Tossing Sand In The Regulatory Gears: Hurdles To Policy Progress In The Supreme Court

Nina A. Mendelson [*] View PDF

In the last few years, the Supreme Court has been a source of seismic change. In Dobbs v. Jackson Women’s Health Organization,¹ the Court overruled Roe v. Wade,² which had protected the right to abortion for nearly fifty years. In Loper Bright Enterprises v. Raimondo, the Court abandoned so-called Chevron deference to particular categories of administrative agency interpretations, a doctrine viewed as bedrock for over forty years.³ Humphrey’s Executor v. United States,⁴ the 1935 ruling validating independent multi-member commissions such as the Federal Trade Commission, Securities and Exchange Commission, and Federal Communications Commission, may soon join the others on the chopping block.⁵

But the Court has been far more skeptical when other governmental institutions seek change to address modern problems, whether it is Congress enacting regulatory legislation, or executive branch agencies seeking to exercise statutory authority. Governmental policy change can, of course, be critical to addressing newly significant problems, including climate change, environmental and health risks from widely used per-fluoroalkyl and polyfluoroalkyl substances (PFAS) found in industry and consumer products and other new pollutants, and innovative means of financial fraud. Policy change may also be needed to incorporate new technologies or other new solutions to societal problems or address new implementation issues that have emerged over time.

Particularly in the last five years, the Court has hobbled numerous executive branch agency efforts to address modern challenges with a new interpretive move that, in 2022, it named the major questions doctrine (“MQD”).⁶ While the MQD has drawn controversy and attention, this essay argues that the MQD’s creation is not an isolated judicial move. It is simply the most visible of a suite of alterations to administrative law and statutory interpretation doctrine, of which the Court’s Loper Bright ruling is the most recent example.

As discussed in greater detail below, between 2019 and 2024, the Court not only developed the MQD, but altered the landscape of judicial deference to agency interpretations. The Court abandoned an attitude that had tolerated, and even lauded, agency policy change in many interpretive settings, replacing it with across-the-board antagonism to change. This included modifying so-called Skidmore deference by significantly strengthening its existing anti-change aspects, making it what we might call Skidmore 2.0.⁷ The Court also has strengthened anti-change aspects of arbitrary and capricious review by requiring agencies to consider the unwieldy factor of generalized reliance.⁸ Finally, the Court is developing additional statutory interpretation rules that disfavor regulatory legislation and limit its application, including a potential “private property” canon.⁹

All of these alterations have made the courts an increasingly powerful barrier to both legislative and executive policy change, ultimately impeding governmental institutions from addressing modern societal challenges in a democratically responsive fashion.

I. The Major Questions Doctrine: The Court’s Most Visible Anti-Change Move

Under the MQD, first named by a majority of the Court in the 2022 decision West Virginia v. EPA, the Court held that an agency’s assertion of economically or politically significant authority would not be within the scope of an otherwise broad statutory grant of power unless the statute provides “clear congressional authorization.”¹⁰ The majority conceded that in major questions cases, the statutory language has often supplied a “colorable textual basis” for the agency’s interpretation.¹¹ But, the Court explained that extraordinary circumstances would prompt it to take a different approach, and instead to require a clear statutory statement.¹² As the cases discussed below show, the statutory authorizing language must not only be clear, but specific.¹³ Broad language seemingly will not do.

Under the MQD, an agency’s claim of significant authority to take action that can be characterized as new or different in kind is particularly vulnerable to challenge. In West Virginia v. EPA, the agency had sought to regulate coal-fired power plant greenhouse gas emissions through a “standard of performance” for existing coal-fired power plants known as the Clean Power Plan.¹⁴ Under the statutory provisions the Environmental Protection Agency (EPA) relied upon, it was then up to the states to issue the particular rules achieving those emissions reductions.¹⁵ “Standard of performance” was defined as a “standard for emissions” reflecting the application of the “best system of emission reduction” that has been “adequately demonstrated.”¹⁶ As the Court explained, EPA’s standard was based on a number of potential measures to reduce emissions, including power plant operators burning coal more cleanly, shifting to cleaner fuels or renewables (so-called generation shifting), or participating in an emissions trading regime with other operators who had succeeded in lowering greenhouse gases, including by generation shifting.¹⁷ The Court conceded that these measures, including generation shifting, could be encompassed within the statute’s language as a “system” capable of reducing emissions.¹⁸ Other sections of the Clean Air Act that more narrowly authorized the agency to set “standards applicable to . . . emission[s],” without referencing a “system,” seemed to confirm the breadth of this particular authority.¹⁹ It is worth emphasizing that the policy’s reliance on the market mechanism of emissions trading was an “unqualified bull’s-eye” in terms of economic efficiency, since it “create[d] economic incentives to have . . . pollution reductions undertaken by those . . . that can do so least expensively,” the kind of regulatory approach also typically “trumpeted by Republican administrations.”²⁰ One might fairly infer that the statutory term “system” had been chosen by Congress not only to enable the agency to respond to technological evolution, but also to evolving thinking on maximizing regulatory effectiveness.

The Court hesitated to read the statute in this way, however, characterizing the Clean Power Plan as the agency asserting a “newfound power” in a “rarely . . . used” provision, because the agency had previously tended to set or call for source-based emissions limits.²¹ Calling the action “novel” was a critical step in the Court’s invocation of the major questions doctrine.²² Using that doctrine, the Court then rejected the broad term “system of emissions reduction” as inadequately specific to authorize the Clean Power Plan.²³

Similarly, in a case decided the previous year on the so-called shadow docket, Alabama Assn. of Realtors v. Department of Health and Human Services,²⁴ the Court alluded to the “unprecedented” nature of the Centers for Disease Control’s assertion of statutory authority to institute a nationwide eviction moratorium in response to the COVID-19 pandemic. Again, the statute’s language broadly authorized the agency to adopt measures “necessary to prevent” the spread of disease, but the Court instead focused on the purported newness of the action. Despite evidence of disease transmission among the unhoused, the Court characterized the action as housing-focused, rather than aimed at combatting the spread of disease.²⁵ Again, the newness of the agency action was a key step in the Court’s application of the MQD. The agency ultimately lost its claim that action was authorized for lack of specific statutory language.

In National Federation of Independent Business v. Occupational Safety and Health Administration, the Court held that the power of the Occupational Safety and Health Administration (OSHA) over “occupational safety and health standards” did not permit it to require large employers to institute a “vaccine or test” rule during the early years of the COVID-19 pandemic, an action the Court also found to be unprecedented.²⁶ Although the statute made reference to “occupational . . . health” as well as “safety,” language facially clear enough to authorize a rule aimed at reducing workplace disease transmission, the Court nonetheless invalidated the rule because the statute’s authorization was not sufficiently specific.²⁷ Similarly, in Biden v. Nebraska, the Court characterized the agency’s claimed use of its power to “waive or modify” student loans to waive repayment obligations as creating a “novel and fundamentally different loan forgiveness program.”²⁸ It then applied the MQD to strike down the action, finding the statutory language inadequately specific to meet the “clear congressional authorization” requirement.²⁹

The major questions doctrine is deeply problematic because it lacks objective criteria for a “major” question or “significant” economic or political impact, thus conferring substantial discretion on courts and making its application unpredictable.³⁰ It seems hard to justify as a substantive interpretive canon that would effectuate the nondelegation doctrine, since that doctrine does not preclude large delegations to agencies, only unprincipled ones.³¹ The MQD also cannot be justified as an interpretive rule that provides Congress with a stable background for legislation, since, even if the MQD could be considered predictable, a dubious proposition, the MQD has been applied to statutes enacted long before the MQD was created, when Congress could not have anticipated it.³² Finally, the MQD also seems difficult to justify as a “common sense” assumption about congressional use of language, a sort of linguistic canon presented by Justice Amy Coney Barrett in her concurrence in Biden v. Nebraska.³³ Survey-based research suggests that ordinary people’s understanding of delegations do not typically include significance limitations.³⁴ Congressional drafters and agency counsel also likely understand broad language in a statute to delegate “wide authority” to agency administrators.³⁵

That said, the MQD is here to stay.³⁶ Its inclusion of newness as a factor that could make an issue of agency authority a “major question” is obviously antagonistic to policy change and innovation. It constrains agencies from adapting policy in response to new problems, new advances in knowledge or technology, and unforeseen implementation problems.

The MQD also can be understood as playing “gotcha” with Congress—even when Congress could not have anticipated an MQD-type requirement—by requiring it to have anticipated and spoken to the agency’s specific policy solution to a modern problem. In the cases discussed above, the challenged policies seemed plainly within both the underlying statute’s broad language as well as its purpose, as with the Occupational Safety and Health Act’s instruction to protect workplace “health” in the face of a global pandemic. (The Act was, of course, enacted in 1970, long before the creation of the MQD, so even if members of Congress anticipated its application in a global pandemic, they might have felt no call to include specific language). Again, such broad statutory language seems aimed at empowering implementing agencies to respond to specific issues Congress understands may arise later.³⁷ But in the MQD era, if the Court considers the issue to be “major,” and Congress has not specifically anticipated that particular issue in statutory language, one statute is not good enough for the Court—Congress must legislate a second time. Courts effectively regulating the legislative process in this way seems problematic at any time, but it is particularly cynical in an era of legislative dysfunction.

The MQD is not only anti-change; it is distinctly anti-regulatory. First, as Professors Deacon and Litman have argued, the Court’s focus on “new” agency actions essentially requires agencies to “use it [early on] or lose it.”³⁸ Second, creating regulation is particularly vulnerable as “new.” Deregulation—even dramatic deregulation—seems less susceptible to being characterized as “new” because in the case of deregulation, the agency has already plowed that subject-matter ground by regulating in the first place. Finally, the MQD burdens regulation through its legislative process impacts. Losers in a major questions doctrine case must return to Congress to obtain specific authorizing language. As if requiring Congress to legislate a second time weren’t enough, MQD-required specific regulatory language will likely be especially difficult to enact since a specific regulatory proposal will be strongly opposed by regulated industry. As public choice theory predicts, regulated industry is likely to be better organized and better-funded than diffuse public beneficiaries, and thus more effective at defeating even legislation that significantly benefits the public.³⁹ Broad, public-facing legislation has long been understood as vulnerable to this sort of behind-the-scenes resource imbalance; the MQD’s requirement of specific language worsens it by further tilting the scales against regulatory legislation.

The MQD is mainly focused on new agency actions that are economically or politically “significant,” which might seem to limit its scope. But judicial opposition to change is not limited to the MQD. The Court has embedded antipathy to “new” agency actions—even those that might not be deemed so “significant”—through several other doctrinal moves.

II. Earlier, Subtler Moves Towards an Anti-Change Regime: Requirements that Agencies Consider Reliance in Arbitrary and Capricious Review

Under the foundational statute governing the federal administrative state, the Administrative Procedure Act, courts are to set aside agency action if they judge it to be “arbitrary, capricious, an abuse of discretion, or otherwise contrary to law.”⁴⁰ While reviewing courts have long demanded reasoned and rational decisions from agencies, the Court also had long indicated its comfort with, and even expectation of, administrative policy flexibility and change.⁴¹

There were, of course, certain limits. In the enforcement setting, fair warning requirements rooted in due process precluded an agency from imposing liability on an entity based on newly announced regulatory requirements.⁴² The Court also previously suggested that agencies should not be able to act through adjudication (as opposed to rulemaking) if the result might create new liability for regulated entity actions taken in good-faith reliance on previous agency regulations, a similarly targeted notion of reliance.⁴³ Similarly, the Court indicated that Congress might itself create a safe harbor for regulated entities who might have relied on earlier versions of regulations.⁴⁴ One might also infer that the presumption against retroactive applications of statutes incorporates reliance concerns, though such a presumption generally can be readily rebutted by language or some other indication that Congress meant for a statute to apply retroactively.⁴⁵ But crucially, none of this precluded agencies from changing policy or instituting new rules that would apply prospectively and broadly.

For example, the iconic 1984 arbitrary and capricious review case, Motor Vehicles Manufacturers Association v. State Farm,⁴⁶ involved regulatory policy change. In 1981, the Reagan Administration repealed a 1977 Carter-era rule that would have required all automotive manufacturers to install passive restraints (automatic seatbelts or airbags) on new cars by automotive model year 1984. The Court established that rule repeals were subject to arbitrary and capricious review,⁴⁷ even though agency inaction might not require justification⁴⁸ or—though the Court did not specifically state this—might not even be subject to judicial review.⁴⁹ The Court noted that an agency’s earlier position should be seen as an “informed judgment” on the best course of implementation, justifying arbitrary and capricious review of a rescission of that judgment.⁵⁰ The State Farm Court’s analysis identified the still-leading factors that compose arbitrary and capricious review, including whether the agency has: 1) considered relevant alternatives, 2) considered relevant factors under a statute (and not irrelevant ones), and 3) articulated a rational connection between the facts it found and the choices it made.⁵¹

The agency’s earlier rule requiring passive restraints did, to some extent, shape the Court’s review of the rescission in State Farm. For example, the agency’s requirement of airbags and automatic belts as the 1977 compliance options led the Court to unanimously set aside the 1981 agency action for failure to consider an airbags-only alternative.⁵² Yet, the Court emphasized that an agency must be given “ample latitude” to adapt its rules and policies to changing circumstances,⁵³ and notably, State Farm did not mention either manufacturer or consumer reliance on the earlier rule as a consideration for the agency. To the contrary, Justice William Rehnquist, in dissent, famously suggested that the Court should have been even more generous with the agency’s rescission decision, stating: “A change in administration . . . is a perfectly reasonable basis for an executive agency’s reappraisal of the costs and benefits of its programs and regulations [including public resistance and uncertainty], as long as the agency remains within the bounds established by Congress.”⁵⁴

Beginning in the 1990s, however, the Court began to embed more anti-change seeds in its arbitrary and capricious rulings. It did so by requiring agencies changing policy to expressly consider reliance.⁵⁵ The Court first flagged reliance as a general concern for agency policy changes in its 1996 decision, Smiley v. Citibank (South Dakota), N.A.⁵⁶ In a unanimous opinion authored by Justice Antonin Scalia, the Court upheld an agency interpretation under Chevron deference and rejected challenges based on the agency’s interpretive change since it found no meaningful change in official position.⁵⁷ The Court reaffirmed that such interpretive change could be ordinary, but—in dicta—identified reliance as an additional factor for an agency to consider. The Court commented, “[C]hange is not invalidating, since the whole point of Chevron is to leave the discretion provided by the ambiguities of a statute with the implementing agency,”⁵⁸ but cautioned that a failure to “take account of legitimate reliance on prior interpretation” could render a decision arbitrary or capricious.⁵⁹

In 2008, in FCC v. Fox Television Stations, Inc.,⁶⁰ the Court, in a majority opinion by Justice Scalia, rejected an arbitrary and capricious challenge⁶¹ to the FCC’s decision to abandon a safe broadcasting harbor for fleeting expletives and to treat them instead as indecent. Although the Court’s majority stated that an agency has flexibility to change policy, it laid out a more detailed explanatory requirement for the agency, and again affirmed that an agency cannot change policy without considering reliance. The Fox Court stated that an agency changing its policy need not demonstrate that a new policy was “better” than the old, but it must display “awareness” that it was changing position and explain (if necessary) why it might disregard “facts and circumstances” associated with the prior policy.⁶² The Court also stated that the agency’s consideration must include “serious reliance interests” potentially engendered by the earlier policy.⁶³

In contrast to other requirements for agencies, the reliance requirement has been developed as free-floating judicial common law not specifically rooted in other legal sources or doctrines. The Court has not mentioned fair warning or fair notice concerns, for example. With respect to other relevant factors an agency must consider, the Court generally has felt compelled to anchor its analysis in the agency’s underlying authorizing statute. For example, in Michigan v. EPA,⁶⁴ the Court set aside an EPA decision as arbitrary and capricious because the agency failed to consider cost in an initial decision that mercury emissions from power plants required regulation.⁶⁵ But it did so only after exhaustively explaining that cost should be understood as an implicitly required factor by the statute’s terms “appropriate and necessary.”⁶⁶ Although the Court has engaged in no such analysis for reliance, courts nonetheless have set aside new agency policy decisions for failure to consider reliance.⁶⁷

The Court has also been unclear about just which sorts of reliance interests an agency must consider, creating additional opportunities for the courts to nitpick agency policy changes. The cases discussed above implicate economic concerns, and the Court repeatedly mentioned the “significant [economic] reliance interest involved” in setting aside, as inadequately reasoned, an agency’s Fair Labor Standards Act regulation interpreting “salesman” to cover service advisors at auto dealerships in Encino Motorcars v. Navarro in 2016.⁶⁸ But in 2020, in Department of Homeland Security v. Regents of the University of California,⁶⁹ the Court set aside as arbitrary and capricious the Department of Homeland Security’s recission of the Deferred Action for Childhood Arrivals, or “Dreamers” program, for failure to consider reliance interests in detail.⁷⁰ The Court expanded the list of plausible reliance interests beyond purely economic interests—though without ranking them, evaluating them, or anchoring them in the underlying statute—to include personal decisions to take jobs, marry, purchase houses, or seek education, as well as money invested in employer training and governmental tax revenue.⁷¹ Such a list may be appealing as broadening legitimate reliance concerns beyond the purely economic. Irrespective of its potential appeal, however, courts have developed this requirement in an ad hoc, unpredictable fashion.

Thus, in addition to offering an extensive explanation of why it sees the world differently, an agency seeking to change policy must address a difficult-to-anticipate set of reliance interests. These requirements represent a subtle new set of judicial roadblocks to agency policy changes.

III. Change Itself as a Reason to Refuse Deference to Agency Interpretations Across the Board

Prior to 2019, two of the three general doctrines governing judicial review of agency interpretations of law incorporated considerable scope for agency flexibility and policy change. The doctrine of Auer⁷² or Seminole Rock⁷³Bowles v. Seminole Rock & Sand Co., 325 U.S. 410 (1945). deference to an agency’s interpretation of its own regulation left space for an agency to change its position. The Chevron doctrine, requiring presumptive deference to certain categories of agency interpretations of statutes, also supported flexible interpretation. Only the framework of the 1944 decision Skidmore v. Swift ⁷⁴ disfavored change by suggesting that other agency statutory interpretations would be more worthy of deference if they were consistently held over time.⁷⁵ But in 2019, in Kisor v. Wilkie,⁷⁶ the Court significantly modified Auer, and of course in the 2024 decision of Loper Bright Enterprises v. Raimondo,⁷⁷ the Court overruled Chevron and reinstated the judiciary as the primary decider of legal questions about an agency’s authority.⁷⁸ Loper Bright also modified Skidmore deference to strengthen its anti-change tilt.⁷⁹ Thus, change in agency interpretations is now judicially disfavored across the board.

A. Skidmore and Chevron Prior to Loper Bright

Before Chevron, the Court had long applied the so-called Skidmore framework to decide whether deference to an agency interpretation might be warranted.⁸⁰ In Skidmore, the Court, in addressing a private employment dispute over pay, followed an agency interpretation of a statute announced in an agency “bulletin.”⁸¹ The open-ended Skidmore framework largely left the deference question to a judge’s discretion, though it expressed the notion that some agency interpretations might be especially deserving, and thus a worthy source of guidance to which a court might resort (or not).⁸² The Skidmore framework afforded “respect” to the agency’s interpretation of the law based on 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.”⁸³ The preference for consistency, of course, disfavored change. Though the Skidmore Court did not explain why it chose consistency as a criterion, conceivably a consistently-held interpretation might have been judged relatively well-considered at the start—or perhaps frequent interpretive changes might be deemed disruptive.

By contrast, under the stronger Chevron deference framework (in which courts deferred to reasonable agency interpretations of ambiguous authorizing statutes), change was no bar to deference. In Chevron,⁸⁴ the Court expressly rejected the notion that change would delegitimize an agency’s legal interpretation, deferring to the agency’s interpretation of a Clean Air Act term even though the agency had changed its interpretation over time. The Court explained: “An initial agency interpretation is not instantly carved in stone. On the contrary, the agency, to engage in informed rulemaking, must consider varying interpretations and the wisdom of its policy on a continuing basis.”⁸⁵ The Court reaffirmed this position—and the agency’s legitimate need to reconsider a policy’s wisdom via reinterpretation of its authority—in National Cable & Telecommunications Ass’n v. Brand X, where it stated that change in an agency interpretation, even if contrary to a Court of Appeals ruling, would not render the new agency interpretation ineligible for Chevron deference.⁸⁶ The Brand X Court reaffirmed the reasons given in Justice Rehnquist’s State Farm dissent, commenting that, at most, unexplained inconsistency might prompt a court to find a changed interpretation arbitrary and capricious.⁸⁷

B. Chevron’s Abandonment in Loper Bright and Skidmore 2.0’s Deepened Opposition to Change

By the time Loper Bright was argued in 2024, members of the Court had expressed significant concern regarding change in an agency’s interpretation of its own authorizing statute. In a dissent to a 2022 certiorari denial in Buffington v. McDonough,⁸⁸ Justice Neil Gorsuch argued that Chevron invited executive officials to interpret the law aggressively—and then, during a change in presidential administration, emboldened new officials to “proceed in the opposite direction with equal zeal.”⁸⁹ Gorsuch suggested that Chevron facilitated so much change that “individuals can never be sure of their legal rights and duties,” potentially leaving them “caught in the whipsaw” of rule changes.⁹⁰ Similarly, at the Loper Bright oral argument, Justice Brett Kavanaugh expressed concern for the prospect that Chevron deference might still apply when the “agency changes position every four years,”⁹¹ a sharp contrast to Justice Rehnquist’s endorsement of policy change initiated by new presidential administrations in State Farm.⁹²

In Loper Bright, the Court overruled Chevron and returned to Skidmore deference for all agency statutory interpretations—but in a way that significantly strengthened Skidmore’s anti-change tilt. Skidmore already incorporated a preference for consistent agency interpretations of statutes so that the announcement of a changed position would apparently be less deserving of deference. The Loper Bright Court endorsed this principle in what we might call Skidmore 2.0.⁹³

Loper Bright strengthened Skidmore’s obstacles to change in two ways. First, the Court suggested that respect should be denied not only to changed agency interpretations, but to new ones as well. Citing pre-New Deal cases, the Court characterized its historical approach as respectful to agency views when an agency interpretation was not just consistent but was issued “roughly contemporaneously” with a statute’s enactment.⁹⁴ It emphasized that in Skidmore 2.0, courts should find especially useful “interpretations issued contemporaneously . . . and which have remained consistent over time.”⁹⁵

The Court did not explain its rationale for this additional limit on interpretations. Conceivably, privileging interpretations contemporaneous with enactment might recognize that an agency official had been involved in drafting or negotiating the statute, and thus might have particular knowledge of legislative context or congressional goals. Or, perhaps it reflects the possibility that Congress has not had an opportunity to respond to later interpretations, and thus cannot be assumed to have acquiesced. But whatever the justification, the Court’s decision means less respect for a later-in-time agency interpretation. Such interpretations can, of course, be important to addressing new problems, new solutions, and new implementation issues that have emerged as a statute has been interpreted. For example, after the Court ruled in 2007 in Massachusetts v. EPA that the core 1970 Clean Air Act term “air pollutant” clearly includes greenhouse gases,⁹⁶ the EPA confronted interpretive difficulties from the fact that greenhouse gases are often emitted in more “vast quantities” than other air pollutants regulated under the Clean Air Act.⁹⁷ Any statute may present a host of issues that Congress could not or did not specifically address.

Second, the Loper Bright Court set forth a statutory interpretation approach, at least in the setting of agency actions, that seems to leave little room for change over time, even when a statute is phrased broadly. The Court stated, “[E]very statute’s meaning is fixed at the time of enactment,” expressly rejecting the notion that statutes could have gaps or ambiguities, which in past opinions has warranted deference to agency interpretations.⁹⁸ Over a dissent by Justice Elena Kagan, the majority denied that “Congress’s instructions” could “run out” at a particular point in time⁹⁹ and insisted that every statute must have a “single, best meaning” that simply awaits judicial detection.¹⁰⁰ This focus on the time of enactment paves the way for future cases to adopt a particularly time-bound approach to statutes, whether it is by limiting application of a statute to circumstances mentioned in dictionaries at the time, or to those specifically contemplated by members of the enacting Congress. Consider, for example, the use of the term “drug” in the Federal Food, Drug, and Cosmetic Act of 1938, defined to include “articles intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease . . . [and those] intended to affect the structure or any function of the body . . . .”¹⁰¹ Antihistamines for human use were not distributed widely until the 1940s; immunotherapy treatments for cancer were developed far more recently.¹⁰² The statute, aimed at protecting the public from unsafe and ineffective drugs, was clearly meant to encompass such substances even if Congress did not, or could not, specifically anticipate them in 1938. The Court’s approach is in tension with Congress’s deliberate selection of capacious words in regulatory statutes.

In contrast, in other settings with no agency interpretation on offer, the Court has acknowledged that a statute could be properly interpreted to encompass situations not specifically anticipated at the time of enactment. For example, in the 2018 opinion quoted in Loper Bright for the proposition that a statute’s meaning is “fixed” at enactment, Wisconsin Central Ltd. v. United States,¹⁰³ the Court held that the term “money remuneration” did not include stock options for purposes of the Railroad Retirement Tax Act of 1937.¹⁰⁴ But it also acknowledged in dicta that “electronic transfers of paychecks,” although unknown in 1937, would still qualify as “money remuneration.”¹⁰⁵ Justice Gorsuch wrote for the majority that although “every statute’s meaning is fixed at the time of enactment, new applications may arise in light of changes in the world.”¹⁰⁶ Whether or not the concept of statutory “meaning” can be convincingly distinguished from “applications,”¹⁰⁷ the Wisconsin Central opinion clearly contemplated that a general term can be appropriately interpreted over time to apply to circumstances the drafters did not specifically anticipate. Similarly, in the 2018 decision Bostock v. Clayton County,¹⁰⁸ a majority of the Court held that the term “sex” in the Civil Rights Act of 1964’s employment discrimination provisions was properly interpreted to encompass sexual orientation and transgender status, since such discrimination is necessarily “based on sex,”¹⁰⁹ although the majority acknowledged that these particular outcomes “might not have [been] anticipated” at the time.¹¹⁰

Perhaps the Loper Bright Court simply forgot to mention that broad statutory terms implemented by agencies could properly apply to circumstances not specifically foreseen at the time of enactment. But it is also conceivable that the Court is laying the groundwork to extend some aspects of the MQD to a much larger class of cases, not just those raising “major” issues. Where a statute authorizes an agency to act, the Court could seek to limit a statute’s scope of authorization just to circumstances specifically addressed in the text or specifically intended at the time of enactment. While the issue is beyond the scope of this essay, this sort of approach is in tension with textualism’s commitments, since it may require looking beyond generally phrased statutory text to specific legislative intent, legislative history, or both.¹¹¹ And the Court might apply such an approach even to statutes enacted long before Loper Bright was decided, irrespective of whether Congress could have anticipated this interpretive move. As with the MQD, such an interpretive approach could force Congress to legislate again to achieve a desired and appropriate policy response to new problems.

C. Restricting Auer Deference for Changed Agency Interpretations

The Court’s shift against change also has extended to its treatment of an agency interpretation of its own regulation. In the 1997 decision Auer v. Robbins, the Court reaffirmed a longstanding doctrine announced in a 1945 case, Bowles v. Seminole Rock, that an agency’s interpretation of its own rules should be accepted as “controlling unless ‘plainly erroneous or inconsistent with the regulation.’”¹¹² In Auer and in a 2007 decision, Long Island Care at Home Ltd. v. Coke, the Court held that Auer deference was available for interpretations both in internal memoranda and amicus briefs as long as they created no “unfair surprise” and were not post hoc rationalizations.¹¹³ In Long Island Care, the Court specifically held that a change in an agency’s interpretation would be no reason to deny deference,¹¹⁴ implying that generalized reliance on previous positions would not bar an agency from adopting a new interpretation. The Court did not amplify the meaning of the sort of “unfair surprise” that would be a reason to deny deference, but did suggest that the use of notice and comment rulemaking to change a policy would be sufficient to eliminate any unfairness.¹¹⁵ Again, the Court’s concern seemed limited to the case in which a party faced liability based on a new agency interpretation, something akin to a lack of fair warning required by due process.¹¹⁶

But in Kisor v. Wilkie,¹¹⁷ the Court more clearly ruled a far larger class of changed interpretations unworthy of deference. Kisor reaffirmed that an agency interpretation of its rule would not warrant Auer deference if it created “unfair surprise,” but equated that surprise to the “disruption of expectations [that] may occur when an agency substitutes one view of a rule for another.”¹¹⁸ The Court added that such disqualifying “upending of reliance” might happen—even if an agency had not previously plowed the interpretive ground—if the agency, through its interpretation, sought to address “longstanding conduct that the agency had never before addressed.”¹¹⁹ If the Court follows through and applies this language as written, it would seem to cover most newly announced agency interpretations of rules, making them ineligible for deference.

Moreover, if this expansive notion of “upending reliance” is extended to other areas in which the Court has decided reliance is relevant, such as in arbitrary and capricious review or review of changed agency interpretations of statutes, it would further impede agency efforts to respond to new problems, technologies, or democratic policy preferences.

IV. Anti-Change Statutory Interpretation Rules that Impede Effective Regulatory Legislation

Finally, the Court seems to be developing interpretive approaches that would limit the scope and effectiveness of regulatory legislation. A handful of anti-change substantive interpretive canons are already in use, most notably the so-called common law canon, stating that “statutes in derogation of the common law are to be strictly construed.”¹²⁰ The canon includes the notion that enacted words with a settled common law meaning take that “common law soil” along with them.¹²¹ But statutes, of course, are often meant to respond to inadequacies of common law.¹²² The common-law canon can thus blunt a statute’s impact. For example, in United States v. Bestfoods,¹²³ the Court narrowly interpreted the Comprehensive Environment, Response, Compensation, and Liability Act (CERCLA) in light of “general principle[s]” of corporate law, which it characterized as common law.¹²⁴ CERCLA imposes strict, joint and several, and retroactive liability for hazardous waste cleanup costs on those responsible for contamination, including facility “operators.”¹²⁵ It was passed in response to environmental disasters, especially abandoned hazardous waste sites, that were not being adequately addressed by common law remedies.¹²⁶ Guided by what it characterized as common-law restrictions on “piercing the corporate veil” to impose legal responsibility on corporate shareholders, the Court nonetheless interpreted the statute to almost never impose liability on a corporate shareholder or parent company, even one with complete ownership of a polluting company, and thus the ability to control its operation and environmental impacts.¹²⁷ Such an interpretive approach can be understood as undercutting a statute’s effectiveness, impeding Congress from making meaningful change.

Two of the doctrinal changes discussed above can be understood as adding to the Court’s arsenal of anti-change interpretive rules. Besides serving as a roadblock to agency assertions of authority, the MQD hinders Congress from using clear, but broad, language to create regulatory programs that can effectively address problems arising over time. Meanwhile, the Loper Bright Court’s dicta that a statute’s meaning is “fixed at the time of enactment” could hamper the application of many broadly-phrased statutes to later-arising issues, even when no “major questions” are implicated.

Finally, the Court more recently alluded to a nascent “private property” interpretive rule in the 2023 decision of Sackett v. Environmental Protection Agency.¹²⁸ In that case, a majority of the Court narrowly interpreted the scope of the central federal water pollution statute, the Clean Water Act. The Court held that the Clean Water Act’s jurisdictional “waters of the United States” language did not cover wetlands other than those with a continuous surface connection to traditionally navigable waters and their tributaries.¹²⁹ It further stated that Congress would have to “enact exceedingly clear language if it wishes to significantly alter the balance between federal and state power and the power of the Government over private property.”¹³⁰ While the Court’s reference to the balance of federal and state power seemed to invoke its late-twentieth-century federalism canons,¹³¹ an extra-clear statement rule to protect private property would be wholly new.¹³² Virtually all federal regulation—economic, environmental, safety, or health—affects private property in the sense that it aims to improve some parties’ well-being or wealth, and changes the way others spend their resources or use their assets. Thus, such an interpretive rule, if entrenched, could limit the application of a wide array of federal regulatory statutes, old and new, unless Congress is able to see into a crystal ball and address individual issues in a highly specific fashion. Ironically, given this seeding of a new anti-regulatory change canon, in adopting a narrow application of the Clean Water Act, the Court fractured longstanding practice by overruling the position of all eight post-enactment presidential administrations on the Clean Water Act’s coverage.¹³³

V. Conclusion

By requiring not just clear, but specific, statutory authorization language, the Court’s major questions doctrine has unquestionably made it harder both for Congress and agencies to respond to significant modern problems. But in a far less visible way, numerous other doctrinal changes embody the Court’s efforts to restrict the scope of modern government. The Court is instituting large and small doctrinal obstacles to legislative regimes that were designed to respond to entire categories of modern challenges such as environmental pollution, occupational health, product safety, and financial fraud. At the same time, the Court is also hampering agencies—to which Congress has entrusted the implementation of particular programs—from responding to new technical challenges and to evolving democratic preferences. The Court’s lack of democratic accountability, as well as policy and technical expertise, make it a poor institution to impose these barriers. Nonetheless, through ongoing doctrinal changes, the Court is exercising its power to send both Congress and the agencies, time and time again, back to the drawing board, with damaging economic and social welfare consequences.

[*] Joseph L. Sax Collegiate Professor of Law, University of Michigan Law School. Thanks to the Cook Fund at the University of Michigan Law School for research support and to Dan Deacon for valuable comments. Thank you as well to the staff of the Harvard Journal on Legislation for their excellent editing of this essay.

Related