-scholar hat, characterized the pre-APA period this way: “[J]udicial review of administrative action was curtailed, and particular agency decisions were frequently sustained with judicial obeisance to the mysteries of administrative expertise.” S. Breyer et al., Administrative Law and Regulatory Policy 21 (7th ed. 2011). And that description extends to review of an agency’s statutory constructions. An influential study of administrative practice, published five years before the APA’s enactment, described the state of play: Judicial “review may, in some instances at least, be limited to the inquiry whether the administrative construction is a permissible one.” Final Report of Attorney General’s Committee on Administrative Procedure (1941), reprinted in Administrative Procedure in Government Agencies, S. Doc. No. 8, 77th Cong., 1st Sess., 78 (1941). Or again: “[W]here the statute is reasonably susceptible of more than one interpretation, the court may accept that of the administrative body.” Id., at 90–91.[6]

Two prominent Supreme Court decisions of the 1940s put those principles into action. Gray v. Powell, 314 U.S. 402 (1941), was then widely understood as “the leading case” on review of agency interpretations. Davis 882; see ibid. (noting that it “establish[ed] what is known as ‘the doctrine of Gray v. Powell’ ”). There, the Court deferred to an agency construction of the term “producer” as used in a statutory exemption from price controls. Congress, the Court explained, had committed the scope of the exemption to the agency because its “experience in [the] field gave promise of a better informed, more equitable, adjustment of the conflicting interests.” Gray, 314 U. S., at 412. Accordingly, the Court concluded that it was “not the province of a court” to “substitute its judgment” for the agency’s. Ibid. Three years later, the Court decided NLRB v. Hearst Publications, Inc., 322 U.S. 111 (1944), another acknowledged “leading case.” Davis 882; see id., at 884. The Court again deferred, this time to an agency’s construction of the term “employee” in the National Labor Relations Act. The scope of that term, the Court explained, “belong[ed] to” the agency to answer based on its “[e]veryday experience in the administration of the statute.” Hearst, 322 U. S., at 130. The Court therefore “limited” its review to whether the agency’s reading had “warrant in the record and a reasonable basis in law.” Id., at 131.[7] Recall here that even the majority accepts that Section 706 was meant to “restate[ ] the present law” as to judicial review. See ante, at 15–16; supra, at 19–20. Well then? It sure would seem that the provision allows a deference regime.

The majority has no way around those two noteworthy decisions. It first appears to distinguish between “pure legal question[s]” and the so-called mixed questions in Gray and Hearst, involving the application of a legal standard to a set of facts. Ante, at 11. If in drawing that distinction, the majority intends to confine its holding to the pure type of legal issue—thus enabling courts to defer when law and facts are entwined—I’d be glad. But I suspect the majority has no such intent, because that approach would preserve Chevron in a substantial part of its current domain. Cf. Wilkinson v. Garland, 601 U.S. 209, 230 (2024) (Alito, J., dissenting) (noting, in the immigration context, that the universe of mixed questions swamps that of pure legal ones). It is frequently in the consideration of mixed questions that the scope of statutory terms is established and their meaning defined. See H. Monaghan, Marbury and the Administrative State, 83 Colum. L. Rev. 1, 29 (1983) (“Administrative application of law is administrative formulation of law whenever it involves elaboration of the statutory norm”). How does a statutory interpreter decide, as in Hearst, what an “employee” is? In large part through cases asking whether the term covers people performing specific jobs, like (in that case) “newsboys.” 322 U. S., at 120. Or consider one of the examples I offered above. How does an interpreter decide when one population segment of a species is “distinct” from another? Often by considering that requirement with respect to particular species, like western gray squirrels. So the distinction the majority offers makes no real-world (or even theoretical) sense. If the Hearst Court was deferring to an agency on whether the term “employee” covered newsboys, it was deferring to the agency on the scope and meaning of the term “employee.”

The majority’s next rejoinder—that “the Court was far from consistent” in deferring—falls equally flat. Ante, at 12. I am perfectly ready to acknowledge that in the pre-APA period, a deference regime had not yet taken complete hold. I’ll go even further: Let’s assume that deference was then an on-again, off-again function (as the majority seems to suggest, see ante, at 11–12, and 13, n. 3). Even on that assumption, the majority’s main argument—that Section 706 prohibited deferential review—collapses. Once again, the majority agrees that Section 706 was not meant to change the then-prevailing law. See ante, at 15–16. And even if inconsistent, that law cannot possibly be thought to have prohibited deference. Or otherwise said: “If Section 706 did not change the law of judicial review (as we have long recognized), then it did not proscribe a deferential standard then known and in use.” Kisor, 588 U. S., at 583 (plurality opinion).

The majority’s whole argument for overturning Chevron relies on Section 706. But the text of Section 706 does not support that result. And neither does the contemporaneous practice, which that text was supposed to reflect. So today’s decision has no basis in the only law the majority deems relevant. It is grounded on air.

III

And still there is worse, because abandoning Chevron subverts every known principle of stare decisis. Of course, respecting precedent is not an “inexorable command.” Payne v. Tennessee, 501 U.S. 808, 828 (1991). But overthrowing it requires far more than the majority has offered up here. Chevron is entitled to stare decisis’s strongest form of protection. The majority thus needs an exceptionally strong reason to overturn the decision, above and beyond thinking it wrong. And it has nothing approaching such a justification, proposing only a bewildering theory about Chevron’s “unworkability.” Ante, at 32. Just five years ago, this Court in Kisor rejected a plea to overrule Auer v. Robbins, 519 U.S. 452 (1997), which requires judicial deference to agencies’ interpretations of their own regulations. See 588 U. S., at 586–589 (opinion of the Court). The case against overruling Chevron is at least as strong. In particular, the majority’s decision today will cause a massive shock to the legal system, “cast[ing] doubt on many settled constructions” of statutes and threatening the interests of many parties who have relied on them for years. 588 U. S., at 587 (opinion of the Court).

Adherence to precedent is “a foundation stone of the rule of law.” Michigan v. Bay Mills Indian Community, 572 U.S. 782, 798 (2014). Stare decisis “promotes the evenhanded, predictable, and consistent development of legal principles.” Payne, 501 U. S., at 827. It enables people to order their lives in reliance on judicial decisions. And it “contributes to the actual and perceived integrity of the judicial process,” by ensuring that those decisions are founded in the law, and not in the “personal preferences” of judges. Id., at 828; Dobbs, 597 U. S., at 388 (dissenting opinion). Perhaps above all else, stare decisis is a “doctrine of judicial modesty.” Id., at 363. In that, it shares something important with Chevron. Both tell judges that they do not know everything, and would do well to attend to the views of others. So today, the majority rejects what judicial humility counsels not just once but twice over.

And Chevron is entitled to a particularly strong form of stare decisis, for two separate reasons. First, it matters that “Congress remains free to alter what we have done.” Patterson v. McLean Credit Union, 491 U.S. 164, 173 (1989); see Kisor, 588 U. S., at 587 (opinion of the Court) (making the same point for Auer deference). In a constitutional case, the Court alone can correct an error. But that is not so here. “Our deference decisions are balls tossed into Congress’s court, for acceptance or not as that branch elects.” 588 U. S., at 587–588 (opinion of the Court). And for generations now, Congress has chosen acceptance. Throughout those years, Congress could have abolished Chevron across the board, most easily by amending the APA. Or it could have eliminated deferential review in discrete areas, by amending old laws or drafting new laws to include an anti-Chevron provision. Instead, Congress has “spurned multiple opportunities” to do a comprehensive rejection of Chevron, and has hardly ever done a targeted one. Kimble v. Marvel Entertainment, LLC, 576 U.S. 446, 456 (2015); see supra, at 14–15. Or to put the point more affirmatively, Congress has kept Chevron as is for 40 years. It maintained that position even as Members of this Court began to call Chevron into question. See ante, at 30. From all it appears, Congress has not agreed with the view of some Justices that they and other judges should have more power.

Second, Chevron is by now much more than a single decision. This Court alone, acting as Chevron allows, has upheld an agency’s reasonable interpretation of a statute at least 70 times. See Brief for United States in No. 22–1219, p. 27; App. to id., at 68a–72a (collecting cases). Lower courts have applied the Chevron framework on thousands upon thousands of occasions. See K. Barnett & C. Walker, Chevron and Stare Decisis, 31 Geo. Mason L. Rev. 475, 477, and n. 11 (2024) (noting that at last count, Chevron was cited in more than 18,000 federal-court decisions). The Kisor Court observed, when upholding Auer, that “[d]eference to reasonable agency interpretations of ambiguous rules pervades the whole corpus of administrative law.” 588 U. S., at 587 (opinion of the Court). So too does deference to reasonable agency interpretations of ambiguous statutes—except more so. Chevron is as embedded as embedded gets in the law.

The majority says differently, because this Court has ignored Chevron lately; all that is left of the decision is a “decaying husk with bold pretensions.” Ante, at 33. Tell that to the D. C. Circuit, the court that reviews a large share of agency interpretations, where Chevron remains alive and well. See, e.g., Lissack v. Commissioner, 68 F. 4th 1312, 1321–1322 (2023); Solar Energy Industries Assn. v. FERC, 59 F. 4th 1287, 1291–1294 (2023). But more to the point: The majority’s argument is a bootstrap. This Court has “avoided deferring under Chevron since 2016” (ante, at 32) because it has been preparing to overrule Chevron since around that time. That kind of self-help on the way to reversing precedent has become almost routine at this Court. Stop applying a decision where one should; “throw some gratuitous criticisms into a couple of opinions”; issue a few separate writings “question[ing the decision’s] premises” (ante, at 30); give the whole process a few years . . . and voila!—you have a justification for overruling the decision. Janus v. State, County, and Municipal Employees, 585 U.S. 878, 950 (2018) (Kagan, J., dissenting) (discussing the overruling of Abood v. Detroit Bd. of Ed., 431 U.S. 209 (1977)); see also, e.g., Kennedy v. Bremerton School Dist., 597 U.S. 507, 571–572 (2022) (Sotomayor, J., dissenting) (similar for Lemon v. Kurtzman, 403 U.S. 602 (1971)); Shelby County v. Holder, 570 U.S. 529, 587–588 (2013) (Ginsburg, J., dissenting) (similar for South Carolina v. Katzenbach, 383 U.S. 301 (1966)). I once remarked that this overruling-through-enfeeblement technique “mock[ed] stare decisis.” Janus, 585 U. S., at 950 (dissenting opinion). I have seen no reason to change my mind.

The majority does no better in its main justification for overruling Chevron—that the decision is “unworkable.” Ante, at 30. The majority’s first theory on that score is that there is no single “answer” about what “ambiguity” means: Some judges turn out to see more of it than others do, leading to “different results.” Ante, at 30–31. But even if so, the legal system has for many years, in many contexts, dealt perfectly well with that variation. Take contract law. It is hornbook stuff that when (but only when) a contract is ambiguous, a court interpreting it can consult extrinsic evidence. See CNH Industrial N.V. v. Reese, 583 U.S. 133, 139 (2018) (per curiam). And when all interpretive tools still leave ambiguity, the contract is construed against the drafter. See Lamps Plus, Inc. v. Varela, 587 U.S. 176, 186–187 (2019). So I guess the contract rules of the 50 States are unworkable now. Or look closer to home, to doctrines this Court regularly applies. In deciding whether a government has waived sovereign immunity, we construe “[a]ny ambiguities in the statutory language” in “favor of immunity.” FAA v. Cooper, 566 U.S. 284, 290 (2012). Similarly, the rule of lenity tells us to construe ambiguous statutes in favor of criminal defendants. See United States v. Castleman, 572 U.S. 157, 172–173 (2014). And the canon of constitutional avoidance instructs us to construe ambiguous laws to avoid difficult constitutional questions. See United States v. Oakland Cannabis Buyers’ Cooperative, 532 U.S. 483, 494 (2001). I could go on, but the point is made. There are ambiguity triggers all over the law. Somehow everyone seems to get by.

And Chevron is an especially puzzling decision to criticize on the ground of generating too much judicial divergence. There’s good empirical—meaning, non-impressionistic—evidence on exactly that subject. And it shows that, as compared with de novo review, use of the Chevron two-step framework fosters agreement among judges. See K. Barnett, C. Boyd, & C. Walker, Administrative Law’s Political Dynamics, 71 Vand. L. Rev. 1463, 1502 (2018) (Barnett). More particularly, Chevron has a “powerful constraining effect on partisanship in judicial decisionmaking.” Barnett 1463 (italics deleted); see Sunstein 1672 (“[A] predictable effect of overruling Chevron would be to ensure a far greater role for judicial policy preferences in statutory interpretation and far more common splits along ideological lines”). So if consistency among judges is the majority’s lodestar, then the Court should not overrule Chevron, but return to using it.

The majority’s second theory on workability is likewise a makeweight. Chevron, the majority complains, has some exceptions, which (so the majority says) are “difficult” and “complicate[d]” to apply. Ante, at 32. Recall that courts are not supposed to defer when the agency construing a statute (1) has not been charged with administering that law; (2) has not used deliberative procedures—i.e., notice-and-comment rulemaking or adjudication; or (3) is intervening in a “major question,” of great economic and political significance. See supra, at 11–12; ante, at 27–28. As I’ve explained, those exceptions—the majority also aptly calls them “refinements”—fit with Chevron’s rationale: They define circumstances in which Congress is unlikely to have wanted agency views to govern. Ante, at 27; see supra, at 11–12. And on the difficulty scale, they are nothing much. Has Congress put the agency in charge of administering the statute? In 99 of 100 cases, everyone will agree on the answer with scarcely a moment’s thought. Did the agency use notice-and-comment or an adjudication before rendering an interpretation? Once again, I could stretch my mind and think up a few edge cases, but for the most part, the answer is an easy yes or no. The major questions exception is, I acknowledge, different: There, many judges have indeed disputed its nature and scope. Compare, e.g., West Virginia, 597 U. S., at 721–724, with id., at 764–770 (Kagan, J., dissenting). But that disagreement concerns, on everyone’s view, a tiny subset of all agency interpretations. For the most part, the exceptions that so upset the majority require merely a rote, check-the-box inquiry. If that is the majority’s idea of a “dizzying breakdance,” ante, at 32, the majority needs to get out more.

And anyway, difficult as compared to what? The majority’s prescribed way of proceeding is no walk in the park. First, the majority makes clear that what is usually called Skidmore deference continues to apply. See ante, at 16–17. Under that decision, agency interpretations “constitute a body of experience and informed judgment” that may be “entitled to respect.” Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944). If the majority thinks that the same judges who argue today about where “ambiguity” resides (see ante, at 30) are not going to argue tomorrow about what “respect” requires, I fear it will be gravely disappointed. Second, the majority directs courts to comply with the varied ways in which Congress in fact “delegates discretionary authority” to agencies. Ante, at 17–18. For example, Congress may authorize an agency to “define[ ]” or “delimit[ ]” statutory terms or concepts, or to “fill up the details” of a statutory scheme. Ante, at 17, and n. 5. Or Congress may use, in describing an agency’s regulatory authority, inherently “flexib[le]” language like “appropriate” or “reasonable.” Ante, at 17, and n. 6. Attending to every such delegation, as the majority says, is necessary in a world without Chevron. But that task involves complexities of its own. Indeed, one reason Justice Scalia supported Chevron was that it replaced such a “statute-by-statute evaluation (which was assuredly a font of uncertainty and litigation) with an across-the-board presumption.” A. Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L. J. 511, 516. As a lover of the predictability that rules create, Justice Scalia thought the latter “unquestionably better.” Id., at 517.

On the other side of the balance, the most important stare decisis factor—call it the “jolt to the legal system” issue—weighs heavily against overruling Chevron. Dobbs, 597 U. S., at 357 (Roberts, C. J., concurring in judgment). Congress and agencies alike have relied on Chevron—have assumed its existence—in much of their work for the last 40 years. Statutes passed during that time reflect the expectation that Chevron would allocate interpretive authority between agencies and courts. Rules issued during the period likewise presuppose that statutory ambiguities were the agencies’ to (reasonably) resolve. Those agency interpretations may have benefited regulated entities; or they may have protected members of the broader public. Either way, private parties have ordered their affairs—their business and financial decisions, their health-care decisions, their educational decisions—around agency actions that are suddenly now subject to challenge. In Kisor, this Court refused to overrule Auer because doing so would “cast doubt on” many longstanding constructions of rules, and thereby upset settled expectations. 588 U. S., at 587 (opinion of the Court). Overruling Chevron, and thus raising new doubts about agency constructions of statutes, will be far more disruptive.

The majority tries to alleviate concerns about a piece of that problem: It states that judicial decisions that have upheld agency action as reasonable under Chevron should not be overruled on that account alone. See ante, at 34–35. That is all to the good: There are thousands of such decisions, many settled for decades. See supra, at 26. But first, reasonable reliance need not be predicated on a prior judicial decision. Some agency interpretations never challenged under Chevron now will be; expectations formed around those constructions thus could be upset, in a way the majority’s assurance does not touch. And anyway, how good is that assurance, really? The majority says that a decision’s “[m]ere reliance on Chevron” is not enough to counter the force of stare decisis; a challenger will need an additional “special justification.” Ante, at 34. The majority is sanguine; I am not so much. Courts motivated to overrule an old Chevron-based decision can always come up with something to label a “special justification.” Maybe a court will say “the quality of [the precedent’s] reasoning” was poor. Ante, at 29. Or maybe the court will discover something “unworkable” in the decision—like some exception that has to be applied. Ante, at 30. All a court need do is look to today’s opinion to see how it is done.

IV

Judges are not experts in the field, and are not part of either political branch of the Government.

  — Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 865 (1984)

Those were the days, when we knew what we are not. When we knew that as between courts and agencies, Congress would usually think agencies the better choice to resolve the ambiguities and fill the gaps in regulatory statutes. Because agencies are “experts in the field.” And because they are part of a political branch, with a claim to making interstitial policy. And because Congress has charged them, not us, with administering the statutes containing the open questions. At its core, Chevron is about respecting that allocation of responsibility—the conferral of primary authority over regulatory matters to agencies, not courts.

Today, the majority does not respect that judgment. It gives courts the power to make all manner of scientific and technical judgments. It gives courts the power to make all manner of policy calls, including about how to weigh competing goods and values. (See Chevron itself.) It puts courts at the apex of the administrative process as to every conceivable subject—because there are always gaps and ambiguities in regulatory statutes, and often of great import. What actions can be taken to address climate change or other environmental challenges? What will the Nation’s health-care system look like in the coming decades? Or the financial or transportation systems? What rules are going to constrain the development of A.I.? In every sphere of current or future federal regulation, expect courts from now on to play a commanding role. It is not a role Congress has given to them, in the APA or any other statute. It is a role this Court has now claimed for itself, as well as for other judges.

And that claim requires disrespecting, too, this Court’s precedent. There are no special reasons, of the kind usually invoked for overturning precedent, to eliminate Chevron deference. And given Chevron’s pervasiveness, the decision to do so is likely to produce large-scale disruption. All that backs today’s decision is the majority’s belief that Chevron was wrong—that it gave agencies too much power and courts not enough. But shifting views about the worth of regulatory actors and their work do not justify overhauling a cornerstone of administrative law. In that sense too, today’s majority has lost sight of its proper role.

And it is impossible to pretend that today’s decision is a one-off, in either its treatment of agencies or its treatment of precedent. As to the first, this very Term presents yet another example of the Court’s resolve to roll back agency authority, despite congressional direction to the contrary. See SEC v. Jarkesy, 603 U. S. ___ (2024); see also supra, at 3. As to the second, just my own defenses of stare decisis—my own dissents to this Court’s reversals of settled law—by now fill a small volume. See Dobbs, 597 U. S., at 363–364 (joint opinion of Breyer, Sotomayor, and Kagan, JJ.); Edwards v. Vannoy, 593 U.S. 255, 296–297 (2021); Knick v. Township of Scott, 588 U.S. 180, 207–208 (2019); Janus, 585 U. S., at 931–932. Once again, with respect, I dissent.